From GNU Radio
Revision as of 01:36, 8 March 2017 by Mbr0wn (talk | contribs) (Imported from Redmine)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Dev Call February 2017[edit]


  • 3.8
  • GRCon
  • hackthemachine, dyspan, fosdem videos

GNU Radio Project Call January 2017[edit]


  • GSoC -- Ideas
  • Releases
  • Plan for Python3 support


  • FOSDEM: Be there! It's free!
  • GSoC: Please post ideas to http://gnuradio.org/redmine/projects/gnuradio/wiki/GSoCIdeas, even if they're still raw
  • Releases: We're in code freeze, bugfix PRs still get accepted
    •, 3.7.11 are coming
  • Python 3: Will be coming this year!
    • Dual-version, single codebase
    • Involves a lot of changes
    • Shoutouts to Douglas Anderson who's done a lot of work so far
    • GRC involves biggest changes. We need to rip out Cheetah though.
    • Help is highly appreciated!


- GRCon 17
- DARPA SDR Hackfest 2017
- FOSDEM 2017
- DARPA Battle of the ModRecs
- IEEE DySPAN 2017

- 2016 Year in Review! https://docs.google.com/a/gnuradio.org/presentation/d/13wbtESpeuJPHljzn8aef17tl5xZ-NZzc8q9HEb_RXUM/edit?usp=sharing

- https://www.openhub.net/p/real-gnuradio

Dev Call October 2016[edit]

- GRCon 16 review
- GRCon 17 not-quite-an-announcement
- Releases (pybombs)
- Livesdr update, Redmine to github transition
- Announcements: FOSDEM, GSoC


GRCon '16[edit]

- Videos editing is finished, and should be uploaded pretty soon (waiting on videographers)
- Attendee survey is now closed, but response rate was pretty good (close to 100 respondents, that's nearly 50%). Overall feedback very positive (over 90% gave 4 or 5 out of 5). Plus, good feedback on how to improve.
- Already people are asking for sponsor slots for next year! We haven't locked down details on sponsors yet, though.

GRCon '17[edit]

- It's time to start planning it.
- Ideas on where to do it: Come back to DC, stay in Boulder (because everyone loved it), San Diego, Canada (maybe Montreal?). It's all in the air, and suggestions are welcome! Most important deciding factor is if there's people on the ground helping us put together the conference.

Dev Call August 2016[edit]


  • GRCon!
    • Everything will be recorded
    • Keynote speakers
    • Review schedule
    • Ticket status
    • Student's reg
    • Developer's summit
    • Posters
  • GSoC
    • Status
  • Stack overflow!


Dev Call July 21st 2016[edit]

Release announcements[edit]

  • Bugfix release
  • 3.7.10


  • Be there!


  • Still ongoing!

Working Groups[edit]


  • WG is updating a lot of the code to use more modern libraries
  • Anyone who can code Python and has time here and there to fix warning, please step up! Contact Sebastian or the mailing list.


  • GRCon
  • SoC
  • Educators
  • releases



- Receive a lot of abstracts!
- Registrations 1/3 sold out, ahead of last year's at the same time
- Friday night: Great Scott Gadgets hosts


- SOCIS is live!


There's a new list. Contact Martin


3.7.10 coming out soon, please review packet3 branch.

GRC Developers Call June 2016[edit]

(Wednesday, June 8, 8pm CEST - an hour later than usual)


  • GUI bindings update (branch: gtk3)
    • progress updates (Seth: Actions, Jarriel: Canvas (cairo), Sebastian (PropsDialog), Glenn (Dialogs))
    • discuss open issues (e.g. custom styling of widgets now seems to require some CSS)
    • identify issues that need to be fixed before we call this stable enough and merge into next_grcwg
  • YAML/Mako block wrapper format
    • progress update


(after the call)

Developer Call May 2016[edit]

- GRCon 2016
- GSoC


  • See the website. Lots of people are registering!


  • Looks like we will soon have a fourth student through SOCIS


  • Lots of movement
  • New variable panel
  • 3.8: Moving to GTK3, mako, YAML


  • Please help test

GRC Developers Call April 2016[edit]

(April 28, 2016, 7pm CEST)


  • GUI bindings update
    • changes we need no make to get rid of deprecated stuff (Actions, draw code)
    • split-up work
  • YAML/Mako block wrapper format
    • this is a rare occasion: we get to define a completely new format and remove some the annoyances of the XML format.
    • Here is a draft from a while back: https://gist.github.com/skoslowski/d7016c703344cc53df2f . This can serve as basis for the discussion.
    • this uses pure python with a special tag outside of the make and import parts (that have to be templates). Good, or stick with the templated valued we have now?
    • since we switch template engine everything breaks for everyone. simple (most) wrappers should be converted automatically but not all (qtgui).


  • recent changes to the code base
    • we now have separate gui and core (formally 'base' and 'python')
      • removed triangle-like class hierarchy
      • enable easier transition to other gui frameworks
      • now we can add tests
    • code refactoring
      • started with core.Platform and core.Flowgraph
      • goal: replace old/clunky/slow/un-pythonic parts (but don't break stuff)
      • next big chunks will be core.Block and core.Param
    • dev branches
      • with the refactored package structure maint and master have diverged
      • preexisting bugs fixes must be manually merged from maint to master
      • bugs from the refactoring only go in master
  • features for 3.7.10
    • variable editor (Sebastian/Seth -> PR -> review)
    • block alignment tools (Glenn --> review)
    • after that we focus on next!
  • gui-bindings (gtk3) (GR 3.8)
    • auto-converted code runs w/o flow graph canvas, but has lot of deprecation warnings
    • Seth: tackle those related to the Actions and the ActionHandler
    • Sebastian: clean-up initial cairo-based canvas code and hand work over to Jarriel
    • as soon as we have a functional version will push it to grcwg/next_grcwg
  • Python 3 (GR 3.8)
    • GRC code base shall become dual-version
    • start using python-six on next_grcwg and any refactoring
  • block wrapper format changes (YAML/MAKO) (GR 3.8)
    • the proposed structure was mostly accepted
    • optional, non-mako based make, import and callbacks
    • Sebastian will write-up a spec
    • we need a auto-conversion tool (Sebastian has a prototype)
      • XML -> YAM is easy
      • Cheetah --> Mako is hard/impossible

Developer Call April 21 2016[edit]


  • Website
  • GSoC
  • GRCon

Youtube link:



- Development mostly done, some minor things not working but that'll get sorted out soon
- Thanks to Marcus and Nathan for making pics and logos
- Next step, more content -- ping us if you have anything interesting


- Registration open
- Deadlines not yet finalized, whatever's up now will be pushed forward
- There will be paper proceedings
- Booth space is running out, so are some sponsor slots, if you want to sponsor, please get in touch with Ben soon

GSoC + SOCIS[edit]

- Nothing new; tomorrow, GSoC stuff will get announced


- GRC call next week (tbd, but probably same time as GR dev call)
- Lot of refactoring happening

March 2016 Dev Call[edit]


  • Org Changes
  • GRCon
  • Website
  • Call re-org
  • GSoC
  • Release Announcements


Org Changes[edit]

Ben new org leader. See the announcement on the mailing list.


See announcement website.

New Website[edit]

Split: Nice frontpage for non-technical stuff (blogs, conference, community, news), based on Wordpress. Redmine gets replaced with gitlab (Issues, Milestone planning, Feature requests...). Roll-out will be soon. If you want to help or just proofread, please do!

Call re-org[edit]

Ben will do more community-related calls, but there'll be more technical calls on the side.

Release schedules[edit] coming up soon (bugfix release). Due next Friday (April 8th).

Issue Tracker[edit]

Is filling up. We need a new issue tracker manager.

Dev Call February 2016[edit]


Dev Call December 2015[edit]


  • FOSDEM Information
  • GSoC Call for Ideas
  • Hackfest updates
  • Release information (v3.7.9)
  • Working Group reporting

Video: https://www.youtube.com/watch?v=fdiKhThmV7s



Hackfest Berlin[edit]

  • Pre-FOSDEM
  • Please contact Martin if you want to come!

Summer of Code[edit]

  • Still need lots of good ideas!


  • Next GNU Radio release 3.7.9
  • Will be released Wednesday

Project Stats[edit]

  • See attachment (attachment:gr2015.pdf)

Working Groups[edit]

  • There will be a VOLK release

Dev Call November 2015[edit]


  • Future Hackfest discussions
    • Berlin
    • East coast
  • SDR Day at WinnComm'16
  • VOLK in-place / const
  • Accepting other copyrighted code
  • v3.7.9
  • Working Group reporting

Video: https://www.youtube.com/watch?v=B0jN_8JRL_g


  • WinnComm'16
    • March 15-17
    • We will have an SDR Day at this year's conference
    • Tom helping to organize
    • Will include talks and demos
  • Summer of Code:
    • We still need lots of good ideas for projects
    • Want a good list by the end of the year
    • Will vet, polish, promote in January/February
  • Embedded Linux Conference (ELC), April 4-6 in San Diego
  • VOLK inplace operations
    • Do we / Don't we support it? How do we or don't we?
    • QA would be hard to check this sort of this
      • maybe test inplace using a sort of a puppet that does memcpy(out, in, size); kernel(out, out);?
    • maybe just mark the allowed args with attributes so you an detect them when generating the QA code.
    • Still working on solution
  • Non FSF-Copyright code contributions to VOLK
    • We do this in GR on rare occasions and don't like to
      • makes it hard to manage, fix, update, etc.
    • VOLK different - more easily defined walls (kernel and proto-kernel level)
    • Tom's recommendation: have the VOLK project with two repos
      • main repo is all FSF code
      • general/third-party/universal repo for other code
      • All still GPLv3
    • Another topic to settle on in the next VOLK call
  • WG


'* Incorporating a few new kernels; some bugs being addressed

'* Looking at the next WG call.

' Co-Processor

'* Progress on custom buffer work

'* Testing with OpenCL

'* Possible bug found; checking into this

'* Want more testers / users!

' Embedded

'* "* | Crofton|work is depressed and lonely in embedded space"

'* Lots of people using embedded systems -- need more people talking and contributing.

'* E310 and OE builds are being kept up-to-date

'* http://gnuradio.org/redmine/projects/gnuradio/wiki/GNURadio_example_status


'* Chugging along nicely

'* Lots of new features, growing contributors

'* Lots of nice new features in 3.7.9.

' Community

'* FOSDEM dev room

'* Possible collaborators to build better educational materials

'* PyBOMBS v2 is coming along nicely -- about ready for beta testers!

' Hope to release by end of year

' pip install


'* Going to try to have regular calls

'* https://github.com/marcusmueller/gnuradio/tree/runtime_add_message_port_wrappers

'* http://gnuradio.org/redmine/projects/gnuradio/wiki/FASSWG_msg_actions

GRC Working Group Call November 2015[edit]

(Start 8pm CET)


  • state and progress of current PRs
  • discuss GSoC project ideas
  • efforts to ease starting GRC development
  • QT and the road to py3k =)


  • no notes for a sad one-man show =)

Dev Call October 2015[edit]


  • Releases: 3.7.9,, 3.8
  • GSoC planning
    • Discuss project ideas -- do we have an overall theme?
    • Discuss updating deliverable requirements
  • Pre-FOSDEM hackfest
  • Working Group reporting

Video: https://www.youtube.com/watch?v=mi9BJbV0XSo


  • releases
    • -- just an accumulation of minor bugs
    • need help with underlying speed of LDPC coders/encoders
    • main issue with 3.8 is python3 support -- means mako for grc files
      • tnt suggests m4ri
      • trondeau tnt: definitely a candidate
      • trondeau you've recommended it before and I've heard others mention it as well
    • https://github.com/gnuradio/pybombs2
    • we're working to make sure that using yaml/mako/etc for grc that we still support current XML for compat with OOTs
    • pybombs2 repo https://github.com/gnuradio/pybombs2
  • PyBOMBS2
    • period of deprecation of pybombs. Work towards complete switch of pybombs2 to pybombs at the same time as the 3.8 release (some time next year).
  • GSoC
  • pre-FOSDEM hackfest and FOSDEM
    • still waiting to hear if our sdr room has been accepted (we have no reason to believe it won't)
    • pre-FOSDEM hackfest happening at legendary C-base in Berlin
    • pre-FOSDEM hackfest monday-thursday
    • hackfest in Berlin. Mon - Thurs before FOSDEM; Friday travel day
    • hacfest space: https://en.wikipedia.org/wiki/C-base
  • working group reports
    • Community: news to come on more educational support
    • GRC: Seth will be helping GRC much more closely with Sebastian
    • embedded python blocks - open editor; build a python block; auto-generate block
    • autogenerate missing hier_blocks

Dev Call September 2015[edit]


  • GRCon15 Wrap-up
  • Review of GRCon15 working group discussions

Recording: http://youtu.be/my_EBYgO8cE


  • GRCon15 wrap-up:
    • We all thought it went smashingly well
    • Big question about where to host next year:
      • DC again, or Boulder, CO?
    • Do we want smaller, regional meetups throughout the year?
      • Short-term: keep finding conferences to attach an event to
      • Longer-term: be more proactive about creating these opportunities
      • Increased participation from other countries
    • Videos are almost now up on the GRCon15 webpage: http://www.trondeau.com/grcon15-presentations/
  • WG
    • Community (Martin)
      • Looking into more areas for Task Forces
      • First one: More tutorials to extend and go more in-depth into topics
    • VOLK (Nathan)
      • Call for tutorials for using VOLK; specifically, using/calling in OOT projects.
      • Work on reproducible builds -- already worked on at GRCon15 Hackfest
      • Talked of building internal library to wrap small chunks that will be reused a lot, like complex multiply
    • GRC (via Martin)
      • Identified a few people to help on certain tasks. Looking to expand and manage a number of people helping
      • Some more discussion QT version
    • Co-Processor (Doug)
      • Discussed callback API for device drivers to control access to DMAing data around
      • Custom buffer work is nearing completion. People seem happy, but looking for a good example (OpenCL)
    • FASS (via Johnathan)
      • Talk about expanding use of Lambda blocks for easier development times
      • Method to move concept of a block's work into a way that makes it easier to use in a stream, tagged stream, or message passing model
      • Simplify the class structure to reduce programming demands of the very simple blocks
  • Mailing list
    • Corgan raised issue of ruby-forums being a difficult way to interact on our mailing list. They have been unsubscribed, but we're still wondering how to best interact with everyone that has questions.
    • Please sign up for the mailing list! http://gnuradio.org/redmine/projects/gnuradio/wiki/MailingLists
    • I would like to see more time spent on answering StackOverflow questions to hopefully offload a lot of questions to proper answers there.
  • Python Blocks
    • We have discussed moving all Python hier_blocks inside GNU Radio to C++
    • They would be identical in structure and config as the current Python, but now also available in C++
    • Would help with documentation issues (in GRC and in the manual both)
    • We will make a new spreadsheet to list all of the Python blocks for record keeping.
    • Waiting for the FOSDEM call for participation
    • We plan to participate again and think it's likely we'll have a room
    • So be prepared to submit!
    • Also, we will have a Hackfest in Europe somewhere the week (Mon - Thurs) before FOSDEM
    • Hopefully more info on both by next dev call

Dev Call July 2015[edit]


  • GRCON15 Prep
  • Release 3.7.8 timeline
  • Working Group Reports


Dev Call June 2015[edit]


  • Review: NEWSDR & W@VT
    • We've been busy at conferences, presented lots of stuff and organized a hack session at NEWSDR
  • Android Support
    • Tom has been busy working on Android + GNU Radio, it's now ready for general testing and we hope people start using our template codes / SDKs for writing some killer apps to showcase the capabilities of GNU Radio on Android
  • Working Groups
    • Most working groups have been quiet
    • All will present at GRCon on what they're plans for the future are
  • GRCon
    • Agenda is up! Check it out, and do come to DC

Video: https://www.youtube.com/watch?v=d4bd4sKC_eE

Dev Call May 2015[edit]


  • Announce NEWSDR + W@VT
  • Working Groups
  • GRCon


NEWSDR happening now @ WPI along with GNU Radio hacking workshop. Thanks to Neel for organizing!

Next week Wireless @ VT will have a large GNU Radio presence.
Phil invited GNU Radio-interested people over to his place for a cook-out.

We can announce localized GNU Radio meetups on the dev call.

  • Denver area with Mike Ossman
  • Cyberspectrum in the Bay Area doing SDR related things. Next meeting at Ettus.

KIT-IEEE Signals Intelligence challenge is ongoing. http://www-ieee.etec.uni-karlsruhe.de/?page_id=147

GNU Radio Con starts August 25. CFP is over-- lots of good content so far. Acceptance
goes out next week.

Not too late to sign up to attend. http://www.trondeau.com/gnu-radio-conference-2015/
for info

Working Groups[edit]

Identify goals and milestones for individual working groups. WG leaders should have a
good idea of things to accomplish soon, especially by the conference.

Working Group Updates[edit]

User experience / community (martin)[edit]

  • CGRAN task-force is incrementally updating CGRAN.
  • Effort to rewrite pybombs is still ongoing. Proper prefix support is working, but still needs work on SDK integration
  • Any volunteers to help with python code to refactor pybombs (hopefully new version by grcon) are welcome

VOLK (nathan)[edit]

  • lots of PRs and new contributors to VOLK doing maintenance (yay!)
  • AVX2/FMA support coming today
  • most groups don't upstream their kernels, but we would like to encourage upstream contributions if they are not application specific
  • Look for a call in early June
  • Successful on Android

grc (Sebastian)[edit]

  • Quiet
  • New feature on bypassing blocks-- useful for quickly bypassing blocks. merge coming


  • Contributors definitely welcome (good way to get attention and name out there for students)

Coprocessors (Doug)[edit]

  • Not much to report
  • Working on custom buffers
  • Allows blocks to create their own special buffers for hardware gains
  • document @ CoProcBufferDesignDoc

embedded (phil)[edit]


  • mostly inactive
  • look for a discussion at wireless@vt

Workin group leaders meeting upcoming-- look for an email.

April 2015 Dev Call[edit]


GRC Working Group Call April 2015[edit]


  • future block wrapper file format


Current situation:

  • Cheetah is used to template both, expr for design time and the actual make template
  • Cheetah is dead and not py3k compatible
  • to replace Cheetah without breaking existing files we need a new format (or format indicator at least).
  • both formats will be supported alongside each other for a while.

New Format:

  • YAML with mako in templates, else pure python
  • harmonize file formats in the GR ecosystem (pybombs recipes, GRC block wrappers, MANIFEST, ....)
  • Easing the transition: We need a converter (warn / double-check with / ask user for Cheetah stuff)
  • detailed documentation of new format
  • new dependency required: PyYAML
  • alternative: pure-python all the way (GRCQT)
    • powerful, but possibly opens up too much GRC internals.
    • makes changes to GRC hard
    • most blocks don't need that
    • maybe use in in-tree blocks, mark as unsupported, provisional API
  • custom parameter widgets as GRC plug-ins. To allow integration with e.g. gr-filter-design


  • Sebastian writes mock-up of YAML/mako format

Dev Call March 2015[edit]



Summer of Code (Martin)[edit]

  • Rejected from google
  • Applying again next year.
  • analyzing how to improve
  • Accepted to SoC in space by ESA
  • program is different feel
  • only 1 student spot available
  • coding period is mostly similar to GSoC

Hackfest (Tom)[edit]

  • gr-mac & pmt on embedded systems
  • pyqwt - figure out plan to fix long-running issues
  • CGRAN website & front page for gnuradio.org

No working groups, but hackfest should revitalize them

CGRAN (Nathan)[edit]

Android (Tom)[edit]

  • Porting gnuradio to Android
  • Soft-release at FOSDEM
  • Released as seperate Android branch from Tom for now
  • wiki page for instructions
  • sink to move data between Java UI and gnuradio via grand
  • alpha release. testers/devs familiar with gnuradio and android
  • instructions are not perfect. feedback welcome

ControlPort (Tom)[edit]

  • ControlPort was not removed; ICE (the only working backend) was removed
  • We've wanted it back, especially with all embedded work going on
  • At GRcon Thrift looked like the best chance at getting a functioning backend working
  • Thrift support is very close to being mergable in to master. One problem remaining
  • If you run QA code 100s of times in a row it locks up (probably race condition, thrift bug?)
  • Thrift is a young project that will likely move quickly.
  • 0.9.2 is min version, which is not yet released by most distros
  • revive controlport-monitor and performance monitor
  • GNU Radio ControlPort API is the same, so existing server-side should still work with new backend
  • Client-side tools probably need to be changed, but there is a new python abstraction layer in gnuradio
  • Unsure if release comes at 3.8 or sooner (discuss at hackfest)

v3.7.7 Release (Tom)[edit]

  • Extracting VOLK as its own library that Nathan is maintaining
  • Still a GNU Radio project
  • Working to make sure everything is in place for smooth transition
  • First day of hackfest the switch a submodule will occur
  • The rest of the week will be a stress test (everybody hopefully bases new work off of this master)
  • April 6 code freeze
  • One week where no new features get merged in to master
  • Focus on bug fixes
  • April 13 release v3.7.7 and v3.7.6.2 (maintenance release)

Dev Call Feb 2015[edit]


  • FOSDEM Recap
  • Hackfest Recap
  • GSoC 2015
  • Issue Tracker issues
  • CGRAN Progress
  • WGs
  • livesdr DVD




  • Big success
  • Videos will be available as soon as FOSDEM uploads them
  • Tom's stuff is already available
  • Massive demand!
  • Please, all authors: Upload your slides to Pentabarf

Delft Hackfest[edit]



  • We have redefined the issue states (see contributor's guide)
  • If you have bugs assigned to you, please fix them or unclaim the issue
  • Maybe we bring back the "issue tracker manager" role


  • We have a new task force remaking CGRAN
  • We've picked our technology for the new website
  • It'll be there soon (order of weeks)


  • Has now been put into its own repo
  • Not yet submoduled in GNU Radio (but that'll happen soon)
  • Now has documentation


  • We held a CoProc WG call on Feb 11 - notes are online at: http://gnuradio.org/redmine/projects/gnuradio/wiki/Call20150211
  • There was a fair amount of progress made on issues #729 and #730 made last Friday at a mini-hackfest hosted by Tim O'Shea at Virginia Tech/Arlington
    • Tom submitted a pull request that resolves #729
    • Doug picked up working on #730 from Jonathan Corgan's GitHub branch, and is making some progress (thanks to snow days in DC and a long weekend!)


  • QT port is coming along
  • WX/QT blocks vs. options is better now
  • Handling of GRC-generated hier blocks needs reworking
  • S.K. would rather switch to regular imports instead of 'execute file'


  • Wiki page is updated


The PyBOMB squad: CGRAN etc.[edit]


  • Rule 0: CGRAN must never have any manual intervention
  • We'll use Ravi's stack (i.e. Pelican)
  • Tables
    • Sorting (by name)
    • Filtering / Character Search must work for: Name + Tags
    • Bootstrap Tables does the job we need it to do
  • Content
    • No landing page (goes directly to the table)
    • A page for every module
    • Anything static we just link to the wiki
  • Theming
    • We'll stick with the GNU Radio colours
    • Base the logo off the septagon



  • #730 - Custom buffers
    • Design document desirable before moving forward: CoProcBufferDesignDoc
      • Use-cases: what should be made possible which isn't possible now
      • Identify what custom allocators (OOT co-processor support) must/may/should implement

Misc notes[edit]

Dev Call 20 January 2015 (VOLK WG)[edit]

The call will take place at 3:00 CST (21:00 UTC) on a G+ hangout. If you want to participate inform Nathan or hang out on IRC just before.

Dev Call January 2015[edit]





FOSDEM + Hackfest[edit]

  • please fill out the hackfest page

GRC Unit Testing[edit]

  • LINK: https://fosdem.org/2015/schedule/track/testing_and_automation/
  • Idea: We need a set of 'unit-test' like .grc files that we can run grcc over automatically, and maybe run automatically (mbr0wn,
  • 18:27:45)
  • Also, we want as many unit tests as possible for GRC
  • Auto-clicking or whatever -> call for action
  • Will be tackled at the hackfest

Blocking in blocks[edit]

  • The actual solution is for a block to signal that work() shouldn't be called until there's a message
  • Blocking is not really an issue, but it sets a bad example
  • ACTION: tnt will code the temp fix


  • Building the DVD is automated now
  • jcorgan is asking if there's interest in publishing the source for the DVD
  • ACTION: The codes will go up, mbr0wn & jcorgan will clean it up

VOLK WG[edit]

Community Stuff: CGRAN![edit]


Packet Radio[edit]

  • trondeau has cool stuff running


  • Nothing new


  • Sebastian is AWOL :)
  • But the QT port is going strong
  • Lots of commits there

Release Schedule[edit]

Meeting ended at 18:59:44 UTC.

PyBOMB Squad - Inaugural Call[edit]

Video: https://www.youtube.com/watch?v=zs-Xoq_Sw18

Action Items:

- Chris: Write manifest files
- Nathan: integrate grist onto gnuradio website, come up with CGRAN 2.0 alpha (bare essentials of current GRIST)
- Ravi, Alfredo, George, Neel volunteers to help with grist (Nathan is head)
- Python Pelican could be used instead of Jekyll
- Find people who are good at designing web!
- For inspiration layers.openembedded.org
- Martin: Add archiving capabilities to PyBOMBS
- George + Johnathan: Coordinate about updating cgran.org

Future way to add OOTs to PyBOMBS:
- We provide an easy way to create a bare recipe file from an OOT, using the manifest etc.
- The author provides the recipe to us via pull request, as before
- We use this to also moderate what gets into PyBOMBS
- The recipe can be maintained by us once it's in PyBOMBS

Dev Call December 2014[edit]


  • Stats of 2014
  • Delft Hackfest and FOSDEM
  • Release Schedule
  • Working Groups



See attachment.

More comments:

  • Chuck Swig the first to draw blood in 2014
  • 2014 yielded 6 releases; we interpret this as higher stability
  • OpenBox says "well-commented source code"
  • multiple Hackfests 2014
  • GRCon '14 was the biggiste converence so far with around 160 attendees
  • Working Groups made great progress
  • GSoC: 5 Students, 4 passed; Tom points out that integration in the Community will be a crucial aspect for acceptance in the coming years
  • Dev Calls: held 11 dev calls (monthly aside from GRCon month)
  • GNU Radio progress based on experience of people actually using GR 3.7



Release Schedule[edit]

Working Groups[edit]

  • Embedded: stuff
  • https://www.dropbox.com/home/GNURadio-embedded
  • Embedded: images for zc702 and zedboards available ^
  • VOLK: Next call ~jan
  • VOLK: submodule idea being pursued
  • VOLK: Work fix-point kernels
  • FASS: tim` is AWOL
  • ...but Tom is improving packet-based comms
  • Coproc: will schedule another call soon. Opened a number of issues related to the api for coprocs that need continued work/attention
  • GRC: continued work on GRC-QT port
  • GRC: connection domains; better handling of how connections are defined underneath and represented on the canvas
  • GRC: first use case is RFNoC

Dev Call November 2014[edit]


  • Hackfest
  • Digital modulation and packet structure review
  • Release Schedule
  • Working Groups

Watch at:


  • Hackfest
    • At TU Delft
    • Invite-Only
  • Digital Modulations (mbr0wn, 18:13:28)
    • We're trying to consolidate all our digital modulation / packet support
    • Want to make it all more coherent
    • For 3.8
    • trondeau built a chat app over-the-air for the Guided Tutorials
    • Clean implementation of modulation, packetization, packet sync etc.
    • ...FEC too
    • We'll have a design document up for review by the community
    • ACTION: We need feedback from everyone to make sure we have all the important things covered
  • Release Schedule
    • We don't have anything major going on, but we might have a release anyway
    • Maybe 3.7.6 before Christmas
    • GNU Radio LiveDVD will definitely come out before that
    • Debian Jessie froze with 3.7.5, stable backports will resume after Jessie release
  • Working Groups
    • Accelerators: Call last month Currently ongoing: Custom Buffer stuff Another call: Probably between Thanksgiving & Christmas
    • http://stats.gnuradio.org/covtable
    • VOLK: Call last Monday, new NEON kernels, build system improvements for submoduling VOLK
    • Embedded: Nothing dramatic, new test images for zynq-based boards, next step is to update wiki page to get people started
    • Embedded: Lowering entry barrier by publishing new SDKs and improving documentation
    • GRC: QT-Port is going on[[File:|]]!
    • GRC: Generator rewrite
    • GRC: Fixes hier block issues
    • Tutorials: Going out of beta soon
    • Want to push the stack overflow thing

GRC Working Group Call November 2014[edit]


  • QT port


Dev Call 17 November 2014 (VOLK WG)[edit]

The call will take place at 3:30 CST (21:30 UTC) on a G+ hangout. If you want to participate inform Nathan or hang out on IRC just before.

  • Embedded build status
  • Kernel/ISA coverage
  • Git subtrees / breakout

Co-Processors WG Call - October 29, 2014[edit]


Dev Call October 2014[edit]

  • GRCon14: Wrap-up
  • GRCon15: a few quick words
  • Hackfest planning
    • TU Delft, end of January
    • FOSDEM
    • Ettus, others?
  • Working Group updates
    • Reports from GRCon14
  • Example data files
    • Work with Ken at CU Boulder for snapshots
  • Release


GRC Working Group Call October 2014[edit]


  • QT port:
    • organize work
    • maybe figure out some milestones
  • Flowgraph model rewrite:
    • new block API
    • Cheetah replacement


  • QT port:
    • Andy will merge his work into the current qtport branch and we all continue from there
    • Seth concentrates on extensions to the main window and its widgets
    • Andy continues his work on the flow-graph display
    • Sebastian does a rewrite of the flow-graph model

Dev Call August 2014[edit]

  • GRCon14: last dev call before conference
  • GSoC: wrapping up this week
  • New Guided_Tutorials announced and in beta
  • Recent updates to the code:
    • Fix for parallel build issue - under test
    • Some outstanding issues in Coverity
    • GRC updates, including thoughts on versioning files
    • VOLK NEON support
    • Improvements to QTGUI: small bug fixes; all examples now in QT
  • Updates on versions
    • Potential for 3.7.5 release for GRCon
    • 3.8: still far off; collecting ideas and identifying new versions of deps
  • Announcing the new Working Group
  • ...and all the other WGs report

Co-Processor WG Call[edit]

August 12, 2014


> * Additional discussion to follow-up the contiguous memory requirements discussed in that thread

  • Lots of people seem interested in co-processor support (e.g. the Zynq, Keystone-type SoC's, also GPGPU via OpenCL) - however other than encouraging various efforts, there isn't much else to report on this. Hopefully GrCon will bring together many of the people working in this area.

Dev Call July 2014[edit]


  • GSoC students: Brief update
  • WG Leaders: Updates
  • To-Dos for 3.7.5

Youtube link: http://youtu.be/uGyJIp2WZig


  • GSoC-Students (mbr0wn, 17:03:17)
    • Gsoc updates: http://gnuradio.org/redmine/news/39
    • Marcus: Measurement Toolbox (mbr0wn, 17:03:31)
    • Jan: Speedup gr-trellis (mbr0wn, 17:09:15)
    • Jan: Fully working trellis decoder as OOT, moved away from gr-trellis (funkylab, 17:11:03)
    • Jan: Compared to generic kernel, SSE4.1 kernels are three to four times faster (funkylab, 17:13:25)
    • Alfredo: Embedding GNU Radio for the Keystone board, and dissecting the GR runtime (funkylab, 17:14:24)
    • Stefan: Radar Toolbox (funkylab, 17:20:00)
    • Stefan: implemented Echo timer to test raw rf roundtrip latency (funkylab, 17:20:24)
    • Stefan: Videos on Blog! https://grradar.wordpress.com/ (funkylab, 17:20:54)
    • Stefan: GUI blocks, using QT, for Radar visualization (funkylab, 17:21:09)
    • Abishek: VOLK improvements for OFDM frame sync (funkylab, 17:26:25)
  • WG Updates (mbr0wn, 17:28:28)
    • VOLK: More Protokernels (mbr0wn, 17:28:34)
    • VOLK: Had a call recently (mbr0wn, 17:29:06)
    • New volk kernels from Abishek's gsoc work: http://abhishekbhowmick.blogspot.in/ (trondeau, 17:29:54)
    • VOLK: Upcoming: Cleanup, then NEON (mbr0wn, 17:29:57)
    • GRC WG: recent call, notes in the wiki (funkylab, 17:30:37)
    • GRC: versioning file versioning planned (funkylab, 17:31:06)
    • GRC: wiki documentation is upcoming work (funkylab, 17:36:03)
    • Embedded stuff: http://gnuradio.org/redmine/projects/gnuradio/wiki/Embedded (trondeau, 17:38:06)
    • Updates to the daisy brnach (Crofton|work, 17:38:41)
    • Hoepfully ICE gets into daisy so we can do control port there (Crofton|work, 17:38:57)
    • If you are tracking master, expect pain :) (Crofton|work, 17:39:12)
    • Buy this book : http://www.amazon.com/gp/product/1783282339/ (Crofton|work, 17:40:31)
    • OFDM: moving to PDUs (Tim O'Shea) (funkylab, 17:41:15)
    • pdu_to_tagged_stream has to CPU overconsumption problem (funkylab, 17:43:14)
    • Planning new WG for digital Modem (PDUs etc, led by Tim) (funkylab, 17:44:08)
    • Tutorials: in beta-testing stage (muniza, 17:44:55)
  • 3.7.5 release plans
  • feature list is not comprehensive (funkylab, 17:45:58)
    • keep message format compatible, so defining a message format makes sense (funkylab, 17:57:29)
    • like pdus (tim`, 17:57:43)
    • standardized PMT formats: put that into the doxygen manual (funkylab, 18:02:00)
  • GRCon

Meeting ended at 18:03:25 UTC.

GRC Working Group Call July 2014[edit]


  • new variable block behaviour
  • versioning grc files
  • documentation
    • of blocks (user vs. api documentation)
    • of GRC itself


  • Version of GRC with a XML comment for backwards compatibility
    • Seth
  • extended variable definitions
    • no more dummy param value
    • instead we use a new tag <var_value> with supports Cheetah templates to get a design time value (thinks of config, filter design, or constellation objects)
    • Sebastian
  • documentation
    • the GRC wiki needs an overhaul
    • Block documentation should also be improved
    • Maybe finally integrate source/doxygen access to buttons
  • message ports
    • we need a <hide> tag for (message) ports
    • some blocks are have ports that are rarely used
    • user realizable blocks would be nice too
    • Sebastian
  • port graphics in general:
    • option to show/hide port labels
    • maybe truncate longer names
  • proposal (Seth): event messages
    • similar to signals and slots
    • no visible connections
    • basically melt virtual sinks/sources into these ports
    • or add GR support directly for such messages



  • Make sure we have a similar understand of what the problem is

- How does GNU Radio currently handle buffers?
- What kind of buffer management do we need for coprocessors?

  • Current projects status (GSoC, gr-liquiddsp, etc.)

- GSoC: work using get_user_pages to perform direct IO without copies to kernel space

  • Changes needed in mainline GNURadio (e.g. buffer management)

- GSoC: get_user_pages requires that we pass the write pointer and buffer size to blocks

  • Unified accelerator API

- GSoC: get_user_pages should support a large number of coprocessors since it operates on pointers

For history GRCon13_Coprocessors_Discussion

  • Notes:
    • Look at trondeau's in-place buffer management branch on github to understand how GR buffer management works (https://github.com/trondeau/gnuradio/commits/sched_inplace)
    • Circular buffer's: mandate all coprocessor's drivers implement double-buffer to support circular buffer structure?
      • Keystone 2: Writing the kernel driver allows the use of get_user_pages, GR enforcement of page alignment helps
    • Ingress/Egress blocks vs. monolithic blocks
      • 'Closely' coupled systems (shared memory: e.g. ARM SoC + DSP's, Zynq, etc.) vs. 'loosely' coupled systems (non-shared memory)
    • How to draw more interested users into the Working Group community

Developer's Call June 19th[edit]



  • All coming along well
  • Students are all trying to come to GRCon!
  • Check out their blogs and repos!
    • Action Item: Students, please join the call next month and talk about your work!


  • Looking good, coming along
  • Lots of sponsors!
  • Working group leaders: Be ready for some action items!
    • Action Item: We'll set up a call with GRCon organizers and

Working groups[edit]


  • All fantastic

DSPs / Accelerators[edit]

  • We finally have a new chair: Doug Geiger!


  • We still need more active participants!
  • We need another call. Action Items: Make a call next month.


  • In the review stage, very close to finishing!


  • We have a new wiki page Embedded that catalogues our current progress.



  • Dealing with new (to VOLK) architectures
  • NEON ARMv7/ARMv8
  • FMA3 (/avx2)
  • Work to do
  • resolve u_generic and a_generic calls
  • a_generic
  • volk_16i_convert_8i
  • volk_16i_s32f_convert_32f
  • volk_16u_byteswap
  • volk_32fc_conjugate_32fc
  • volk_32fc_deinterleave_64f_x2
  • volk_32fc_magnitude_32f
  • volk_32fc_magnitude_squared_32f
  • volk_32f_convert_64f
  • volk_32fc_s32fc_multiply_32fc
  • volk_32fc_x2_conjugate_dot_prod_32fc
  • volk_32fc_x2_dot_prod_32fc
  • volk_32fc_x2_dot_prod_32fc
  • volk_32fc_x2_multiply_32fc
  • volk_32fc_x2_multiply_conjugate_32fc
  • volk_32f_s32f_convert_16i
  • volk_32f_s32f_convert_32i
  • volk_32f_s32f_convert_8i
  • volk_32f_s32f_multiply_32f
  • volk_32f_x2_add_32f
  • volk_32f_x2_dot_prod_32f
  • volk_32f_x2_multiply_32f
  • volk_32i_s32f_convert_32f
  • volk_32u_byteswap
  • volk_64f_convert_32f
  • volk_64u_byteswap
  • volk_8i_convert_16i
  • volk_8i_s32f_convert_32f
  • u_generic
  • volk_32fc_index_max_16u
  • volk_32f_index_max_16u
  • volk_8u_conv_k7_r2puppet_8u
  • volk_8u_conv_k7_r2puppet_8u
  • volk_8u_x4_conv_k7_r2_8u
  • Same thing, fill in the gaps with unaligned/aligned proto kernels. Examples
  • grep -H a_sse3 * | cut -d: -f1 | xargs grep -L u_sse3
  • volk_16ic_magnitude_16i.h
  • volk_16ic_s32f_magnitude_32f.h
  • volk_32fc_index_max_16u.h
  • volk_32fc_s32f_magnitude_16i.h
  • volk_32fc_s32f_power_spectrum_32f.h
  • volk_32fc_s32f_x2_power_spectral_density_32f.h
  • volk_32fc_x2_s32f_square_dist_scalar_mult_32f.h
  • volk_32fc_x2_square_dist_32f.h
  • volk_32f_x3_sum_of_poly_32f.h
  • grep -H u_sse3 * | cut -d: -f1 | xargs grep -L a_sse3
  • volk_32fc_x2_conjugate_dot_prod_32fc.h
  • Other things that came up
  • breaking out VOLK and versioning
  • Abhishek's work so far
  • Conference working group sessions -- going for more involvement

Developer's Call May 15th, 2014[edit]


access to then let us know

    • Embedded
    • Accelerators
    • Tutorials

Developer's Call April 17th, 2014[edit]


  • Releases
    • Upcoming 3.8: Rough timeline, discussion of API breakages (see also ProposedChangesV3_8)
    • What releases before that? 3.7.x?
    • When do we diverge next from master?
  • Tracker tool
  • GSoC
    • No status update, will move that to next call, because then, everything will be announced
  • GRCon
    • Status Update
  • Working Group Reports
    • GRC (Sebastian)
      • It is still basically just me
      • Added a menu option to hide disabled blocks (idea: John, Balint)
      • New Tools menu to call external tools like fdesign. Ideas?
      • Added Tabs to Param Dialog
      • Ability to base a params' XML definitions on another
      • When opening a flowgraph missing blocks are now shown (instead of being excluded)
      • Block and settings file paths can be set via ENV variables (Balint)
      • Param evaluation now happens on focus-out - on a trail basis (before: on change)
      • made a little progress on hard part of qtgrc: the display and interaction with blocks, ports and connections
    • Coprocessor (Justin)
      • Still not strongly organized
      • Active development effort is largely ARM/FPGA-related through embedded WG and Zynq
      • Hope to start new DSP-related effort and put energy into the coprocessor working group soon
    • VOLK

YouTube video link:



The call will take place on March 5th. 6PM UTC (noon US central time)


  • Review significant commits since last call
  • Review open issues
  • Stats
  • Upcoming projects
  • GSoC


You can replay the meeting here: http://www.youtube.com/watch?v=wutYjOz8jZ0

Notable features[edit]

volk regex[edit]

Use volk_profile -R multiply to run only kernels matching regex of multiply. Currently it does not overwrite profile config, but soon it will update the file (not overwrite) with latest results

volk malloc/free[edit]

Solve the issue of not having generic aligned memory calls. There's several existing solutions that are not necessarily platform independent or can be relied on. Use posix_memalign when available, falls back to Tom's custom memory pool. Always use volk_free with volk_malloc

  • Nathan has action item to replace qa_utils mem pool with volk_malloc/free.

volk_profile -b now has intrinsic value of 1, so just the presence of b is required now.

Review of issues[edit]

582/583 arm hardfloat problems[edit]
  • Several closed issues since past meeting
  • Constellation_receiver stuff crashes inside of VOLK, but not related to VOLK
  • Most kernels are

Still open; traced back to ORC overwriting the tol value in QA. This is critical for armhf, and not just a QA bug since it could corrupt memory at runtime. We should check in with ORC devs to see if they support hardfloat and whether they have seen similar issues before. Nick Foster knows people behind ORC. We may rip out ORC once we get NEON kernels in place.

  • Tom has action item to check in with Nick and ORC devs about armhf issues.


Tim and Tom are presenting a paper at WSR on their profiling work.

Upcoming Projects[edit]

  • Nathan to look at cache performance (hackfest?)
  • OFDM schmidl & cox kernel -- Abhishek is interested and applying for GSoC
  • Florian from OpenAirInterfaces is adding a GSoC project to our wiki to move their SIMD code in to VOLK

Developer's Call February 2014[edit]

Video: https://www.youtube.com/watch?v=VIxgeO3T_a8

It's pretty short right this time (22 Mins).



Martin explains what's happening.

  • We haven't been accepted yet!
  • Everything is documented
  • Once we're accepted, it's up to students
  • This year we should operate by rules of conduct.

Working groups (9:50)[edit]

Phil / Embedded[edit]

  • MetaSDR layer is not a place for fixes for recipes
  • Phil is looking for more actual recipes that are SDR related (and WX)
  • Help Tim w/ performance statistics
    • This uses volk_profile, anyone can help provide statistics
    • Allows us to track base performance and changes to VOLK kernels through time

Sebastian / GRC[edit]

  • We have a new error display for XML syntax
  • With the new feature, we had to change the XML file loading (figure out if block description or -tree)
    • Now, a block tree file has to be called as such
  • There will be a call soon, probably in March
  • GRC working group needs more participants!

Tim / VOLK[edit]

  • We need moar data (volk_profile!)

Martin / Tutorials[edit]

  • Good progress on the structure
  • We still need volunteers for 3 tutorials

Coming Up[edit]

  • WSR + Hackfest (Beer event on March 11th)
  • Hackfest in March

January 27 2014[edit]

  • Repo Manifest Updated. (Includes UHD 3.6.2, current master GNU Radio, adds Freescale and TI boards)
  • Meta-sdr, added gqrx, gr-osmosdr, update UHD, GNU Radio
  • Building ICE (and eventually, gr-ctrlport) in OE
  • Watch https://bugs.launchpad.net/gnuradio/+bug/1272024
  • Install GNU Radio toolchain files as part of GNU Radio
  • In early May, the repo manifest will be updated to follow OE-Core release.

January 16, 2014[edit]


  • FOSDEM Final Preparations
    • Merchandise?
  • Release Schedule
    • Next bug fix release (
    • Next minor version release (3.7.3)
      • Pending work
      • Probably 1 month off
  • Volunteer to curate signal repository
  • gr-atsc progress
  • QTGUI: What do we need to make this the default?
  • GSoC: Call for ideas
  • Working Group Status Update
    • Wiki Link
    • GRC (Nothing much to report, SK won't make it)
    • Embedded
    • VOLK
    • Tutorials
    • Accelerators

YouTube link



Is February 1st + 2nd, on 2nd, we have a SDR track with lots of speakers from GNU Radio and other projects.
Also, we have cool GNU Radio merchandise which will be available soon.

Upcoming releases[edit]

Mainly maintenance + stability.
Only pressing issue: VOLK issues on ARM with hard-float.
Fixes are in progress.


Andrew Davis is working on this. There will be a new, faster gr-atsc available, possibly in 3.7.3 if stable enough.
Question is how to put that in: Replace the old gr-atsc (what we normally don't do, 'cause API breakage), put in a gr-atsc2 (ugly) or wait until 3.8 to replace.

This depends on who uses gr-atsc -> Please write us an email.

Side topic: We'd like a curator for signals uploaded to gnuradio.org. Volunteers welcome!

Nathan has pointed out there's a torrent of RF captures available: academictorrents.com/collection/gnu-radio-rf-captures

QT GUI[edit]

No real actions here, but we want to push stuff more towards QT.
WX will stay where it is, unless it stops working for some reason.


Call for action! Please head to the GSoC page and post idea. We also need mentors, so volunteers, please!

WG leader: Have a look at what you can make a GSoC project!

Working Groups[edit]

Most WG's were quiet over the break, except for Phil's:

19:43 | Crofton|work > I've been doing a bunch of emebdded stuff
19:44 | Crofton|work > in the repo manifest
19:44 | Crofton|work > and I want to schedule a call
19:44 | Crofton|work > otherwise I will do this for my max convenience

December 19, 2013[edit]


  • Tagged Stream Blocks: Discussion
    • How can we optimally handle rate changing non-TSB blocks? Will Tim's scaling block for stream tags be enough for now?
    • Does the current implementation need rethinking? (Buffer problems)
  • Modulators (digital modulation, PSK etc.)
    • Call for action: Port everything modulator-related to C++ (e.g. 16-QAM etc.)
    • Possible entry point for new devs?
  • Torrent for data?
  • 2013 wrap-up
  • Next hackfest
  • Working Group status report

Youtube link:


Tagged stream blocks[edit]

  • tagged_stream_blocks have a problem if the packet is larger than the buffer size (trondeau, 18:25:40)
  • jcorgan: tagged_stream_blocks written before we had the message passing structure in place. (trondeau, 18:26:08)
  • LINK: http://gnuradio.org/doc/doxygen/page_tagged_stream_blocks.html (mbr0wn, 18:26:32)
  • moving from tagged_stream_blocks to message passing of PDUs. We have no flow control on message passing; no back-pressure. (trondeau, 18:27:53)
  • LINK: http://gnuradio.org/redmine/projects/gnuradio/wiki/Call20131219 (trondeau, 18:28:16)
  • LINK: https://www.youtube.com/watch?v=oLijSnXGK-I (mbr0wn, 18:28:17)
  • IDEA: jcorgan: make a new working group to look at tag streams? (trondeau, 18:32:48)
  • IDEA: mbr0wn: maybe wait until the hackfest to really set this up; start with some preliminary questions that need addressing. (trondeau, 18:34:05)
  • Martin will set up a task force.

Digital Modulation, Python 3[edit]

  • lots of modulation work done in Python --> should be moved to C++ (and swig'd) (trondeau, 18:37:28)
  • IDEA: easy entry point for new devs to move from Python to C++. (trondeau, 18:37:49)
  • the less Python we have, the easier we will find it to move to Python 3. (trondeau, 18:40:27)
  • concern over Cheetah -- no (announced) plans to move to 3. jcorgan: can we get away from Cheetah with the rework of grc? (trondeau, 18:41:55)
  • modtool also uses Cheetah
  • worries in general about dependency support for Python 3. Supporting both Python 2/3 may or may not be possible (probably not...) (trondeau, 18:42:50)
  • OE has python3 in process. SHouldn't be the limiting factor for GNU Radio using python 3 (Crofton|work, 19:05:49)

Hosting signal data sets via bitorrent etc[edit]

  • useful data sets for reference data for modulators/demodulators (LTE, wifi, ATSC). (trondeau, 18:44:50)
  • IDEA: use torrent to move data around. Start with data on gnuradio.org. (trondeau, 18:45:14)
  • ACTION: jcorgan and trondeau will start setting this up (trondeau, 18:46:15)

Working group status reports[edit]

  • Martin on User Experience; specifically, new tutorials that are well-structured from us.
  • Sebastian on GRC: a number of improvements have already been accomplished and in the works. (trondeau, 18:48:15)
    • work going on making "ghost blocks" that fill in when a block is missing in a GRC file. (trondeau, 18:48:36)
    • sebastian as submaintainer of GRC for gnuradio (trondeau, 18:49:30)
    • new staging repo for GRC fixes/improvements: https://github.com/gnuradio/gnuradio-wg-grc (trondeau, 18:50:04)
  • philip talking about the embedded WG.
  • Nathan on volk issues (trondeau, 18:53:10)
  • new stats collecting and display about volk capabilities on different processors/archs. (trondeau, 18:53:49)
    • new kernels in development; work on allowing asm (for NEON) in the volk system (trondeau, 18:54:21)
    • bugs on issue tracker that we are working on. Refactoring qa code that we think might be related to the bugs we're seeing more than the kernels. (trondeau, 18:54:52)
    • LINK: https://github.com/bistromath/vrt-dissector (bistromath, 18:57:54)


  • at WSR (march 12, 13) followed by 2-day hackfest (14, 15)
  • Next year: tba

VOLK Developer's Call December 2013[edit]


  • Progress with VOLK on ARM
  • QA issues
    • ARM failures.
    • misc (sum_of_poly)
  • New VOLK profiling tools
    • extending volk_profile with options
    • data collection

VOLK on ARM[edit]

A couple of mixed reports of failing kernels on ARM. Namely deinterleave and 32fc_magnitude with mixed success. Magnitude is off by 1 and appears that the tolerance given to PROFILE isn't propagating to the check correctly. The deinterleave was a typecast problem. Tom submitted fixes, but the bugs may still be alive.

Nathan also suspects that sum_of_poly may not be getting properly tested since passing. This came up while working on newer NEON kernels, which may have other QA issues as well.
Action item for Tom, Philip, and Nathan to schedule time on IRC to compare notes and code to resolve all of the above.

New profiling tools[edit]

Tim added a benchmark option to volk_profile to test all kernels available, including kernels where there is only a generic available. Nathan has an action item to add an option to run kernels by name.

stats.gnuradio.org is up and running. Likely a student project to clean up the interface.

Tim and Tom working on collecting CFLAGs from builds in GNU Radio to be saved along with gnuradio-config-info and a new volk-config-info. This along with /proc/cpuinfo and /proc/version provides a mostly complete view of a system. One missing bit is collecting actual clock speed of ARM processors and clock speeds across platforms. BOGOMIPS are reported in Linux for both Intel and ARM, but this is not available on OS X. Michael Dickens brought up code to manually test clock speed to be used across all platforms. Perhaps we write a function that we run regardless of platform to give us some indication of speed that we can use for comparison.

We want to make sure we capture as much information about the processor and platform as possible. Tell us about any missing information that might be useful.

New features in gnuradio-config-info and (a new tool) volk-config info to get compiler version and cflags used when building. volk-config-tool also reports the machines the library was built to include, all machines available on the current platform, and the machine volk will use when running.

Open Fora[edit]

Martin brought up kernels doing operations in place. There is no (known to members of call) restriction against this and it's probably possible to give the input and output argument the same pointer, but note that they will be aliased in the VOLK kernel.

A release is coming up in a few weeks. Nathan has an action item to push avx branch

Developer's Call November 2013[edit]


  • Next Hackfest
  • Do we want to scrap Sphinx?
  • Logo and Merchandise
  • OS X 10.9 macports building efforts
  • OpenEmbedded instructions
  • Working group monthly reports

YouTube Video


We are hoping to do a hackfest twice a year, roughly every 4 months with the conference making up a third event. We need to organize these for location and timing.

We want to do one early 2014, Jan - Mar time period. FOSDEM is in early February. WSR at Karlsruhe and WinnForum's SDR conferences are both the week of March 10. WSR is hoping to hold their own hackfest after WSR.

We could conceivably (if KIT can do it) have a week-long hackfest at KIT after WSR. An alternative is Ettus Research after SDR or something local in Chicago.

Sphinx Documentation[edit]

The Sphinx documentation was designed to help us understand the Python API and Python-specific calls. However, it doesn't always do a great job of pulling in all of the documentation from the C++ code, and it tends to be confusing to users who see 2 manuals. Sphinx also takes a bit of hand-holding to keep it up-to-date with new blocks and changes to the code.

We'd like to get rid of it, but need to first make sure we're not losing anything. Since we export the Doxygen markup as docstrings to the Python code and can access that straight through Python, we have most of what we need through there. Other aspects of the code in Python could be put into a page in the manual, as well.

The suggestion was to make a page describing how to marry the API in the Doxygen manual with working in Python as well as how to get access to the documentation in Python directly. We can then deprecate Sphinx.

Logo and Merchandise[edit]

If you haven't seen it, we have a new GNU Radio logo.

Part of the reason for the logo is to allow us to make merchandise to promote the project. We're hoping to have hoodies, shirts, stickers, etc. ready by FOSDEM with the new logo.

Tom Rondeau will produce a survey for the type of merchandise to prepare.

OS X 10.9 macports building efforts[edit]

Michael Dickens has fixed GNU Radio 3.7.2 to build on OS X 10.9

Some behind-the-scenes patching was done to get things to behave nicely. One VOLK patch for older versions of Clang will likely make it into GNU Radio.

OpenEmbedded instructions[edit]

Philip Balister has put together a repo manifest and instructions for building OE for ARM systems.

There are still a few issues to work out, but this is a relatively clean and easy way to get started on ARM-based embedded systems. Merged with Jonathon Pendlum's Zynq instructions to get running on Zedboards, etc.

Working group monthly reports[edit]

We want to remember to get updates on activities and progress by the Working Groups each dev call.

Working Group Call "User Experience" - November 14th, 2013[edit]

This will be the first call of the "user experience" working group. Please have a look at GRCon13Community.


  • Website
    • Redesign of front page
    • FAQ rehaul
    • Collect requirements. Does Redmine do all we want? Do we want an official GR blog (success stories, case studies...)?
    • Make sure the web site has enough info for anyone to get involved
  • Tutorials
    • We want an introductory course


  • Website will be split into a static part and the wiki (powered by Redmine)
  • The static part will include
    • Obvious link to technical stuff
    • Donations form
    • Twitter feeds
    • News streams
    • Sliders
    • Case studies / success stories
    • "Elevator Pitch", crafted towards different types of users (hackers, academia, industry)
  • A group will work on a default set of tutorials, which will go on to the static part
  • Another group will rehaul the FAQ, which is then made less prominent
  • Sphinx might go away (up to discussion)
    • We need to find out what's only in Sphinx

The recording is available at:

Developer's Call October 2013[edit]


  • Thoughts on GRCon13
  • Working groups (Embedded, VOLK, DSPs, User Experience, GRC) - What now?
    • Do we want separate mailing lists for each of these?

YouTube Video

Working Groups[edit]

  • ACTION: trondeau solve all the communication problems (Crofton|work, 17:29:45)
  • Working groups continue working
  • Communication will run via mailing list for now; perhaps hangouts. Working group leaders decide
  • Tags in subject lines (e.g. [Volk]) will help


Dev Call September 19, 2013[edit]

Time: 5pm UTC (1pm EDT)
Where: Google+ (check our G+ community page)


  • Porting issues 3.6./3.7 -- which version is the "official" version?
  • GRCon 2013


Porting issues and official version[edit]

  • 3.7 is official version, 3.6 is deprecated and updates should be done
  • ACTION: Buildgnuradio supports 3.7 by default (Crofton|work, 17:24:19)
  • ACTION: Update installling gr web page (Crofton|work, 17:24:32)

GNU Radio conference[edit]

  • we're sold out[[File:|]]

Dev Call August 16, 2013[edit]

Time: 5pm UTC (1pm EDT)
Where: Google+

YouTube: http://www.youtube.com/watch?v=G8HW7Vc0u74&feature=youtu.be

Dev Call July 18, 2013[edit]

Time: 5pm UTC (1pm EDT)
Where: Google+

Dev Call Part 1: http://www.youtube.com/watch?v=Iyxv4H6QSwE
Dev Call Part 2: http://www.youtube.com/watch?v=Uk2_M_w0Tnc

We went off air in the middle of the first call due to bandwidth issues. Tracie Perez's project overview was unfortunately cut off.


  • GSoC: The students have been invited to give a very brief presentation on their status.
    • Aneela Yasmeen: Improving the GNU Radio companion (GRC) (Sebastian Koslowski)
    • Jonathon Pendlum: GnuRadio FPGA Co-processing with the Xilinx Zynq System-on-Chip (Philip Balister)
    • Manu T S: LDPC codes and more FEC in GNURadio (Jens Elsner)
    • Shashank Gaur: GNU Radio: 802.11a/n/g Receiver and Wireshark Connector (Ankit Kaushik)
    • Tracie Perez: Implementation of LDPC encoder and decoder in GNU Radio (Jens Elsner)
  • GRCon updates
    • Call for Presentations is live and was announced. Need more advertising.
    • Registration page should be live soon.
  • 3.7
    • 3.7.0 has been released
    • Any current feedback? What's next?
    • Porting projects?

Dev Call June 20, 2013[edit]

This call will take place via Google Hangout, make sure to join our community


Dev Call May 16, 2013[edit]

This call will take place via Google Hangout, make sure to join our community


  • Donations
  • Status 3.6 and 3.7
  • GRCon
  • Hackfest


Meeting started by Crofton|work at 17:10:42 UTC.

3.7 and 3.6 releases[edit]

  • Hope we have 3.7.0rc0 by the hackfest (mbr0wn, 17:15:57)
  • We can work off the RC at the hackfest, will test it (mbr0wn, 17:16:09)
  • Lack of tests of OOT-Modules against pre-3.7 (mbr0wn, 17:18:18)
  • ACTION: Everyone Test pre-3.7 with your own stuff! (mbr0wn, 17:18:45)
  • Stuff missing from 3.7: Ben's documentation and Python QA branches (mbr0wn, 17:21:45)
  • Also: Sean's gr-uhd branch (mbr0wn, 17:22:33)
  • Missing for 3.6: OFDM over the air (mbr0wn, 17:24:21)

GRCon 2013[edit]

  • John's looking for affordable venues (mbr0wn, 17:26:35)
  • GRCon shall be at least a week after DARPA SC Prelims (mbr0wn, 17:27:48)
  • Around Sep 30th (mbr0wn, 17:28:00)

Hackfest @ VT[edit]

  • Right now around 15 participants (mbr0wn, 17:30:39)
  • Monday we'll get organized and do presentations (mbr0wn, 17:33:22)
  • VT students will have opportunity to see real GNU Radio developers (mbr0wn, 17:34:40)
  • Maybe even talk to them (mbr0wn, 17:34:45)
  • Monday night: Working Dinner (you may enjoy the food, though) (mbr0wn, 17:36:41)


Are now possible.


Coming along as intended. Many, many participants.

Next Call[edit]

  • June 20th, on G+ again (mbr0wn, 17:47:59)

Meeting ended at 17:48:49 UTC.

Action Item[edit]

  • Everyone: Test pre-3.7 with your own stuff!

Dev Call April 18, 2013[edit]

Meeting started by mbr0wn at 17:04:32 UTC.

Meeting summary[edit]

3.6 (mbr0wn, 17:05:20)[edit]

  • mbr0wn working on completing OFDM redo for 3.6.5. (trondeau, 17:06:28)
  • OFDM is still missing (mbr0wn, 17:07:47)
  • ACTION: mbr0wn fix that (mbr0wn, 17:07:58)
  • Also, some bug fixes are missing, but they would probably be a (mbr0wn, 17:08:19)
  • Also, Sean's UHD stuff is missing (mbr0wn, 17:08:39)
  • marco is interested in GSoC (Crofton|work, 17:10:00)
  • ACTION: take mait's bsd bug patches into GR. (trondeau, 17:23:39)

Google Hangout for dev calls? (mbr0wn, 17:13:03)[edit]

3.7 (mbr0wn, 17:23:03)[edit]

  • One final API change left (includes) (mbr0wn, 17:23:48)
  • Then, next branch is stable enough for using (mbr0wn, 17:24:08)
  • There will be a RC0 (mbr0wn, 17:24:23)
  • ACTION: Crofton|work will test next on non-x86 (mbr0wn, 17:26:22)
  • ACTION: Crofton build next for embedded (Crofton|work, 17:26:33)
  • ACTION: mait will try and build for Debian (mbr0wn, 17:27:13)

Deb packages (mbr0wn, 17:30:18)[edit]

  • Ubuntu grabs packages from Debian testing (mbr0wn, 17:31:55)
  • ACTION: mbr0wn will change the wiki to promote the binaries over the source build (mbr0wn, 17:37:46)
  • ACTION: mait will try to package rtlsdr, gqrx etc. (mbr0wn, 17:38:21)

GSoC (mbr0wn, 17:39:21)[edit]

  • We have enough proposals such that we'll probably have good projects (mbr0wn, 17:41:27)
  • ACTION: mbr0wn Crofton organize some form of application triage after applications are closed (Crofton|work, 17:45:14)
  • IDEA: try to make access to the wiki easier. (trondeau, 17:54:46)
  • Next call: May 16 (mbr0wn, 17:59:06)

Meeting ended at 17:59:16 UTC.

Action Items[edit]

  • mbr0wn fix that (OFDM)
  • mbr0wn email people about dev call on G+ Hangout
  • Crofton|work checks G+
  • take mait's bsd bug patches into GR.
  • Crofton|work will test next on non-x86
  • Crofton build next for embedded
  • mait will try and build for Debian
  • mbr0wn will change the wiki to promote the binaries over the source build
  • mait will try to package rtlsdr, gqrx etc.
  • mbr0wn Crofton organize some form of application triage after applications are closed

Action Items, by person[edit]

  • Crofton|work
    • Crofton|work checks G+
    • Crofton|work will test next on non-x86
    • Crofton build next for embedded
  • mait
    • take mait's bsd bug patches into GR.
    • mait will try and build for Debian
    • mait will try to package rtlsdr, gqrx etc.
  • mbr0wn
    • mbr0wn fix that
    • mbr0wn email people about dev call on G+ Hangout
    • mbr0wn will change the wiki to promote the binaries over the source build
    • mbr0wn Crofton organize some form of application triage after applications are closed

Developer's Call March 21, 2013[edit]

Time: 1700 UTC, 10 AM Pacific, 1PM Eastern, 18:00 CET.

Meeting summary[edit]

  • 3.6.5 (mbr0wn, 17:19:01)
    • issue #518 is pretty much complete for 3.6.5 (trondeau, 17:22:14)
    • ACTION: mbr0wn will finish this and close the issue. (trondeau, 17:22:32)
    • OFDM blocks are also pretty much done (mbr0wn, 17:22:40)
    • for issue #503, we're close on this restructuring. Some problems with tag propagation. (trondeau, 17:25:30)
    • Major points for 3.6.5: OFDM tag-bug and better use of logger (mbr0wn, 17:31:59)
    • ACTION: trondeau open a ticket for the use of the logger. (trondeau, 17:32:54)
  • GSoC (mbr0wn, 17:33:20)
    • Two students have contacted Phil and Jens, respectively (mbr0wn, 17:41:53)
    • Parallella have contacted Tom (mbr0wn, 17:41:53)
  • GR-Con (mbr0wn, 17:44:30)
    • About 90 attendees (mbr0wn, 17:46:09)
    • IDEA: support a poster session and a demo day/session (trondeau, 17:53:31)
    • IDEA: demos to Boston free software community (mait, 17:55:26)
    • AGREED: We need to nail the dates/location ASAP (mbr0wn, 17:58:17)
    • IDEA: Let's have an additional call for the GRCon (mbr0wn, 18:01:11)
  • Hackfest (mbr0wn, 18:02:20)
    • Happening directly after the VT Wireless Symposium (mbr0wn, 18:02:35)

Meeting ended at 18:05:30 UTC.

Action Items[edit]

  • mbr0wn will finish this and close the issue.
  • trondeau open a ticket for the use of the logger.

Generated by `MeetBot`_ 0.1.4

.. _`MeetBot`: http://wiki.debian.org/MeetBot

Dev Call Feb 21, 2013[edit]


  • 1800 UTC
  • 10 AM Pacific
  • 1 PM Eastern
  • 19:00 CET


As recorded by MeetBot, thanks to Phil.

Meeting started by mbr0wn at 18:01:17 UTC. The full logs are available as attachment.

Meeting summary[edit]

  • LINK: http://wiki.debian.org/MeetBot (trondeau, 18:05:19)
  • Dev Calls (mbr0wn, 18:06:06)
    • IDEA: Asians might want their own dev call (mbr0wn, 18:07:48)
    • IDEA: Pay for a professional audio conference call service (runexe, 18:08:47)
    • IDEA: Best use of API/version numbering in closing tickets (runexe, 18:11:54)
    • IDEA: Encouraging people on the mailing list to open tickets for reported issues (runexe, 18:12:54)
    • IDEA: Keeping original reporter included discussion (ala patch- mailing list) (runexe, 18:15:26)
    • IDEA: Point people to RSS feeds on redmine issue list to follow tickets opened/closed/etc. (runexe, 18:18:52)
  • FSF Funding (mbr0wn, 18:19:55)
  • GRHack (mbr0wn, 18:23:25)
    • ACTION: OFDM/PDU tutorials must be added to the wiki (mbr0wn, 18:28:32)
    • ACTION: Need to investigate possible bug in scheduler/stream tag interaction (particularly rate-changing blocks). Look into off-by-one errors. (runexe, 18:30:14)
    • ACTION: Merging OFDM work into either v3.6 or v3.7 (sooner is better) (runexe, 18:34:09)
    • IDEA: Maybe tags are pruned too soon (mbr0wn, 18:36:41)
  • 3.6.4 release (mbr0wn, 18:37:21)
    • AGREED: 3.6.4 w/o OFDM, make 3.6.5 w/ OFDM (mbr0wn, 18:38:15)
    • AGREED: 3.6.4 release time: around next week (mbr0wn, 18:38:51)
    • AGREED: "Feature freeze" for 3.6.4 starts now (mbr0wn, 18:40:47)
    • IDEA: extra point-release for bug-fixes:, then 3.6.4 is new features (runexe, 18:41:41)
    • IDEA: always release a patch fix version whenever we release a new minor version (like have a when we release a 3.6.5). (trondeau, 18:45:31)
    • Tom's working on the performance counters (mbr0wn, 18:51:48)
    • That's the last feature before we can freeze gnuradio-core (mbr0wn, 18:52:00)
    • Which is necessary to convert that to gr-runtime (mbr0wn, 18:52:25)
    • The performance counters are pretty cool (mbr0wn, 18:54:06)
    • performance counters turned on/off via standard gnuradio config files (runexe, 18:54:27)
    • and can be exported via gr-ctrlport (runexe, 18:54:45)
    • counters are platform independant (runexe, 18:55:20)
    • ACTION: runexe Puts down a new wiki page on that (mbr0wn, 18:56:06)
    • ACTION: martin + ben: Get OFDM merged ASAP, even if it's not really working (mbr0wn, 18:59:08)
    • ACTION: make sure to work with benreynwar to fix Python documentation before 3.7 release. (trondeau, 19:01:35)
  • gsoc (mbr0wn, 19:01:59)
    • IDEA: gsoc ideas: better linking between GRC and the code. (trondeau, 19:03:56)
    • Please add more ideas to the GSoC wiki page (runexe, 19:05:21)
    • Sign up early to be a mentor! (runexe, 19:07:26)
  • Hackfest 2013? (mbr0wn, 19:09:51)
    • IDEA: Hold one in conjunction with VT Wireless Symposium this summer (runexe, 19:11:35)
    • LINK: http://wireless.vt.edu/symposium/2013/ (Crofton, 19:13:55)
    • IDEA: Alternate locations (East Coast: Philly, DC, Boston) (runexe, 19:14:34)
    • ACTION: Tom contacts people from VT (mbr0wn, 19:19:08)
  • grcon (mbr0wn, 19:20:41)
    • GrCon'13: tentatively in Sept. (runexe, 19:21:53)
    • Tentative location is Boston (based on discussions from GrCon'12) (runexe, 19:26:15)
    • ACTION: mbr0wn and trondeau get in touch with Matt for more planning on grcon. (trondeau, 19:28:00)
  • Misc (mbr0wn, 19:28:48)
    • One-day SDR/GNURadio info-day in Rio de Janeiro this spring (runexe, 19:29:59)

Action Items[edit]

  • OFDM/PDU tutorials must be added to the wiki
  • Need to investigate possible bug in scheduler/stream tag interaction (particularly rate-changing blocks). Look into off-by-one errors.
  • Merging OFDM work into either v3.6 or v3.7 (sooner is better)
  • runexe Puts down a new wiki page on that: PerformanceCounters
  • martin + ben: Get OFDM merged ASAP, even if it's not really working
  • make sure to work with benreynwar to fix Python documentation before 3.7 release.
  • Tom contacts people from VT
  • mbr0wn and trondeau get in touch with Matt for more planning on grcon.

Action Items, by person[edit]

  • benreynwar
    • make sure to work with benreynwar to fix Python documentation before 3.7 release.
    • martin + ben: Get OFDM merged ASAP, even if it's not really working
    • Merging OFDM work into either v3.6 or v3.7 (sooner is better)
  • mbr0wn
    • mbr0wn and trondeau get in touch with Matt for more planning on grcon.
    • OFDM/PDU tutorials must be added to the wiki
  • runexe
    • runexe Puts down a new wiki page on that
  • trondeau
    • mbr0wn and trondeau get in touch with Matt for more planning on grcon.
    • Tom contacts people from VT
    • Need to investigate possible bug in scheduler/stream tag interaction (particularly rate-changing blocks). Look into off-by-one errors.

Developer's Call, July 19, 2012[edit]

Annotated Agenda[edit]


Time: 4 PM (1600) EDT or 2000 UTC
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda Items[edit]

  • GR Conference
    • Student travel application announced and out
    • Call for presentations has ended and notifications have been sent
      • 9 accepted talks
      • Will be posting abstracts on website soon
    • Need list of final invited talks
      • Will build schedule around this
  • 3.7 Development work
    • gr-filter is almost done and all components transferred over (on next)
    • Progress moving on gr-blocks
    • gr-qtgui has been having a significant make-over; just updated to 3.7 structure/style; tons of new features and capabilities.
    • Still hoping to be ready by the conference for release.
  • Volk
    • Stil lintermittent problems
    • Anything major outstanding?
    • New work on alignment and changing nbytes to nitems for all blocks needs to be handled.
      • this will happen in next
    • Thinking about making a stand-alone library;
      • Lots of issues to think about with this, but it's where we want to be.
  • gr-osmo
    • Can we / Should we absorb the gr-osmo hardware abstraction into GNU Radio?


Developer's Call, June 21, 2012[edit]

Annotated Agenda[edit]


Time: 4 PM (1600) EDT or 2200 UTC
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda Items[edit]

  • New structure
    • gr-filter has been merged with master
      • Uses new structure
      • Basic documentation and QA code for all blocks
      • Will be removing these blocks on next
    • gr-blocks is progressing
  • Conference
    • Flyer available - please distribute!
    • NSF Student travel funding approved!
    • Waiting to see about abstract submissions



Annotated Agenda[edit]


Time: 6 PM (1800) EDT or 2200 UTC
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda Items[edit]

  • New structure
    • Roll-out schedule
    • Comments on approach
    • what to do about apps (per Andrew Davis' question)
  • Conference
    • Working on schedule
    • Need advertising!
    • Call for presentations available
      • Due date is June 29
    • Still waiting on NSF travel grant go/no-go
    • Finish flyer


  • Aiming to get version 3.7 out before GRCon2012
  • Sphinx (Python documentation)
    • Difficulty in building Sphinx -- looks at fully-installed Python mods
    • Some possible solutions:
      • build local dir to install and run sphinx against prior to system install
      • make two install targets; install, then run sphinx, then install sphinx docs
      • build a script to run and install sphinx
    • The last idea seems to have lots of benefits for how to handle it.
    • Tom will be working on this.
  • gr-apps
    • We have some top-level components (gr-pager, gr-noaa, gr-atsc) that are really GNU Radio applications.
    • Does it make sense to have them located the same as others components?
    • We could:
      • Move them to a gr-apps (or just apps) directory
      • Move them to their own projects either on CGRAN or hosted on gnuradio.org
    • Moving them out of GNU Radio opens up possible bitrot issues.
    • What's the motivation for wanting to do this?
  • Packag VOLK into stand-alone library
    • VOLK has tons of possible uses outside of GNU Radio. Making it a stand-alone library might help its adoption.
  • Conference
    • Conference flier by May 28th (Wireless@VT Symposium)
    • Need to publicize more!

Developer's Call, March 15, 2012[edit]

Annotated agenda[edit]


Time: 4 PM (1600) EDT or 2000 UTC
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items[edit]

  • Discuss any further points on the 'pimpl'
  • GR Conference Updates
    • News on location
    • Need to start getting word out, finding speakers
  • New version schedule
    • Volk has some 32-bit machine issues
      • Some will be fixed today - cut a
    • Should add cmake build to releases
      • How soon do we need a 3.5.3?
    • Discuss what's left for 3.6

IRC Log[edit]

* voip-gr (~voip-gr@digitalbazaar.com) has joined #gnuradio
  (IAX2/diamondcard-12761) is in the conference.
 Tom (SIP/xxx.xxx.xxx.xxx-0000001c) is in the conference.
  (SIP/xxx.xxx.xxx.xxx-0000001e) has joined the conference.
* benreynwar (~ben@ip24-251-204-81.ph.ph.cox.net) has joined #gnuradio
 Ben Reynwar (SIP/ekiga.net-0000001f) has joined the conference.
 Philip Balister (SIP/xxx.xxx.xxx.xxx-00000020) has joined the conference.
 Looking at the debian build logs: https://buildd.debian.org/status/package.php?p=gnuradio
 Looks like a bit of work for some of these build systems.
 Patrik Tast (SIP/poes-weather.com-00000021) has joined the conference.
 Lg Conference Room (SIP/xxx.xxx.xxx.xxx:5825-00000022) has joined the conference.
 voip-gr: 1e is mait
 mait is now associated with SIP/xxx.xxx.xxx.xxx-0000001e.
* guruofquality (~jblum@xxx.xxx.xxx.xxx) has joined #gnuradio
 agenda: Discuss any further points on the 'pimpl'
* ptast (~patrik@av83.netikka.fi) has joined #gnuradio
 Patrik Tast (SIP/poes-weather.com-00000021) has left the conference.
 voip-gr, 5825 is HokieTux 
 voip-gr: 22 is HokieTux 
 HokieTux is now associated with SIP/xxx.xxx.xxx.xxx:5825-00000022.
 I've package Alex's gr-fcd - http://bugs.debian.org/660570
 hopefully that's just like what Jonathan is merging.
 I hear echo, and my phone is muted.
 I muted myself for a bit
 GR Conference Updates
 News on location
 Sheraton Gateway Hotel ATL 9/24 - 9/27
 well, I would talk about actual gnuradio content :)
 there is 500 per student for mentor orgs
* manu1 (~chatzilla@pool-74-107-166-95.ronkva.east.verizon.net) has left #gnuradio
 Crofton|work: you can do whatever you want on that :)
 FYI, looking way ahead, mentor summit is Oct 20-21
 confirm 500 for the mentor org, and 5K per student
 HokieTux (SIP/xxx.xxx.xxx.xxx:5825-00000022) has left the conference.
 Lg Conference Room (SIP/xxx.xxx.xxx.xxx:21859-00000023) has joined the conference.
 man thats annoying
 March 16 @ 1900 UTC list of mentor orgs published
 voip-gr: 23 is HokieTux 
 HokieTux is now associated with SIP/xxx.xxx.xxx.xxx:21859-00000023.
 set your timer for 15 minutes :)
 Crofton|work, are there requirements on attending the mentor summit if we get funding?
 HokieTux, we'll figure that out later, I just wanted it in the notes to make sure we do not clash
 Crofton|work, roger
 New version schedule
 Volk has some 32-bit machine issues
 I can talk, but no one hears me
 Philip Balister (SIP/xxx.xxx.xxx.xxx-00000024) has joined the conference.
 I'll push my volk fixes for 32-bit machines. We'll create a xxx.xxx.xxx.xxx
 3.5.3: Should add cmake build to releases?
 HokieTux (SIP/xxx.xxx.xxx.xxx:21859-00000023) has left the conference.
 Lg Conference Room (SIP/xxx.xxx.xxx.xxx:56876-00000025) has joined the conference.
 EXTRA_DIST the CMakeLists.txt files. Have to add the cmake dir, too.
 Will then be packaged when doing a 'make dist'
 Discuss what's left for 3.6
 and do we want to think about a timeline for its release?
 Andrew Davis (SIP/xxx.xxx.xxx.xxx-00000026) has joined the conference.
 We did a check for installing with autotools and cmake to see what files are installed and what are different.
 We can do the same thing when Python is disabled to make sure.
 There's a case to be made to just compile the Python parts - to build separate packages for all supported versions of Python in Debian.
 The Debian packages will debug any of those issues too.
 Lg Conference Room (SIP/xxx.xxx.xxx.xxx:56876-00000025) has left the conference.
 Lg Conference Room (SIP/xxx.xxx.xxx.xxx:37125-00000028) has joined the conference.
 was the cmake includes issue w/ osx a gr-discuss thread? I would like to get a hold of the output of make VERBOSE=1
 no, phone call with mdickens
 i've got an email from him with the output i think
  (IAX2/diamondcard-12761) has left the conference.
 Lg Conference Room (SIP/xxx.xxx.xxx.xxx:37125-00000028) has left the conference.
 Andrew Davis (SIP/xxx.xxx.xxx.xxx-00000026) has left the conference.
 Tom (SIP/xxx.xxx.xxx.xxx-0000001c) has left the conference.
 mait (SIP/xxx.xxx.xxx.xxx-0000001e) has left the conference.
* benreynwar has quit (Remote host closed the connection)
 So I learned to get mic attached via BT before starting ekiga
 Recap: will push fixes to Volk for xxx.xxx.xxx.xxx (now); we can then cut a new release for that. maint will tag that for a Debian release.
 I think I need to make sure include_directories() for local stuff is listed last
 We can then work on 3.5.3 in the next week or two with cmake as a build option in the release.
 under 24 hours for gsoc announcement
 Crofton|work: yeah, excited to hear back about that.
 I think we ahve a reasonable application
 I do, too. And lots of good ideas.
 there is a lot of interest at KIT from people that would like to participate as students and mentors
 I will be about to get on a plane for CO around the annoucement time
 That's good. I'm sure KIT has a great pool of talent for it.

Developer's Call, February 16, 2012[edit]

Annotated agenda[edit]


Time: 4 PM (1600) EST or 2100 UTC
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items[edit]

  • Issues
    • Quick Burns
      • Bug #478 - [[|docs: file generation synchronization issue]] Any problems since?
      • Bug #468 - [[|volk - make all unit tests pass]] extensive tests in Linux on various processors give me no problem. Anyone else seeing issues?
    • Issues to Discuss
      • Bug #481 - [[|qa_vector_sink_source.py failure - RuntimeError thrown, expecting ValueError]] We don't want to catch all exceptions since we're looking for a particular one. This indicates another problem with BSD. Any ideas?
      • Feature #444 - [[|volk: handle alignment]] my safe_align branch should take care of this. What about platforms that don't care about alignment? We need to make sure scheduler alignment gets set to 1 in these cases.
    • Other bugs/features to discuss?
  • GR Conference
    • Working well with TAPR and the DCC to co-locate our conferences.
      • Likely to be at the Atlanta Airport Sheraton from Sept. 17 - 19 or 20.
    • Should the conference be 2 days + hackfest or 3 days + hackfest?

IRC Chat Log[edit]

Feb 16 15:43:44 *   jcorgan (~jcorgan@gw-sj1.corganenterprises.com) has joined #gnuradio
Feb 16 15:51:38 *   mait (aa4hs@pool-108-18-8-132.washdc.east.verizon.net) has joined #gnuradio
Feb 16 15:57:29    Josh Blum (SIP/xxx.xxx.xxx.xxx:23725-000000be) has joined the conference.
Feb 16 15:59:59    Tom (SIP/xxx.xxx.xxx.xxx-000000bf) has joined the conference.
Feb 16 16:00:14     (SIP/xxx.xxx.xxx.xxx-000000c0) has joined the conference.
Feb 16 16:00:38     (IAX2/diamondcard-4411) has joined the conference.
Feb 16 16:02:00   Conference call starting
Feb 16 16:02:06   Agenda: http://gnuradio.org/redmine/projects/gnuradio/wiki/Call20120216
Feb 16 16:04:30   Bug #478 - "docs: file generation synchronization issue": Any problems since?
Feb 16 16:05:10   We think this provides a solution.
Feb 16 16:05:19   Bug #468 - "volk - make all unit tests pass": extensive tests in Linux on various processors give me no problem. Anyone else seeing issues?
Feb 16 16:06:02 *   CarlesFernandez (4f9bb439@gateway/web/freenode/ip.xxx.xxx.xxx.xxx) has joined #gnuradio
Feb 16 16:06:06   a rounding error issue in Windows 7 still exists.
Feb 16 16:07:19    Martin Braun (SIP/xxx.xxx.xxx.xxx-000000c1) has joined the conference.
Feb 16 16:08:48   nick is going to take a look at it.
Feb 16 16:08:52     voip-gr: c1 is mbr0wn
Feb 16 16:08:52    mbr0wn is now associated with SIP/xxx.xxx.xxx.xxx-000000c1.
Feb 16 16:09:26   voip-gr: c0 is mait
Feb 16 16:09:26    mait is now associated with SIP/xxx.xxx.xxx.xxx-000000c0.
Feb 16 16:09:39     the guess is that it's something in the C stuff in the volk routine where it's being cast in a way that windows interprets differently than linux
Feb 16 16:09:46   Bug #481 - "qa_vector_sink_source.py failure - RuntimeError thrown, expecting ValueError": We don't want to catch all exceptions since we're looking for a particular one. This indicates another problem with BSD. Any ideas?
Feb 16 16:10:32     Bug #468 I saw some problems with volk tests in macos (reported to the distribution list)
Feb 16 16:11:01   CarlesFernandez: i don't think I saw that. Could you repost with verbose mode turned on?
Feb 16 16:11:11    of course
Feb 16 16:11:22   Feature #444 - "volk: handle alignment": my safe_align branch should take care of this. What about platforms that don't care about alignment? We need to make sure scheduler alignment gets set to 1 in these cases.
Feb 16 16:13:26    Josh Blum (SIP/xxx.xxx.xxx.xxx:23725-000000be) has left the conference.
Feb 16 16:13:43   your 15 minutes are up.
Feb 16 16:13:53    Josh Blum (SIP/xxx.xxx.xxx.xxx:18480-000000c2) has joined the conference.
Feb 16 16:20:50   GR Conference
Feb 16 16:20:54   Working well with TAPR and the DCC to co-locate our conferences.
Feb 16 16:21:39   Likely to be at the Atlanta Airport Sheraton from Sept. 17 - 19 or 20.
Feb 16 16:22:31   Should the conference be 2 days + hackfest or 3 days + hackfest?
Feb 16 16:24:09    Javier Arribas (SIP/xxx.xxx.xxx.xxx-000000c3) has joined the conference.
Feb 16 16:24:41     ECHO
Feb 16 16:24:59   ok I'm in
Feb 16 16:25:13   Google Summer of Code:
Feb 16 16:25:20   Website full of ideas: http://gnuradio.org/redmine/projects/gnuradio/wiki/GSoC
Feb 16 16:25:26 *   yootis (~matt@xxx.xxx.xxx.xxx) has joined #gnuradio
Feb 16 16:25:29   Pare down to 5-6 ideas
Feb 16 16:25:30    Javier could you mute your microphone unless speaking?
Feb 16 16:26:19   voip-gr: mute c3
Feb 16 16:26:19    SIP/xxx.xxx.xxx.xxx-000000c3 has been muted.
Feb 16 16:29:49    Josh Blum (SIP/xxx.xxx.xxx.xxx:18480-000000c2) has left the conference.
Feb 16 16:29:57  sorry to interrupt your meeting, but how do I post an issue?
Feb 16 16:30:34    Josh Blum (SIP/xxx.xxx.xxx.xxx:45656-000000c4) has joined the conference.
Feb 16 16:30:42  gnuradio doesn't seems to be very "friendly" with other apps trying to use audio devices... I can't manage to have audio sink, in my flow, and play a mp3 at the same time (using rythmbox, for example)
Feb 16 16:31:32   tavl|work: use the 'pulse' device in a GR audio app (usually the -O option) to work with pulse audio and be more sharing friendly.
Feb 16 16:31:45   voip-gr: connections?
Feb 16 16:31:46    Conference connections:  (mait) [SIP/xxx.xxx.xxx.xxx-000000c0], Javier Arribas [SIP/xxx.xxx.xxx.xxx-000000c3], IAX2/diamondcard-4411, Josh Blum [SIP/xxx.xxx.xxx.xxx:45656-000000c4], Martin Braun (mbr0wn) [SIP/xxx.xxx.xxx.xxx-000000c1], Tom [SIP/xxx.xxx.xxx.xxx-000000bf]
Feb 16 16:32:03  voip-gr: connections?
Feb 16 16:32:03    Conference connections:  (mait) [SIP/xxx.xxx.xxx.xxx-000000c0], Javier Arribas [SIP/xxx.xxx.xxx.xxx-000000c3], IAX2/diamondcard-4411, Josh Blum [SIP/xxx.xxx.xxx.xxx:45656-000000c4], Martin Braun (mbr0wn) [SIP/xxx.xxx.xxx.xxx-000000c1], Tom [SIP/xxx.xxx.xxx.xxx-000000bf]
Feb 16 16:32:05  oooo
Feb 16 16:33:14 *   GLneo (~GLneo@c-98-223-35-49.hsd1.in.comcast.net) has joined #gnuradio
Feb 16 16:33:15  trondeau, what about flows built with GRC?
Feb 16 16:33:54  trondeau, (sorry, didn't see the "device" option in the sink block)
Feb 16 16:34:00    Andrew Davis (SIP/xxx.xxx.xxx.xxx-000000c5) has joined the conference.
Feb 16 16:34:18   it is my fault sorry
Feb 16 16:34:24  sorry
Feb 16 16:34:29   I can't enable the microphone
Feb 16 16:36:38   voip-gr: unmute c3
Feb 16 16:36:38    SIP/xxx.xxx.xxx.xxx-000000c3 has been unmuted.
Feb 16 16:38:25     I'm sorry I can't join the voice call today, but I do want to chime in on the GSoC discussion:
Feb 16 16:38:57     first, I'd like to suggest a possible project: updating George Nychis latency measurements tools to work with UHD (and therefore the rest of the USRP family)
Feb 16 16:39:53     second, I'd like to offer at least some amount of my time to help with mentoring whomever might be working on a GSoC project (I'm willing to negotiate on what that might mean)
Feb 16 16:40:46    Patrik Tast (SIP/gnuradio@digitalbazaar.com-000000c6) has joined the conference.
Feb 16 16:43:09   runexe: thanks, we're discussing it.
Feb 16 16:44:11  alot of GNURadio can be used with sound card/softrocks
Feb 16 16:44:30  so some projects dont need real hardware
Feb 16 16:46:31    Josh Blum (SIP/xxx.xxx.xxx.xxx:45656-000000c4) has left the conference.
Feb 16 16:46:57    Josh Blum (SIP/xxx.xxx.xxx.xxx:27550-000000c7) has joined the conference.
Feb 16 16:47:01    Sorry, my sip client just does not work. We are the ones that proposed the GNSS receiver (GPS + Glonass + Galileo + ...) for the GSoC 
Feb 16 16:47:52  http://t-filter.appspot.com/fir/index.html
Feb 16 16:52:48 *   mait has been looking for VU3RDD's gnurokr code. Anyone have a copy?
Feb 16 16:53:04   It's a softrock interface
Feb 16 16:54:23    Carles, we discussed it briefly, some interest but the feeling was there was a limited amount that could be done in a summer by a student
Feb 16 16:56:21    yes, it's true. We already have a working GPS L1 receiver, and maybe we could extend it in several ways: algorithms, maybe Glonass... 
Feb 16 16:56:25   Hi Johnathan, I'm also working with Carles.
Feb 16 16:56:57    If there is some interest, our code is at http://sourceforge.net/projects/gnss-sdr/
Feb 16 16:57:31    oh, that's you guys?
Feb 16 16:57:38   Yes but I can't enable
Feb 16 16:57:48    ... but maybe GSoC is not the best framework. Anyway, it was just a suggestion ;-)
Feb 16 16:58:18     CarlesFernandez: you're idea is pretty good
Feb 16 16:58:25   I cant't the microphone... I have huge echo... with my laptop microphone.. but I'm hearing
Feb 16 16:59:22    we are also setting up a webpage (not already finished) here: http://gnss-sdr.org/
Feb 16 16:59:25   jarribas, gotcha. You should be unmuted on the call at least
Feb 16 17:00:19     https://github.com/gps-sdr/gps-sdr/tree/master/usrp
Feb 16 17:00:43     ^^^that's the one i was thinking of
Feb 16 17:00:50    oh, i had them confused
Feb 16 17:01:12    gps-sdr is also a nice project
Feb 16 17:01:24    but it does not make use of GNU Radio
Feb 16 17:01:42     CarlesFernandez: sounds like you have more familiarity than i do with it. is gnss-sdr Gnuradio-based?
Feb 16 17:01:49   This is the GPS receiver from Gregory
Feb 16 17:02:14    yes bistromath, gnss-sdr uses GNU Radio
Feb 16 17:02:19     excellent!
Feb 16 17:02:53    Josh Blum (SIP/xxx.xxx.xxx.xxx:27550-000000c7) has left the conference.
Feb 16 17:03:04     hahahaha
Feb 16 17:03:06   FYI, current Debian binary library packages have names like: libgnuradio-core-3.5.1-0.0.0_3.5.1-1_amd64.de
Feb 16 17:03:10    Josh Blum (SIP/xxx.xxx.xxx.xxx:4798-000000c8) has joined the conference.
Feb 16 17:03:14   b a bit long.
Feb 16 17:04:14    a UHD port of gps-sdr would be nice
Feb 16 17:05:19    yes horiz0n, that would be a project that fits in a summer
Feb 16 17:07:09    CarlesFernandez: btw, why are you guys using jam and not something more common to gnuradio (like cmake)?
Feb 16 17:09:44    well, we started with bjam and found it reasonably easy. But we are considering to try cmake (after all they can live together)
Feb 16 17:09:53   well, in the time that we started the project (GNURadio 3.2.0) GNURadio was using automake.. 
Feb 16 17:10:52   and it was not so easy to re-use the existing gnuradio environment... we decided to switch to bjam
Feb 16 17:10:59    cmake and git are those things we want to do sooner or later :-)
Feb 16 17:11:17   We've gotten a lot better with that in general; better even with cmake.
Feb 16 17:11:26    but at this moment bjam and svn work fine for us
Feb 16 17:12:47    ok, understand
Feb 16 17:12:51    but yes, you are right that those are superior tools
Feb 16 17:14:02    it's exciting stuff, will check out your code more closely tomorrow :)
Feb 16 17:15:04    we are actively working on it; please do not expect something even close to stable :-)
Feb 16 17:15:43   tavl|work: as for the issues you asked about, use our redmine website (gnuradio.org). You'll have to sign up for an account and then ask me to add you to the GNU Radio project. You can then make a "New Issue" from there.
Feb 16 17:16:16    we want to post some sample files for testing, we are looking for some server for big files (several Gb)
Feb 16 17:16:53   CarlesFernandez: we _might_ be able to host them on gnuradio.org. Have to see about space issues.
Feb 16 17:18:33    oh, that would be great. Thanks for the try!
Feb 16 17:18:52   I have to run soon, but follow up with me in email about this (so I don't forget...)
Feb 16 17:19:02    CarlesFernandez: yes, i know, i am more interested in general implementation techniques of blocks
Feb 16 17:19:06    Josh Blum (SIP/xxx.xxx.xxx.xxx:4798-000000c8) has left the conference.
Feb 16 17:19:23    Josh Blum (SIP/xxx.xxx.xxx.xxx:32751-000000c9) has joined the conference.
Feb 16 17:20:14    thanks trondeau
Feb 16 17:31:28 *   jcorgan has quit (Remote host closed the connection)
Feb 16 17:31:48    Andrew Davis (SIP/xxx.xxx.xxx.xxx-000000c5) has left the conference.
Feb 16 17:35:19    Josh Blum (SIP/xxx.xxx.xxx.xxx:32751-000000c9) has left the conference.
Feb 16 17:35:34    Josh Blum (SIP/xxx.xxx.xxx.xxx:61176-000000ca) has joined the conference.
Feb 16 17:35:50    Tom (SIP/xxx.xxx.xxx.xxx-000000bf) has left the conference.
Feb 16 17:35:50     (IAX2/diamondcard-4411) has left the conference.
Feb 16 17:35:52    Josh Blum (SIP/xxx.xxx.xxx.xxx:61176-000000ca) has left the conference.
Feb 16 17:35:53    mait (SIP/xxx.xxx.xxx.xxx-000000c0) has left the conference.

Developer's Call, January 19, 2012[edit]

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items[edit]

  • Issues
    • Probably no discussion required:
      • Bug #479: resolved, nothing further to say
      • Feature #476: added
      • Bug #471: should be fixed; may need external verification
    • Issues to be Discussed
      • Bug #478: swigdocs file gen; fixed in nosync, good solution?
      • Bug #468: everything working for me now; any more problems?
      • Bug #463: only happens in this test case; in digital examples, I don't see it.
      • Feature #454: what's left to be done here?
      • Bug #422: can we add this quickly?
      • Bug #416: is this still relevant?
  • GR Conference
    • Are we ok to go with Atlanta.
    • Tom to provide latest news re discussions with TAPR and DCC
  • Google Summer of Code
    • Developing an idea for trying to improve integration into Eclipse; specifically, can we embed or recreate GRC in Eclipse for a more full user experience?
    • We'll need a mentor for this

IRC Log[edit]

 Starting Jan 19 conference call.
 Going over issues resolved and assigned.
 Bug #479: resolved, nothing further to say
 Feature #476: added
 Bug #471: should be fixed; may need external verification
 Note: look into automatically linking repo and issues.
 Bug #422: can we add this quickly?
 put the # sign before a bug to get redmine to link automagically
 jcorgan has already fixed this.
 Bug #416: is this still relevant?
 jcorgan also fixed this recently.
 Feature #454: what's left to be done here?
 do we not have the automatic script for this?
 Should just need a new tarball for this one. Maybe think about rewriting script a bit better.
 will close this; open a new one  to relook at making script better.
 Bug #468: everything working for me now; any more problems?
 Let's hold off on any resolution with this until we hear more.
 Bug #478: swigdocs file gen; fixed in nosync, good solution?
 not sure
 some how the makefiles are calling the target multiple times
 I've spent a while looking at it and can't figure out why
 So my fix uses a try/catch method to catch the errors and try again.
 I'll merge my branch and we'll see if anyone has issues with that.
 Bug #463: only happens in this test case; in digital examples, I don't see it.
 will check in with this again if I or anyone else can make sense of it.
 GR Conference
 TAPR and the DCC (digital comm conference).
 being held in Atlanta this Sept.
 Going to wait till TAPR gives me the full details. Should be any week now.
 If they find a reasonable package and hotel, we're likely to go with them.
 See if GA Tech has anyone working or interested in GR.
 Google Summer of Code
 Developing an idea for trying to improve integration into Eclipse; specifically, can we embed or recreate GRC in Eclipse for a more full user experience?
 Crofton|work to start keeping us on task for getting a SoC going.

Developer's Call, December 15, 2011[edit]

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items:

  • Volk alignment issues - checking
    • What's left before merging gr-blocks into master?
  • Roadmap for v 3.6
  • Bugs, features, and issues to close or to be closed
  • Conference 2012 preliminary discussion

IRC Log:

<mait> voip-gr: 51 is mait<br />
<voip-gr> mait is now associated with SIP/<br />
<Crofton|work> http://www.wired.com/threatlevel/2011/12/internet-suitcase-dc/<br />
<trondeau> Discuss volk alignment and merge gr-blocks.<br />
<voip-gr> Josh Blum (SIP/ has left the conference.<br />
<Crofton|work> quitter<br />
<guruofquality> every 15 mins<br />
<voip-gr> Josh Blum (SIP/ has joined the conference.<br />
<trondeau> jblum/gr-basic looks good. Should be ready to merge.<br />
<trondeau> We need to fix the alignment issue instead of using output_mutliple. Can fix this later.<br />
<voip-gr> <unknown> (SIP/ has joined the conference.<br />
<lindi-> voip-gr: 53 is lindi-<br />
<voip-gr> lindi- is now associated with SIP/<br />
<lindi-> did you just finish? can't hear a thing<br />
<trondeau> talking about some Volk test issues failing on different machines/archs. ORC had some issues -- debian unstable seems to work fine.<br />
<trondeau> lindi-, still talking.<br />
<mait> Debian squeeze has liborc 0.4.6<br />
<lindi-> ok, too tired to start debugging sip :(<br />
<trondeau> looking at resolved/fixed bugs to see about closing them.<br />
<mait> Debian unstable has liborc 0.4.16<br />
<trondeau> issue #467<br />
<trondeau> Josh fixed, merged to master. Closing.<br />
<lindi-> re orc I have one patch which I didn't send yet: http://lindi.iki.fi/lindi/gnuradio/0001-cmake-require-liborc-0.4.9-since-we-need-.floatparam.patch<br />
<trondeau> issue #466<br />
<lindi-> without this you get build failures on e.g. ubuntu lucid.<br />
<Crofton|work> lindi-, you on the call?<br />
<Crofton|work> ah, Isee<br />
<lindi-> Crofton|work: yes but doesn't seem to be working<br />
<Crofton|work> sorry bout your sip issue<br />
<trondeau> Minor issues; fixed and merged. Closing<br />
<trondeau> Still open. Need to look at how history deals with this.<br />
<trondeau> That was for issue #462.<br />
<lindi-> Crofton|work: no problem. I usually use mumble for conference calls<br />
<lindi-> Crofton|work: works bit more reliably since there's only one client and only one tcp connection<br />
<trondeau> issue #460. Fixed in master.<br />
<lindi-> I do have tcpdump of this so maybe I can decode your conversations later with scapy<br />
<trondeau> issue #443: fll buffering.<br />
<trondeau> I need to focus on this soon.<br />
<voip-gr> Josh Blum (SIP/ has left the conference.<br />
<trondeau> May want to reintroduce the old way (costas -> mpsk_receiver) until we fix this.<br />
<Crofton|work> 15 minutes is up<br />
<HokieTux> argh<br />
<voip-gr> Josh Blum (SIP/ has joined the conference.<br />
<trondeau> issue #446. Stuck in swig file to resolve namespace issues. Merged. Closing.<br />
<trondeau> issue #373. aligned malloc in mingw<br />
<trondeau> fixed a long time ago. Just closing it out.<br />
<trondeau> issue 203: fixed with update to UHD. closing.<br />
<trondeau> issue #151: fixed in cmake with shebanging.<br />
<trondeau> closing<br />
<trondeau> Try to make it a (loosely enforced) policy to write 2.7 Python code that is correctly translatable to Python 3.0. Makes the future conversion easier.<br />
<trondeau> conference 2012<br />
<trondeau> Maybe in Philly again next year or possibly with TAPR/ARRL DCC in Atlanta. Mid September most likely.<br />
<trondeau> Last issue is v3.6<br />
<jcorgan> sorry, have to take a customer call right now<br />
<trondeau> Right now, gr-blocks, Python blocks<br />
<voip-gr> <unknown> (IAX2/diamondcard-7633) has left the conference.<br />
<trondeau> they will go into 3.5.x<br />
<voip-gr> Josh Blum (SIP/ has left the conference.<br />
<trondeau> discussing swig 2.0 issues.<br />
<trondeau> We haven't made any effort to support 2.0, so I'm not surprised it's not working.<br />
<trondeau> Or at least that there are issues.<br />
<voip-gr> Josh Blum (SIP/ has joined the conference.<br />
<trondeau> Do we focus on getting 2.0 to work?<br />
<voip-gr> <unknown> (SIP/iptel.org-00000056) has joined the conference.<br />
<Crofton|work> rofl<br />
<trondeau> Wait until ubuntu/fedora make the move to make swig the official package.<br />
<voip-gr> <unknown> (SIP/iptel.org-00000056) has left the conference.<br />
<trondeau> need to have cmake do a swig version check (swig -version >= 1.3.31).<br />
<mait> Debian stable has swig 1.3.40<br />
<mait> Debian unstable standard swig is 2.0<br />
<trondeau> maint, that shouldn't be a problem right now. When is unstable going stable? (that is, how long do we have?)<br />
<mait> I don't know. But I am trying to build for unstable now.<br />
<voip-gr> Andrew Davis (SIP/ has joined the conference.<br />
<trondeau> 3.6 roadmap: remove autotools (already been done on next).<br />
<trondeau> support for inplace blocks.<br />
<trondeau> new/improved message passing infrastructure.<br />
<trondeau> bye bye

= Developer's Call, November 17, 2011 =

== Annotated agenda ==

Time: 2200 UTC, 6 PM EDT, 3 PM PDT, etc.<br />
SIP call: sip:gnuradio@digitalbazaar.com<br />
IRC channel: #gnuradio on Freenode

Agenda items:

* Documentation swigging through Python (almost there, just need to top it off)
* Volk alignment issues
* Bugs, features, and issues to close or to be closed

IRC Log:

<pre> worked on more_swigness_docs to complete docstring work
 jblum/volk_work should be ready to merge
 jblum/cmake_packaging branch is a cleanup of packaging effort.
 Talking about how to host deb and yum repos. CPack in cmake might be working better to build proper deb packages.
 Might be able to work better with Debian package managers to get newer version in their repo. We would also host a repo on gnuradio.org with more up-to-date versions.
 volk alignment issues
 I appeaerd
 just do not miss a branch prediction :)
 premature optimization is the root of all evil
 I don't think this is premature. We have implementations, and now I'm thinking about timing and testing and then optimization.
 But yes, I agree with you.
 I would love not to hanlde tail cases
 he fell off
 what about letting a work function say I processed N itmes and returning the ones it did not process?
 You could do that, too. Definitely the simplest.
 What if you have no more data to be processed after that, so you never get enough in your buffer to make the volk call?
 yeah, need to think
 but we know that systems like certain block sizes
 mod % operation?
 that is evil
 it is basically a divide and burns a hole in the cpu
 we can make that a mask
 with &
 Looks like I am getting two streams properly now, after applying the settings to all channels instead of just 0. 
 Simple mistakes are much better. 
 Crofton|work, yes, can you think of another way to get that value? May be able to mask it, actually.
 no worries
 we should force the alignment to be a power of 2 and use a mask for it instead.
* Crofton|work beats the crap out of trondeau and guerby 
 er guruofquality 
 Crofton|work, :)
 ther are taling about long doubles on the call
 which would hose the arm 
 We are not serious about the long doubles!
 they are, they are lying!
 Too many long doubles being processed. The call could not keep up.
 I think my audio in a fialing
 tails suck
 For volk tails, we tell the scheduler to give us N items but actually make sure N+n are available where n is the minimum number of items that a vector instruction can process.
 So the volk call will process all the items as though they were vectors, but the last n could be junk and not processed. Everyone will just ignore those.
 discussing FLL issues
 still there
 damn it
 I need to look at the FLL to see what's going on. No reason that I can see why it should be a problem
 my audio in is not working
 this is bug #463 for the record.
 Talking about gr-basic and gr-filter. Some more volk work for gr-filter.
 gr-basic is pretty much ready for use; want to get the Volk alignment issues handled first before we merge it.
 i'll find some time to fix the broken blocks and alter test criteria for blocks that work  but produce different results
 action items: I'll look into swig docs; merge jblums branchs (mentioned in log earlier)
 guruofquality will work on volk alignment issues. bistromath on volk unit tests.
 once those are done; we should be able to merge gr-basic.
 we should make an agenda for the gnuradio uses meeting
 and we will be drunk
 Discussing GNU Radio users meeting at WinnComm'11
 Location TBD, but at 8PM in the Hyatt as usual.
 Let's talk about getting more people to commit work
 Crofton|work agree
 there are all these people claiming to run gnuradio on their hardware
 We can use some of Corgan's presentation from the conference for this.
 tunnel.py must die
 CORBA yay
 son of corba
 one of the guys wrote the corba book i used
 ICE! for remote command/control of GNU Radio.
 Sam knows him
 I have a webpage that can control a digital transceiver.
 I'll send out an email about the agenda for the user's group and some proposed talks/structure.

Developer's Call, October 20, 2011[edit]

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EDT, 3 PM PDT, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items:

  • Bugs and Issues
    • Discuss current state of Issues assigned for 3.5 (UHD examples, cmake progress)
    • File sink issue (#448)
  • Progress with 3.5 release
  • Preliminary discussion of 3.6

IRC Log:[edit]

<trondeau> Issues to look into closing: 118, 373, 395, 403, 424, 425, 428, 445, 447
<trondeau> Issue: 118
<voip-gr_> HokieTux (SIP/ has left the conference.
<voip-gr_> Josh Blum Wireless (SIP/ has left the conference.
<voip-gr_> Lg Conference Room (SIP/ has joined the conference.
<HokieTux> voip-gr_, 00b2 is like 4 people
<Crofton|work> can we close #415
<Crofton|work> what number are we talking abuot now?
<trondeau> Issue 118: closed
<trondeau> On to issue 373
<trondeau> Something on Macs that was an issue.
<trondeau> Skipping any decision on that.
<trondeau> Issue 395
<trondeau> comedi compiles again; no one has been able to use it, though.
<trondeau> Going to close the ticket; whoever needs it can open another ticket.
<trondeau> Issue 403
<trondeau> libusb is not a dependency in GNU Radio anymore. Closing the ticket.
<trondeau> Issue 415
<trondeau> Build on ARMEL
<trondeau> Crofton|work, anything to say about this issue?
<Crofton|work> hello?
<Crofton|work> do you guys here me?
<bistromath> no
<trondeau> No
<bendnscom> bistromath: I see two recent updates I don't have. I will try those first. Thanks.
<HokieTux> Crofton|work, we haven't been able to hear you the whole time
<voip-gr_> Philip Balister (SIP/ has left the conference.
<Crofton|work> I am goign to try to reconnect
<voip-gr_> Philip Balister (SIP/ has joined the conference.
<trondeau> Closing 415, we think it's not a problem any more.
<trondeau> Issue 424 now.
<Crofton|work> [balister@moose gnuradio (josh-next)]$ grep -r sync_fetch_and_add_4 *
<Crofton|work> [balister@moose gnuradio (josh-next)]$
<trondeau> The tags in UHD sources were added with an example using it.
<trondeau> Closing 423 and 424
<trondeau> Issue 425
<trondeau> this is handled in GRC now.
<trondeau> closing it.
<voip-gr_> Lg Conference Room (SIP/ has left the conference.
<trondeau> Issue 428, was fixed yesterday. Can now enable/disable components under cmake.
<Crofton|work> get back here
<trondeau> Issue 445, swig exception handler
<Crofton|work> hey
<trondeau> Some swig handling code is included in gnuradio.i
<Crofton|work> get back here
<trondeau> What's up?
<Crofton|work> trondeau, call back
<Crofton|work> we cannot hear you
<bistromath> hahahahaha oops
<bistromath> we thought you dropped, not us
<trondeau> be more clear next time.
<Crofton|work> no
<trondeau> Please?
<voip-gr_> Lg Conference Room (SIP/ has joined the conference.
<Crofton|work> jcorgan, and I are talkign shit about you
<trondeau> Issue 447: removing usrp and usrp2, done yesterday.
<trondeau> Closing
<trondeau> Need to have cmake use version.sh to set the version in cmake instead of git describe.
<guruofquality> from distutils import sysconfig
<guruofquality> print sysconfig.get_python_lib(plat_specific=True, prefix=)
<trondeau> Figure out how we decide to use site-packages or dist-packages; autotools and cmake are doing it differently now.
<trondeau> Push off the Python docstrings to 3.5.1.
<trondeau> Todo for 3.5: finish gr-noaa and gr-pager conversion to Uhd; finish up the minor differences left between cmake and autotools.
<trondeau> Expect to have a 3.5.0rc1 out Saturday.
<voip-gr_> Lg Conference Room (SIP/ has left the conference.
<jcorgan> you guys just dropped
<Crofton|work> get back here
<voip-gr_> Lg Conference Room (SIP/ has joined the conference.
<bendnscom> bistromath: thanks, upgrade seemed to have worked.

  • bobster has quit (Quit: Leaving)

<trondeau> Ideas for 3.5.1: add docstring work and Python grblocks
<trondeau> We should also start to be able to add Volk functions to blocks.
<trondeau> I want to have Volk handle the type conversion blocks. That's an easy set of examples, but it could also make a nice dent.
<trondeau> Add gr-filter in 3.5 but don't remove filter from gnuradio-core. Remove the gnuradio-core for 3.6.
<trondeau> Move more from gr-digital out of Python and into C++ (packet utils, etc.)

Developer's Call, August 18, 2011[edit]

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items:

  • Conference
  • Status of cmake
  • Status of next branch and current work

Developer's Call, July 21, 2011[edit]

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items:

  • Conference
  • Current work in 'next' branch
    • new gr-vocoder
    • new gr-digital (working on adding to this)
    • Future additions: gr-filter, gr-signals
  • USRP support
    • looking to remove usrp, usrp2, gr-usrp, and gr-usrp2
    • Need to move all examples to gr-uhd
    • want to decouple GR examples from USRP - switch in different source/sinks
  • Anything else for release 3.5?

IRC Log[edit]

<bistromath> wooo

  • yootis (~matt@108-65-78-2.lightspeed.sntcca.sbcglobal.net) has joined #gnuradio
  • guruofquality (~jblum@108-65-78-2.lightspeed.sntcca.sbcglobal.net) has joined #gnuradio

<Crofton> you know how this works!

  • ttsou|call (~ttsou@108-65-78-2.lightspeed.sntcca.sbcglobal.net) has joined #gnuradio

<csboyer_> one ping only.....
<trondeau> Discussing conference
<yootis> There are still a few open slots for people to give talks at the conference
<yootis> Ettus Research will pay your conference registration fee if you are a student making a presentation
<trondeau> A few more slots are open for speakers
<trondeau> Need to advertise more
<yootis> For the hams out there -- TAPR DCC is on the weekend right after the conference, and it is in Baltimore, which is pretty close by
<trondeau> Need ideas for a topic in the panel session on Friday
<yootis> Much of the really interesting SDR work out there revolves around how to use all these interesting new computational devices
<yootis> GPU, DSP, FPGA, massive multicore with simple processors, large clusters of low power mobile phone processors
<ttsou|call> there are interesting issues with dsp, which I'm sure apply to other processors as well
<ttsou|call> external memory access for one
<Crofton> I registered
<voip-gr> Conference Room (SIP/xxx.xxx.xxx.xxx:35633-0000008c) has left the conference.
<bistromath> oops
<trondeau> we'll wait.
<voip-gr> Conference Room (SIP/xxx.xxx.xxx.xxx:41769-0000008d) has joined the conference.
<yootis> I think a panel session on the future of interesting processors and heterogeneous processing would be good
<yootis> Plenty of room for more attendees
<yootis> Ettus Research will have 5-6 people there.
<ttsou|call> academic year starts in about a month
<Crofton> update web page so people find the schedule
<yootis> One of the Joshes will be signing autographs
<Crofton> WIll Matt Ettus be signing USRP's
<yootis> Sure. Wouldn't be the first time. It might be the second, though.
<yootis> Will there be a big VT contingent attending?
<ttsou|call> any idea on the timeline for gr-digital?
<trondeau> A few from VT have signed up or expressed interest.
<Crofton> Al?
<ttsou|call> gr-dsp -> C64x
<ttsou|call> gr-dsp -> gr-c64x
<ttsou|call> or gr-dsp -> gr-dsplink may be more appropriate
<yootis> ahem... cmake... ahem...
<bistromath> i heard about this new build tool called cmake
<bistromath> phil nobody can fix your problems
<yootis> I hear it fixes everything and comes with a free pony
<Crofton|work> Action for yootis, work on cloning Josh again
<yootis> The clones never come out quite the same. Like copying a VCR movie.

  • xpot-mobile has quit (Quit: I'm Out, Later.)

<Crofton|work> hitting mute to avoid laughing into phone
<bistromath> hey the underwater sound went away
<bistromath> trondeau: while you're pulling josh's cmake work in, the volk_mutex branch of git://github.com/bistromath/gnuradio.git has a Volk fix, you can cherry pick 414d921a
<Crofton|work> unmute
<Crofton|work> mute
<voip-gr> Conference Room (SIP/xxx.xxx.xxx.xxx:41769-0000008d) has left the conference.
<Crofton|work> dorks
<voip-gr> Conference Room (SIP/xxx.xxx.xxx.xxx:48511-0000008e) has joined the conference.

Developer's Call, June 16, 2011[edit]

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items:

  • Discuss development plans
    • Merge jblum's cmake branch into 'next' and have parallel cmake and autotools build systems
    • Merge trondeau's 8psk branch from github into 'master' to add gr-digital
    • What is the roadmap for 3.4.1?
    • What is the roadmap for 3.5?
  • Conference
    • Plans are falling in to place
    • Registration will be open soon
    • Need to fill out schedule
      • Who wants to give a presentation?
      • Who should we invite to present?
    • Not all sessions will be standard presentations; what else do we want to see?
      • Think "foo/bar camp" style sessions
  • Website update
    • About to roll out updated website
    • New versions of everything, including Redmine (hopefully more stable, more features)
    • Current git and website will be locked for a few hours while everything is transitioned
    • Talk about getting rid of our git server and moving entirely to github. RFC here.

Developer's Call, May 19, 2011[edit]

Johnathan Corgan ran this call as Tom Rondeau was unavailable.

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

Agenda items:

  • Begin work on next for GNU Radio 3.5:
    • Add cmake build structure.
      • We want to start exploring using cmake instead of autotools. We should be able to have parallel build systems to work with until we figure out which one "wins" and then remove the other. This might take a few version cycles.
    • add gr-digital. I've been working on some enhancements to some of the digital modulation blocks. We will add this to gr-digital and deprecate their counterparts in gnuradio-core. This is the first step in splitting up the core into more appropriate components.
    • adding uhd examples/apps. Start relying less on USRP code for all our examples; have gr-usrp/apps, gr-usrp2/apps, and gr-uhd/apps. I made some start on this and more needs to be done.
    • Work on making a consistent API and naming scheme. Based on M. Dickens' suggestion. As part of the coding standard guide, we should make the block names and certain aspects of the API for each block clear (http://gnuradio.org/redmine/wiki/gnuradio/Coding_guide).
    • Start using Volk for gnuradio-core blocks. Need to ensure it works on the E100 and then start integrating Volk into blocks to improve efficiency of our code.
  • Some updates on the GR conference.

IRC Log[edit]

<jcorgan> ok first up is a 3.4.0 status
<Crofton|work> logging
<jcorgan> in short, we're ready to tag and release
<Crofton|work> crap
<Crofton|work> raw log sucks
<jcorgan> there a little matter of the release notes for a 2 year dev cycle, though :)
<jcorgan> both tom and I have some rough copy for the major API changes from 3.3->3.4
<jcorgan> but they're not what you'd call release ready
<douggeiger> Does there need to be a separate 'Porting your application to the 3.4.x API' wiki page or something similar?
<jcorgan> still, the 3.4.0 tarball will be up tomorrow or Saturday
<jcorgan> doug: probably a good idea
<jcorgan> it's not like there is a lot of breakage to contend with
<jcorgan> most of the changes occurred when we merge the previous 'next' branch into master, there didn't seem to be many issues
<jcorgan> I think a lot of the people that were using git had already ported to the next branch anyway, though
<Crofton|work> a result of a two year dev cycle
<jcorgan> yeah, let's not repeat that
<jcorgan> speaking of dev cycle, 3.4.x and 3.5 will be more granular
<jcorgan> I expect that most new development will happen on the master branch
<jcorgan> and we'll be able to pinch off 3.4.1, 3.4.2, 3.4.3, more frequently
<Crofton|work> so 3.4 will have a maintenance branch and we will develop against master?
<jcorgan> while the new 'next' branch will only be for things that require deep surgery
<jcorgan> not quite
<jcorgan> i can go over the plan here if people want
<Crofton|work> ok
<douggeiger> I think that would be good
<jcorgan> ok
<jcorgan> the basic idea is the release in the master branch won't ever break someone's source code
<jcorgan> so if you start writing an app for 3.4.x, none of the new features in 3.4.1, 3.4.2, 3.4.3, etc., would force you to go back and change your app code
<jcorgan> so we can develop most new capabilities in master, as long as they don't "change the API"
<fischer_1> seems reasonable ;-)
<Crofton|work> so how do we handle API changes?
<jcorgan> really it's to allow people to safely update to newer releases
<jcorgan> it's actually how the 3.0, 3.1, 3.2, and 3.3 series have been managed for many years now
<jcorgan> this last cycle, during the 3.3 release series, we let things hang out there too long on the next branch
<jcorgan> and it got to the point where all the new development had to happen on next
<jcorgan> that wasn't supposed to be the way it worked
<fischer_m> yeah, but looking at the refactoring part of the roadmap, a lot will happen on next this time again, if you don't want to break stuff on master
<jcorgan> correct
<jcorgan> but i want to pay close attention to what is happening on next so we can more easily merge it back into master and cut 3.5
<jcorgan> in the meantime, bug fixes, new blocks, new examples, etc., can all go onto master
<jcorgan> and become part of 3.4.1, 3.4.2, 3.4.3, etc.
<jcorgan> and as usual we'll keep merging up master into next frequently, to identify merge issues right away
<jcorgan> git makes doing this quite easy
<fischer_m> yeah
<Crofton|work> git makes everything easy ....
<jcorgan> we'll move on to actual 3.5 plans in a minute, but I want to make sure everyone gets a chance to ask about our workflow, since we don't talk about it much
<fischer_m> I have some questions for the 3.5 plans that are not fitting in the schedule ;-)
<jcorgan> balister: to answer your question, API changes must happen on next
<jcorgan> ok, let's move on to 3.5 stuff
<Crofton|work> yay
<jcorgan> first on the agenda is...exploring a transition to cmake from autotools
<Crofton|work> yay, only because the gnuradio autofoo makes me crazy
<fischer_m> indeed ;-)
<guruofquality> so I have done some legwork here: https://github.com/guruofquality
<jcorgan> I knew it became a problem when I googled for advanced autotools usage, to help debug a gr build issue, and I started getting links back to gr itself
<guruofquality> thats special
<fischer_m> haha
<guruofquality> google does tailor its searches now
<Crofton|work> yeah
<Crofton|work> the gnuradio autofoo was written by people that know to much
<fischer_m> I managed to port our spectral estimation toolbox to cmake in more or less one evening including linking fortran code, and before I had no idea about cmake ... very much in favor of cmake now
<guruofquality> http://4.bp.blogspot.com/_iWTpxPfBguU/THxh6DZpwgI/AAAAAAAAEWA/q8eYZSzhQ8g/s1600/AutoTools.png
<Crofton|work> guruofquality, which branch has your cmake work?
<jcorgan> of course a major advantage of cmake is the Windows support...
- Crofton|work pukes
<guruofquality> its on my github
<Crofton|work> in master
<Crofton|work> ?
<Crofton|work> wb, douggeiger
<guruofquality> yea so I went through the effort of getting the code compiling under windows while I was at it
<guruofquality> yea master branch
<Crofton|work> we should not have branch name conflicts when we use github ...
<fischer_m> ;-)
<jcorgan> Josh's branch has some initial cmake coding for a few top-level directories
<guruofquality> git clone https://guruofquality@github.com/guruofquality/gnuradio.git
<Crofton|work> git remote add ......
<jcorgan> and some changes to GNU Radio itself to work under MSVC
<guruofquality> gruel, audio, qtgui, grc, and gnuradio-core (which is massive)
<guruofquality> I have backported many of the changes for MSVC support (minis the API exports) and they have since been merged
<Crofton|work> I'd like to be able to test the cmake builds in OE
<Crofton|work> asap
<guruofquality> sure, have fun, mind you the build system might need a tweak here and there
<guruofquality> I have only tested this one direcly on windows and linux
<jcorgan> what I propose is to push that branch to gnuradio.org, but keep it separate from next for now
<guruofquality> have not tried mac, cygwin and mingw
<jcorgan> with the idea that we'd build up to replicating the entire current build in cmake on that branch
<Crofton|work> can we split cmake from windows branchwise?
<Crofton|work> topic branches is my thinking
<guruofquality> sounds reasonable, I'd like to squash my changes and merge that over in steps
<guruofquality> oh yea, we can split
<guruofquality> the windows changes are the API exports, I want to keep that in one commit, and the cmakelists as their own commit
<jcorgan> separate branches for cmake and windows would be great, but that takes more work and discipline from Josh and anyone else working that
<guruofquality> is that what you mean by separate branches?
<guruofquality> http://gcc.gnu.org/wiki/Visibility
<Crofton|work> pretty much
<jcorgan> we have to careful about commit flow, though
<guruofquality> the api export things, I would really like to have them as well since they are useful outside of MSVC as well (hence the link)
<guruofquality> sure
<Crofton|work> this way I can test cmake with out fallig over windows :)
<guruofquality> its not an issue, really
<guruofquality> there is no extra "windows" code in there
<jcorgan> i think the GR changes to accommodate MSVC have been pretty benign
<guruofquality> you can easily remove them with a sed script
<guruofquality> harder to add them
<guruofquality> we should be well defining the symbols we export from the API (IMHO)
<Crofton|work> k
<Crofton|work> next ...
<fischer_m> may I interrupt with one short question, as it's getting quite late over here?
<jcorgan> go ahead
<fischer_m> I was wondering whether there's a special reason we stick with cppunit
<jcorgan> inertia
<fischer_m> as we're already heavy in dependencies, and anyway link boost, I was thinking we could use boost-unittest
<bistromath> the new volk uses boost::unit_test
<jcorgan> i'm sure josh will have it coded up by tomorrow
<fischer_m> jcorgan: haha
<guruofquality> we could replace cppunit w/ boot unit test
<Crofton|work> HokieTux, says nice things about boost::unit_test
<bistromath> having that dependency isn't exactly super heavyweight but yeah, cppunit can be replaced
<guruofquality> boot unit test is nice, the binaries have very useful options for logging and verbose
<bistromath> so, moritz: are you volunteering? ;)
<Crofton|work> I wonder how they work cross?
<jcorgan> not opposed, but it's a lot of work
<fischer_m> ummm I'm quite busy at the moment with exams and stuff
<fischer_m> but I'll have a look
<guruofquality> mostly, you can replace CPPUNIT_CHECK w/ BOOST_CHECK
<guruofquality> I did this partially on a branch and gave up
<douggeiger> Just a quick wave to all - I'm boarding my flight, take care everyone
<Crofton|work> l8r
<bistromath> yeah last time i looked at it i figured you could wrap CPPUNIT_CHECK into a BOOST_CHECK macro and get fairly close that way
<Crofton|work> so we would not be opposed to people using boost unit test for new work?
<jcorgan> so exploring cppunit->boost-unittest is something to add to the todo list
<bistromath> well, i'd hate to rewrite the Volk test suite...
<fischer_m> yeah, I agree, it's a lot of work, just it was something that came to my mind playing around with CTest
<jcorgan> it's a lot of work for only moderate benefit, i'd put it fairly low on the priority scale
<guruofquality> definitly not working: http://gnuradio.org/cgit/jblum.git/log/?h=wip/use_boost_unit_test
<jcorgan> let's move on
<fischer_m> ok
<jcorgan> but it's at least worth exploring, we may get surprised
<jcorgan> Tom has a branch where he has implemented a new top-level component, gr-digital
<fischer_m> alright, I'll go to bed ... gn8 everyone
<Crofton|work> gn
<jcorgan> Right now, our digital comms stuff is scattered all over the place
<jcorgan> C++ blocks in gnuradio-core, python stuff in blks2, apps in gnuradio-examples
<Crofton|work> I am a huge fan of organizing all the application stuff
<jcorgan> the idea is to migrate all these things into the new top-level component
<jcorgan> and at the same time, use the opportunity to clean up how we do some things
<Crofton|work> I see a coupe of issues
<Crofton|work> things in digital may dpend on other top level groups
<Crofton|work> like filters
<jcorgan> gr-digital would still depend on gnuradio-core
<bistromath> would filters be part of gnuradio-core?
<Crofton|work> so the "configure" will need to see dependencies between top level stuff
<Crofton|work> I hope not
<bistromath> or would -core be strictly what "runtime" is now?
<jcorgan> this is actually step 1 of a multi-step reorganization
<Crofton|work> sure, I just want to make sure we are moving to a world where the core is limtied to block infrastructure and no actual blocks
<jcorgan> yes, i actually did this on a branch about a year and a half ago, where i separated runtime out from gnuradio-core into its own gr-runtime
<jcorgan> but we're getting ahead of ourselves
<bistromath> so, philosophically, what do you envision gnuradio-core containing?
<jcorgan> I see most of gnuradio-core being eventually refactored into smaller things
<Crofton|work> the existing core
<jcorgan> yes
<jcorgan> it's become a dumping ground for all new blocks, and we really should have more logical organization of things
<jcorgan> the first step that Tom and I are working is to hoist out all the digital comms stuff into a gr-digital
<Crofton|work> I sort of liek the name runtime for the block structure, connection structure and the scheduler
<jcorgan> agree
<Crofton|work> and it avoids the overlap with the word core
<jcorgan> I actually had that working at one point for C++ only, but most blocks became dependent on gr-runtime and not gnuradio-core
<guruofquality> are we planning a gr-runtime?
<jcorgan> we were at one pont
<jcorgan> point
<guruofquality> would like a gr-runtime, and gr-core-blocks that just implement data type, stream manipulations and simple math
<jcorgan> but for really painful reasons, it wasn't possible to have a single gr namespace that spanned gr-runtime and gnuradio-core
<jcorgan> though for C++ only coding it was great
<guruofquality> so it was a problem for python?
<jcorgan> yes
<Crofton|work> there will be pain, but I think it will really help people understand guradio
<guruofquality> is this because of swig?
<guruofquality> I think we can do this
<jcorgan> no, it boiled down to not being able to construct the gr namespace from to different directory hierarchies
<jcorgan> during 'make check' only
<jcorgan> it would have worked once installed into the system, as the files all get installed to the same place
>ttsou_mobile< ping
<guruofquality> I see how that could be, maybe we need to do it more differently
<jcorgan> well, one solution would be to give up the ability to test code prior to installing it
<guruofquality> or not having core blocks in the gr/g.runtime namespace
<jcorgan> that is, you'd compile gr-runtime, make install it, then you'd compile gnuradio-core, then make install it
<guruofquality> you know more namespaces for python
<Crofton|work> the other problem I see is doing this work in a dual cmake/autofoo situation
<jcorgan> yeah, it gets complicated
<guruofquality> gr.runtime gr.core gr.foo
<jcorgan> we should focus on one thing at a time
<guruofquality> sure, we should discuss sometime
<Crofton|work> my concern is the cmake/autofoo decision is blocking lots of work
<guruofquality> your right, we should just switch it all over to cmake immediately
<jcorgan> but Erics and my plan back then was to migrate to a gr-runtime that most blocks would depend on, with gnuradio-core having the most commonly used low-level blocks, and gr-digital having the higher level modulators, packetizers, etc.
<guruofquality> just kidding
<jcorgan> josh, you need to prove you can replicate the current autotools 100%
<guruofquality> Lol no problem
<Crofton|work> I'm just pointing out this is a critical item
<jcorgan> when the existence proof is in hand, we'll switch :)
<guruofquality> did I tell you how easily we can build the compoentized debs
<jcorgan> tom mentioned it, it sounds great
<guruofquality> sexy I know
<jcorgan> ok, let's move on
<guruofquality> okay well we will take our time and see what happens
<jcorgan> Another planned set of development is adding UHD based utilities and examples to the 3.4.x series
<jcorgan> Since adding stuff doesn't break anything, we can do all this on master
<Crofton|work> when should we make the old usrp stuff go away?
<jcorgan> But here is where git let's us do some cool tricks
<jcorgan> the idea is to add a new UHD based example on master
<jcorgan> then merge that up to next
<jcorgan> then delete the usrp one only on next
<jcorgan> if we work that way, once we get all the new uhd stuff added on master, on next we can then delete usrp/usrp2/gr-usrp/gr-usrp2
<Crofton|work> will 3.5 include the old usrp stuff?
<jcorgan> that would let us target 3.5 for having all the old stuff gone
<jcorgan> I don't think it's a great deal of work, and it's something we can get started on right away
<ptast> if usrp is removed, will the "old" TVRX & DBSRX work using uhd?
<jcorgan> tom has already made a uhd_fft and uhd_rx_cfile
<jcorgan> jost?
<jcorgan> josh?
<guruofquality> yea
<guruofquality> all dboards are in uhd
<guruofquality> :-)
<jcorgan> yay
<Crofton|work> even old ones that are no longer made?
<guruofquality> yea we love those boards just as much as the new ones
<Crofton|work> :)
<guruofquality> no favorites
<ptast> good
<Crofton|work> it is getting late here too
<Crofton|work> as in my wife came home
<jcorgan> one other thing we'll probably do during the process, is create the new uhd_* stuff in gr-uhd/apps, and move some of the newly ported gnuradio-examples/python/usrp/* stuff there as well
<guruofquality> wait, we have a debate about not calling them apps
<guruofquality> instead we call them examples
<guruofquality> tom thought it would make tech support easier
<jcorgan> ok, we haven't synced up on that yet
<guruofquality> okay sure
<guruofquality> an app is like a finished polished product
<jcorgan> i think i know a good way to handle that
<Crofton|work> an app runs on my phone
<jcorgan> gnuradio has no finished, polished applications :)
<guruofquality> then we should never see the word app :-)
<guruofquality> its all example
<guruofquality> okay uhd_fft.py can be an app
<guruofquality> tunnel.py is an example
<guruofquality> okay I give up
<jcorgan> okay, let's table it
<jcorgan> let's talk vold
<jcorgan> volk
<Crofton|work> voldemort
<jcorgan> lol
<bistromath> the lib-which-must-not-be-named
>bistromath< the authors how must not be named
<jcorgan> it's obviously in 3.4.0, but nothing in gnuradio is using it
<jcorgan> we need to get it to the point where it can compile generic code on the E100
<bistromath> jcorgan: i did that this afternoon
<jcorgan> excellent
<bistromath> not that it's tested but in principle it should go fine
<jcorgan> famous last words
<bistromath> =)
<Crofton|work> native and cross ....
<bistromath> either one, if you compile it with -mneon it should be fine
<jcorgan> then we can start migrating internals of gnuradio-core stuff to using libvolk routines
<jcorgan> and since it's all under the API, this can happen on master as part of the 3.4.x series and not have to wait until 3.5
<jcorgan> anyone see any issues with that
<bistromath> well
<bistromath> you'll have to pull the major volk changes into the master branch of 3.4.x
<bistromath> i don't have a problem with that but that's a fairly major chunk of changes
<jcorgan> where is it now?
<bistromath> i think the latest is my github
<bistromath> git@github.com:bistromath/gnuradio.git
<bistromath> volk_cmake branch
<bistromath> josh and i trade back and forth
<jcorgan> what is the state of the volk component on current master
<bistromath> lame
<jcorgan> okay, i'll tag 3.4.0 prior to looking at merging your new volk stuff
<bistromath> none of the volk_machine changes got pulled into that, so it can't cross-compile or do runtime machine evaluation
<jcorgan> the E100 will still need --disable-volk then for the gr build
<bistromath> yep
<Crofton|work> I just patch out the autofoo, --diable doesn't work
<bistromath> crofton: can you publish that patch
<jcorgan> i'd really like to know why --disable-volk doesn't work for you
<jcorgan> that's the whole point of having --disable-foo in the first place
<Crofton|work> http://git.openembedded.org/cgit.cgi/openembedded/tree/recipes/gnuradio/gnuradio-git
<Crofton|work> my motivation to work on the autofoo is low
<jcorgan> understood
<Crofton|work> since guruofquality will save us with cmake
<jcorgan> but solving that is needed before we can start using volk internally in GR
<Crofton|work> yes
<Crofton|work> well the native build works, which is even weirder
<bistromath> the native build SHOULD work
<bistromath> native e100 volk has worked for a long time
<jcorgan> alright, well, keeping patching it out of configure.ac for E100 for the 3.4.0 tagged release
<jcorgan> and we can merge in whatever fix you guys come up with after that
<Crofton|work> yep
<Crofton|work> no worries
<Crofton|work> OE is good at carrying fixes for annoying upstream issues
<jcorgan> ouch
<Crofton|work> The Yocto project crowd is getting motivated to upstream more patches though
<jcorgan> is OE/bitbake really the best way to do stuff for devices like the E100?
<Crofton|work> yes
<Crofton|work> http://www.youtube.com/watch?v=DqSy0KNAo2g
<jcorgan> i'll take your word for it
<Crofton|work> alot of big players backing the technology now also
<Crofton|work> moving on
<jcorgan> ok, last thing on the agenda is the GR conference
<jcorgan> tom's update to me was brief, but it looks like he'll have a paypal solution sometime next week so we can start getting attendanc estimates through registration
<Crofton|work> what about arranging talks etc?
<jcorgan> no update from him
<jcorgan> but I'm sure it's on his mind already
<Crofton|work> how many days are we looking at?
<Crofton|work> yeah
<jcorgan> 3 days of the event, followed by an optional 4th day as an open hacking day
<jcorgan> Wed-Fri, with hackfest on Sat
<Crofton|work> we are going to have to kick ass to fill 3 days :)
<jcorgan> ?
<Crofton|work> three days is a lot of talking
<Crofton|work> although this will be single track, so that helps
<jcorgan> my training classes are 3-4 days, and that's all just me blabbing
- Crofton|work is insanely curious to see how people come
<Crofton|work> yes, but I am assuming these will be more focused on interesting stuff :)
<jcorgan> we're planning for 40 attendees, plus Ettus/Corgan/Rondeau staff
<Crofton|work> ok, we done?
<jcorgan> with the agenda, yes
<jcorgan> happy to answer general Q&A for awhile if people want
<Crofton|work> if you want your questions in the log speak now :)
<Crofton|work> saving

Developer's Call, April 15, 2011[edit]

IMPORTANT UPDATE: We have decided to move the monthly calls from the third Friday to the third Thursday of the month. The next few months calls will be: May 19, June 16, July 21, August 18. We will probably not have a call in September due to the GNU Radio Conference.

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

  • Review of the state of GR 3.4
    • Outstanding issues? Windows, OSX, E100?
    • Volk?
    • Testing/testers
    • Timeline (maybe)
  • GR Conference
    • I will update on recent developments
    • Getting down to the wire to make a few decisions

== IRC Log
We had SIP trouble with this call, so the entire conversation was held on IRC. The full transcript: ==

Apr 15 18:02:18 <voip-gr> Tom (SIP/xxx.xxx.xxx.xxx-000000be) has joined the conference.
Apr 15 18:02:31 <trondeau> Anyone on the SIP bridge?
Apr 15 18:02:40 <voip-gr> Tom (SIP/xxx.xxx.xxx.xxx-000000be) has left the conference.
Apr 15 18:02:41 <jcorgan> can't seem to get on it
Apr 15 18:02:48 <guruofquality> its not just me
Apr 15 18:02:53 <guruofquality> :-)
Apr 15 18:03:05 <voip-gr> Tom (SIP/xxx.xxx.xxx.xxx-000000bf) has joined the conference.
Apr 15 18:03:06 <trondeau> I'm on.
Apr 15 18:03:09 <jcorgan> i've been saying that to myself for years :-)
Apr 15 18:03:22 <trondeau> Not sure if my mic is working.
Apr 15 18:03:37 <voip-gr> Conference connections: <unknown> (trapicki) [SIP/xxx.xxx.xxx.xxx:1026-000000a6], Josh Blum [SIP/xxx.xxx.xxx.xxx:8682-000000b8], Andrew Ge [SIP/xxx.xxx.xxx.xxx:5060-000000a5], Tom [SIP/xxx.xxx.xxx.xxx-000000bf]
Apr 15 18:04:05 <trondeau> It says Josh is on.
Apr 15 18:04:06 <guruofquality> why does it say I'm in there
Apr 15 18:04:43 <guruofquality> kill that connection
Apr 15 18:05:23 <trondeau> voip-gr: disconnect b8
Apr 15 18:05:24 <voip-gr> Josh Blum (SIP/xxx.xxx.xxx.xxx:8682-000000b8) has left the conference.
Apr 15 18:06:27 <voip-gr> Tom (SIP/xxx.xxx.xxx.xxx-000000bf) has left the conference.
Apr 15 18:06:33 <guruofquality> my my echo test works
Apr 15 18:07:13 <voip-gr> Tom (SIP/xxx.xxx.xxx.xxx-000000c1) has joined the conference.
Apr 15 18:07:35 <trondeau> Strange. When I change which mic input to use, it also changes my audio out... Is anyone on that can hear me?
Apr 15 18:07:36 <jcorgan> for whatever reason I can't get sip working from here. I'll stick around, though
Apr 15 18:07:56 <trondeau> Well this is turning into a disaster...
Apr 15 18:08:45 <trondeau> We really need to get a real phone number set up for this, but it's going to cost us some money.
Apr 15 18:09:38 <jcorgan> who is here on irc
Apr 15 18:09:44 <guruofquality> can you reboot the server
Apr 15 18:09:46 <trondeau> Maybe just work on IRC today
Apr 15 18:10:09 <trondeau> No, I don't have access to the server. I can email the guy who does, though.
Apr 15 18:10:53 <trondeau> manu1, are you around?
Apr 15 18:11:25 <trondeau> By the way, guruofquality and other Ettus people, great job getting the SBX out.
Apr 15 18:12:09 <jcorgan> perhaps those who are present are those who both care and can make a difference?
Apr 15 18:12:21 <guruofquality> cool thanks
Apr 15 18:12:54 * bistromath (~bistromat@108-65-78-2.lightspeed.sntcca.sbcglobal.net) has joined #gnuradio
Apr 15 18:13:06 <bistromath> are we still having sip difficulties
Apr 15 18:13:24 <guruofquality> so are we going to wait a bit for the technical support or just do this text based?
Apr 15 18:13:42 <trondeau> We are. I'm on, but not sure about my mic. No one is speaking that I know of, so I can't confirm any audio.
Apr 15 18:14:04 <trondeau> I say we do this text based today. I have to run at 7.
Apr 15 18:14:25 <guruofquality> when I installed ekiga it gave me an option to make a free account on some service, if you do that you can use the echo test, very helpful
Apr 15 18:15:17 <guruofquality> so whats on the order of business today?
Apr 15 18:15:21 <trondeau> Testing mic with Skype. Wish Twinkle had something similar.
Apr 15 18:15:35 <trondeau> Skype works.
Apr 15 18:15:44 <trondeau> Ok, anyway, not helpful. Need to move on.
Apr 15 18:16:15 <trondeau> First order of business, I want to talk about changing the time of these calls. More specifically, the day. How about Thursday's at 11 PM UTC?
Apr 15 18:17:09 <trondeau> It's just that every now and then, I have something like a life and it gets in the way of this stuff. For instance, I'm going to be at a wedding in Ireland during our next scheduled conference call.
Apr 15 18:17:39 <trondeau> I don't think I'll be much in a mood for this at 11 PM that night. Then again, it could make for some really interesting chatter...
Apr 15 18:18:08 <bistromath> that's fine, working hours whenever are OK for us, don't know about others
Apr 15 18:19:28 <trondeau> I'll send out an email about it and see. I think Balister will be happier at least. Now we don't have to hear his complaint about not having beer.
Apr 15 18:20:13 <trondeau> Ok, great. We'll tentatively move these calls to the third Thursday of the month unless we get a torrent of emails saying no.
Apr 15 18:20:29 <trondeau> Next up! We think we are about ready to release GNU Radio 3.4!
Apr 15 18:20:56 <trondeau> jcorgan and I have been talking a lot about this, trying to figure out the last things we need to get done for this. We think everything is worked out.
Apr 15 18:21:11 <trondeau> There's one outstanding issue in Windows that we have a patch for that we believe will work.
Apr 15 18:21:28 <trondeau> Also, bistromath, we're ok with --disable-volk for the E100 builds?
Apr 15 18:21:52 <bistromath> pretty sure, forgot to ask balister when he was in earlier this week
Apr 15 18:21:59 <bistromath> i'll email now
Apr 15 18:22:31 <trondeau> Thanks. Let us know what he says. Are you having any issues with it, or is there something different with what Balister is doing?
Apr 15 18:22:48 <bistromath> issues with volk on e100? he's the only one doing gnuradio compiles on e100 right now
Apr 15 18:24:13 <trondeau> So you're just using what he's building then?
Apr 15 18:24:49 <bistromath> yep
Apr 15 18:25:13 <trondeau> Gotcha. Ok, we'll wait to hear what he says.
Apr 15 18:26:09 <trondeau> Anyone else have anything they think we need to do for 3.4?
Apr 15 18:26:40 <guruofquality> I have a branch uhd pkg config file
Apr 15 18:27:16 <trondeau> What does it solve and is is ready to merge?
Apr 15 18:27:41 <guruofquality> yea, we didn't have a pkg config file for gnuradio-uhd
Apr 15 18:27:49 <guruofquality> we usually make one for all libraries
Apr 15 18:28:25 <guruofquality> minor thing
Apr 15 18:29:29 <trondeau> Ah, yes, good, thanks!
Apr 15 18:29:49 <trondeau> I'll merge it now, distcheck, and maybe even have it pushed before the end of the chat.
Apr 15 18:30:14 <trondeau> Though I don't see any problems from the diff; just being safe with the distcheck.
Apr 15 18:30:36 <trondeau> Anything else?
Apr 15 18:31:15 <guruofquality> my high res timer branch is for qtgui is not going to merge nicely, are we not doing that anymore?
Apr 15 18:31:52 <trondeau> Yes, I know. I should have gotten that done before I made all of the other changes, but didn't think about it until it was too late.
Apr 15 18:32:27 <trondeau> I think what you've done is still the right way to do it. Let me fix this tomorrow. I think I can do it pretty cleanly and relatively quickly.
Apr 15 18:32:46 <guruofquality> okay, so its not misfeature? but maybe a "next" feature
Apr 15 18:33:05 <trondeau> I don't think it needs to be a next feature, really.
Apr 15 18:33:30 <guruofquality> so we are cutting a release, so my noisy branches should probably stay out of that
Apr 15 18:33:46 <guruofquality> so I'm happy, other than the gr-uhd pc file, I think its good
Apr 15 18:34:01 <trondeau> Some of the noisier ones, yes, but from what I recall, the high_res_timer wasn't too bad.
Apr 15 18:34:43 <trondeau> Then again, upon reflection, why don't we cut the release after the gr-uhd.pc file addition, but I'll add the high_res_timer stuff for 3.4.1
Apr 15 18:34:59 <trondeau> Just to be safe and not hold things up any longer.
Apr 15 18:35:04 <guruofquality> ah, well it wansnt so widely tested as what is in the currently master
Apr 15 18:35:07 <guruofquality> yes agreed
Apr 15 18:35:12 <trondeau> Great.
Apr 15 18:35:21 <trondeau> bistromath, do you want to say anything about your work with Volk builds?
Apr 15 18:35:26 <bistromath> sure
Apr 15 18:36:21 <bistromath> background for anyone who isn't in on it, i'm taking Volk and giving it a structure which will allow cross-compiling and running optimized functions on machines other than the one on which the lib is built
Apr 15 18:36:48 <bistromath> i got a lot of stuff done in the last 3 days, the code generation is basically done and now i just have to do some Automake wrangling to get it to play nice with autotools
Apr 15 18:37:18 <bistromath> i haven't tried x-compiling it yet so i'm sure there will be some bugfixes necessary
Apr 15 18:37:53 <bistromath> i should be done with it this weekend
Apr 15 18:37:58 <bistromath> so that's it
Apr 15 18:38:40 <trondeau> Great! Thanks!
Apr 15 18:38:45 <guruofquality> wheres the code published
Apr 15 18:38:49 <bistromath> oh right
Apr 15 18:38:54 <bistromath> github/bistromath/gnuradio.git
Apr 15 18:39:06 <trondeau> Hopefully another 3.4.1 addition.
Apr 15 18:39:48 <guruofquality> 3.4.1 sounds exciting
Apr 15 18:39:49 <trondeau> Once we confirm that this works and we can compile and use Volk on all of our main targets (like the E100, which I think is the main issue), we can start using Volk inside GNU Radio blocks.
Apr 15 18:40:47 <guruofquality> this is exciting
Apr 15 18:40:50 <trondeau> I know of a number of places we can make significant gains with Volk in qtgui, and I've already seen a 10% improvement by vectorizing the internal multiplication of the FFT filters with Volk.
Apr 15 18:41:08 <trondeau> Yeah, this is going to be really nice.
Apr 15 18:41:17 <bistromath> wow, over the SSE FFT stuff that was already in there?
Apr 15 18:41:23 <guruofquality> and even better, it seems to work in practice
Apr 15 18:41:32 <guruofquality> are you calling into FFTW?
Apr 15 18:42:12 <trondeau> It still uses the FFTW for the (i)FFTs, but then the taps and signal (post FFT) are multiplied in a for-loop. This was just replacing the for-loop with a Volk 32fc multiply.
Apr 15 18:42:28 <bistromath> cool
Apr 15 18:42:34 <trondeau> It's about as simple a change as you can think of and still made gains.
Apr 15 18:42:49 <bistromath> yeah the difference between what you're using now and what we'll have shortly is you won't do any volk_init before calling into volk
Apr 15 18:43:02 <bistromath> so you can just #include <volk.h>, and then link against volk.so
Apr 15 18:43:38 <trondeau> Yeah, that's great, too. This will be like a 2-line change to this code. One to include the file, one to call the vector multiply function.
Apr 15 18:44:17 <bpadalino> oh wow, lots of talk
Apr 15 18:44:27 <trondeau> Ok, I think we're all set on 3.4 for now; just waiting to hear back from Balister about the E100 issue.
Apr 15 18:44:37 <trondeau> I want to say a few things about the GR conference now.
Apr 15 18:45:12 * patrik (~quassel@av83.netikka.fi) has joined #gnuradio
Apr 15 18:45:27 <trondeau> The dates are set, Sept. 14 - 16 with an optional "Hack Fest" on the 17th.
Apr 15 18:45:39 * patrik is now known as Guest72298
Apr 15 18:46:09 <trondeau> But the rooms we have reserved on the Univ. of Penn campus only seat about 50 people, max.
Apr 15 18:46:53 <trondeau> I've had only about a dozen responses for people who want to make it; only a few have said that the definitely can. It's hard to judge from this the size we will actually get there.
Apr 15 18:48:06 <trondeau> Can anyone help me predict if 50 is too small or just about right?
Apr 15 18:48:45 <bpadalino> judging from the activity of the irc channel, the mailing list, and throwing in a fudge factor - you'll probably be fine
Apr 15 18:49:00 <guruofquality> The conference at sdr forum always seemed to fill up more than expected
Apr 15 18:49:23 * Guest72298 is now known as ptast
Apr 15 18:49:37 <trondeau> That's what I'm thinking, too, guruofquality. But that's a) colocated in a place where interested parties already are and b) is free.
Apr 15 18:50:03 <trondeau> I think we will be charging a nominal fee (like $100) for attendance to pay for food and various other expenses.
Apr 15 18:51:13 <bpadalino> the 51st person can pay $200 then, and you can refund someone else's seat when you kick them out
Apr 15 18:51:33 <trondeau> Not a bad idea :)
Apr 15 18:52:40 <bpadalino> i'm still tentative, but equipment wise - any idea who might be bringing what?
Apr 15 18:52:52 <trondeau> I will also be blocking off a set of rooms at on of the Philly hotels. There's one in Center City that'a few blocks away (about a $6 cab ride in traffic) that's apparently nice and that UPenn has a great relationship with.
Apr 15 18:53:30 <bpadalino> that's a good idea
Apr 15 18:53:41 <trondeau> The other hotel is about a block and a half away, but more expensive and not in as nice a place (not a bad place, but Center City is where most of the shopping, restaurants, and bars are).
Apr 15 18:54:37 <trondeau> Not sure who's bringing what, yet. I imaging people will be bringing USRPs and a handful of other radios. I'm hoping to see at least one Funcube dongle represented :)
Apr 15 18:56:17 <trondeau> For food, we will be providing a small breakfast (basics like coffee, tea, some bagels and muffins) and lunch on Thurs and Fri. Wednesday we will be starting after lunch, so some small snacks around noon for when people arrive.
Apr 15 18:56:43 <bpadalino> do you think other test equipment might be useful? wondering if i should make a request in with work to bring down a vsg, vsa, etc
Apr 15 18:56:48 <trondeau> I hope we can also get something organized as a nice dinner out some night, like a conference banquet, but that needs to be established.
Apr 15 18:58:21 <trondeau> Interesting thought. If someone could bring a hand-held spectrum analyzer, then definitely. I wouldn't expect we need anything major, though.
Apr 15 18:58:40 <trondeau> Still, a good point, and I'll keep that in mind as we know more of who's coming, what they are bringing, and what we might want.
Apr 15 18:59:02 <jcorgan> i'll bring a spec an
Apr 15 18:59:05 <bpadalino> does the school have any equipment that might be able to be moved for the time ?
Apr 15 18:59:17 <trondeau> FYI, guruofquality, I just pushed your gnuradio-uhd stuff to master.
Apr 15 18:59:55 <trondeau> The school actually doesn't have much in the way of that kind of equipment. We have a number of USRPs that are in use on various projects, so may or may not be available during the conf.
Apr 15 19:00:17 <bpadalino> oscopes or anything ?
Apr 15 19:00:24 <trondeau> But the group I work with does not have access to good test equipment like that. Then again, they are more networking and computer security guys, really.
Apr 15 19:00:41 <trondeau> I will try to make friends with more EE types over the next few months and see what comes up.
Apr 15 19:01:54 <trondeau> We have a pretty decent IEEE presence in Philly, too. I will work on getting to know them more. Perhaps a few people and companies would be willing to lend some stuff. Less of a guarantee, but a good excuse to get to know some more engineers here.
Apr 15 19:02:09 <bpadalino> absolutely
Apr 15 19:02:59 <trondeau> Alright, it's 7 and I need to jet. There are still lots of logistics about the conference to work out, but these were the main areas I was hoping for some input on.
Apr 15 19:03:16 <bpadalino> enjoy your "life"
Apr 15 19:03:21 <trondeau> I will also try to capture this in an email and I hope to make a website for the conference information soon.
Apr 15 19:03:39 <trondeau> Yes, thanks. Sometimes I pay at being a real person.
Apr 15 19:03:46 <trondeau> s/pay/play/
Apr 15 19:03:52 <ptast> hi all, did somebody here attend the Miami Satellite Direct Readout Conference? Here are the presentations ftp://ftp.osd.noaa.gov/SDRC_2011/Presentations/Final%20Presentations/
Apr 15 19:03:58 <trondeau> Or maybe that was a Freudian slip...
Apr 15 19:04:07 <bpadalino> when you say one thing but meant your mother?
Apr 15 19:04:21 <trondeau> I almost wrote that but resisted. Thanks, Cliff.
Apr 15 19:04:46 <ptast> lots of GR stuff to try..
Apr 15 19:05:56 <trondeau> Thanks for the presentations, ptast, will have to look through them.
Apr 15 19:07:03 <trondeau> Ok. I think that's it. Thanks everyone! Hopefully the SIP issues will be worked out next time.
Apr 15 19:07:29 <trondeau> Remember, we're changing to a third-Thursday schedule (which rolls off the tongue nicer, anyways).
Apr 15 19:07:44 <jcorgan> time for beer
Apr 15 19:07:57 <trondeau> Indeed. Cheers!
Apr 15 19:08:17 <voip-gr> Tom (SIP/xxx.xxx.xxx.xxx-000000c1) has left the conference.

Developer's Call, March 18, 2011[edit]

Annotated agenda[edit]

Time: 2200 UTC, 6 PM EST, 3 PM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

  • Discussion of the 'next' -> 'master' transition
  • Release schedule
  • Debian and Fedora releases

== IRC Log
No IRC log for this call. IRC room was quite and no notes were taken. ==

The conversation consisted of a brief description of what has been happening with the transition of 'next' into the 'master' branch.
- There is no more 'next' branch for now
- The 'master' branch is the old 'next' branch and now represents our main development
- This 'master' branch is for the 3.4 series
- Looking for people to test, test, test the master so we can cut 3.4 as quickly as possible
- Need more info on bugs in OSX, Windows, E100
- Volk is currently broken in E100 and --disable-volk still leaves some remnants; needs to be fixed

Talked about a more frequent release schedule. We want to get to a more regular series of releases to get new features out faster. Part of this is to develop a stronger roadmap of features to work on and introduce.

We also had a long discussion about what to do about providing Debian/Ubuntu and Fedora packages. We would like to improve the speed of getting the releases to the distributions, but we are debating on how and the best way to package these. We came to no good conclusions and will revisit this later.

Developer's Call, March 7, 2011[edit]

Annotated agenda[edit]

Time: 1700 UTC, 12 PM EST, 9 AM PST, etc.
SIP call: sip:gnuradio@digitalbazaar.com
IRC channel: #gnuradio on Freenode

  • Details of the GNU Radio 'next' transition
    • What bugs are left to work out?
      • keep_1_in_n
      • Volk build issues
  • gr-digital revolution
    • introduction to the "grand-restructure"
    • some inside-baseball on the migration
  • gr-conference
    • kicking around a few more details and ideas

IRC Log[edit]

<trondeau> Agenda: 1. discuss 'next' -> 'master' move
<trondeau> Agenda: 2. gr-digital concept and the future rebuild
<trondeau> Agenda: 3. Conference [opt.]
<voip-gr> Conference connections: SIP/xxx.xxx.xxx.xxx:1026-000000a6, SIP/vps.corganenterprises.com-000000a2, Tom (trondeau) [SIP/xxx.xxx.xxx.xxx-000000a3], Andrew Ge [SIP/xxx.xxx.xxx.xxx:5060-000000a5], Josh Blum [SIP/xxx.xxx.xxx.xxx:5060-000000a4]
<trondeau> trapicki, issues getting on the SIP bridge?
<voip-gr> Martin Braun (SIP/xxx.xxx.xxx.xxx:13015-000000a7) has joined the conference.
<unknown> (SIP/xxx.xxx.xxx.xxx:1026-000000a8) has joined the conference.
<trapicki> got it
<voip-gr> Conference connections: Martin Braun [SIP/xxx.xxx.xxx.xxx:13015-000000a7], SIP/xxx.xxx.xxx.xxx:1026-000000a6, SIP/vps.corganenterprises.com-000000a2, SIP/xxx.xxx.xxx.xxx:1026-000000a8, Tom (trondeau) [SIP/xxx.xxx.xxx.xxx-000000a3], Andrew Ge [SIP/xxx.xxx.xxx.xxx:5060-000000a5], Josh Blum [SIP/xxx.xxx.xxx.xxx:5060-000000a4]
<trapicki> I'll have to check the headset* benreynwar (~ben@ip24-255-25-229.tc.ph.cox.net) has joined #gnuradio
<trapicki> listing, check my mic...
<voip-gr> Ben Reynwar (SIP/xxx.xxx.xxx.xxx:5060-000000a9) has joined the conference.
<voip-gr> Martin Braun (SIP/xxx.xxx.xxx.xxx:13015-000000a7) has left the conference.
<trondeau> for those on the call, can you at least hear us?
<jcorgan> we can hear jcorgan, jblum, ben, and trondeau
<trapicki> trondeau: hear you all
<trapicki> I can't see my voip id, can someone tell me
<trondeau> trapicki, looks like you are xxx.xxx.xxx.xxx:1026-000000a6
<trondeau> Discussing build issues for Volk on other systems (e100, windows, osx)
<voip-gr> trapicki is now associated with SIP/xxx.xxx.xxx.xxx:1026-000000a6.
<guruofquality> ./gnuradio-core/src/lib/general/malloc16.c
<trondeau> in windows: no posix_memalign in Windows.
<trondeau> Use malloc16Align/free16Align ?
<trondeau> Only in QA code; make small fix here. Maybe Nick F. can fix it while he's working on other code.
<trondeau> e100: compile time checks doesn't work.
<trondeau> Nick F. is working on building a new compile structure for Volk to help here.
<trondeau> The other issue with volk is change in documentation of build
<trondeau> Make next->master, then start working on updating docs, cleanup, and build instructions.
<guruofquality> complex.h still an issue?
<trondeau> Note to self: update dependency list in next:README. Put dependency list on Wiki for both master/maint and next.
<trondeau> I believe that Don said that a new update to Cygwin fixed the complex.h issue. Must look back at email thread to confirm.
<trondeau> Discussing keep_1_in_n block issue with very large n.
<trondeau> Can we rewrite keep_1_in_n to use set_relative_rate but not require so much memory?
<trondeau> Going to look into how to fix this. Needs to get fixed for WX FFT block to work properly.
<trondeau> Talking gr-digital: the start of the GNU Radio rebuild structure.
<trondeau> Splitting out and reorganizing the scattered items into known components.
<trondeau> gr-digital is going to be the test for doing this.
<voip-gr> Martin Braun (SIP/xxx.xxx.xxx.xxx:37833-000000aa) has joined the conference.
<trondeau> Put all digital modulation/demodulation (including OFDM) into it's own top-level block.
<trondeau> The concept of a top-level block is that it will be a stand-alone directory with it's own autotools project structure (bootstrap, configure, etc.). The GR root will then call each top-level block in turn (much like Volk and usrp2/firmware today).
<trondeau> This will be started after the next->master transition.
<trondeau> Aiming this as being a part of 3.5 release.
<trondeau> Corgan: we won't have a next branch until we release 3.4.0; then cut new 'next' from there and start on these kinds of build issues.
<trondeau> Ben: suggesting to break into more submodules (gr.digital into gr.digital.receiver, gr.digital.transmitter, gr.digital.ofdm?)
<trondeau> Discussion of documentation, esp. in Python documentation.
<trondeau> Top-level blocks should help this. We can also take this as an opportunity to do the Python documentation the "right" way.
<trondeau> We can work with gr-digital on master (since it won't change the old way of doing things).

Developer's Call, February 18, 2011[edit]

Annotated agenda[edit]

  • GNU Radio release 3.4

' what are the new features

'* gr-uhd

' almost ready to freeze gr-uhd so we can have 'master' that won't have API changes

'* volk

' we always want to have a generic kernel

' add an Orc kernel for those functions to give a first, easy access to SIMD (even if its not 100%)

' Balister has issues with cross compiling (to E100); need to address this before we start integrating volk

'* stream tags

' look into bug in keep_one_in_n

' need more examples and documentation

'* message passing


' accessor methods are not thread-safe; Should convert to message passing for this (which is thread safe).

'* handling the transition from git master 3.4 to 3.5

' investiage RPM and DEB

' deprecate libusrp and libusrp2

'* force new development on next with UHD

'* start migrating examples over to UHD

' Think about developing all examples with GRC

'* use this to force us to think about features in GRC to make this.

  • GNU Radio conference

'* work on getting specifics nailed down (time, date, format, etc.)

'* mid September, probably after PIMRC, Wed-Fri with maybe an open day on Saturday for hacking.

  • OpenBTS

'* skip this for time

IRC Log[edit]

Feb 18 17:03:45 <jcorgan> voip: connections?
Feb 18 17:03:45 <voip-gr> Conference connections: SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:5060-0000005d, SIP/XXX.XXX.XXX.XXX:1315-00000059, Philip Balister [SIP/XXX.XXX.XXX.XXX:5060-00000051]_
Feb 18 17:04:18 <voip-gr> Johnathan Corgan (SIP/corganenterprises.com-0000005e) has joined the conference.
Feb 18 17:22:22 <voip-gr> Tim (SIP/sip2sip.info-0000005f) has joined the conference.
Feb 18 17:23:41 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-0000005d) has left the conference.
Feb 18 17:23:44 <voip-gr> Tim (SIP/sip2sip.info-0000005f) has left the conference.
Feb 18 17:24:47 <voip-gr> Conference connections: SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, Johnathan Corgan [SIP/corganenterprises.com-0000005e], Philip Balister [SIP/XXX.XXX.XXX.XXX:5060-00000051]
Feb 18 17:25:04 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000060) has joined the conference.
Feb 18 17:29:45 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000060) has left the conference.
Feb 18 17:38:04 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000061) has joined the conference.
Feb 18 17:38:30 <voip-gr> Tom (SIP/ has joined the conference.
Feb 18 17:41:35 <trondeau> voip-gr: connections?
Feb 18 17:41:35 <voip-gr> Conference connections: SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, Tom [SIP/], Johnathan Corgan [SIP/corganenterprises.com-0000005e], Philip Balister [SIP/XXX.XXX.XXX.XXX:5060-00000051], SIP/XXX.XXX.XXX.XXX:5060-00000061
Feb 18 17:45:06 <voip-gr> Josh Blum (SIP/XXX.XXX.XXX.XXX:3813-00000063) has joined the conference.
Feb 18 17:48:44 <voip-gr> Manu Sporny (SIP/6000-00000064) has joined the conference.
Feb 18 17:49:18 <jcorgan> voip: connections?
Feb 18 17:49:19 <voip-gr> Conference connections: Josh Blum [SIP/XXX.XXX.XXX.XXX:3813-00000063], Manu Sporny [SIP/6000-00000064], Tom [SIP/], Johnathan Corgan [SIP/corganenterprises.com-0000005e], SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, Philip Balister [SIP/XXX.XXX.XXX.XXX:5060-00000051], SIP/XXX.XXX.XXX.XXX:5060-00000061
Feb 18 17:51:30 <Crofton> hmm, I should n't be there
Feb 18 17:51:50 <trondeau> voip-gr: disconnect 51
Feb 18 17:51:51 <voip-gr> Philip Balister (SIP/XXX.XXX.XXX.XXX:5060-00000051) has left the conference.
Feb 18 17:51:54 <trondeau> there
Feb 18 17:52:21 manu1 I think you killed Balister, tom :P*
Feb 18 17:52:42 <trondeau> he just said he shouldn't be there. So I made sure he wasn't :)
Feb 18 17:52:50 <manu1> oh, my bad :)
Feb 18 17:52:57 <trondeau> I hope that's what he meant...
Feb 18 17:53:34 <voip-gr> Conference connections: Josh Blum [SIP/XXX.XXX.XXX.XXX:3813-00000063], Manu Sporny [SIP/6000-00000064], Tom [SIP/], Johnathan Corgan [SIP/corganenterprises.com-0000005e], SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, SIP/XXX.XXX.XXX.XXX:5060-00000061
Feb 18 17:57:17 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000065) has joined the conference.
Feb 18 17:57:29 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000065) has left the conference.
Feb 18 17:57:40 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000066) has joined the conference.
Feb 18 17:57:52 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000066) has left the conference.
Feb 18 17:58:23 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000067) has joined the conference.
Feb 18 17:58:35 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000067) has left the conference.
Feb 18 17:59:25 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000068) has joined the conference.
Feb 18 17:59:38 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000068) has left the conference.
Feb 18 18:00:12 <trondeau> voip-gr: connections?
Feb 18 18:00:13 <voip-gr> Conference connections: Josh Blum [SIP/XXX.XXX.XXX.XXX:3813-00000063], Manu Sporny [SIP/6000-00000064], Tom [SIP/], Johnathan Corgan [SIP/corganenterprises.com-0000005e], SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, SIP/XXX.XXX.XXX.XXX:5060-00000061
Feb 18 18:01:09 * spino327 (~spino327@ has joined #gnuradio
Feb 18 18:02:08 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000069) has joined the conference.
Feb 18 18:02:20 <voip-gr> Alexandru Csete (SIP/ekiga.net-00000069) has left the conference.
Feb 18 18:02:26 * AlexCsete (~AlexCsete@1508894357.dhcp.dbnet.dk) has joined #gnuradio
Feb 18 18:02:42 <trondeau> AlexCsete, having issues with audio when you connect?
Feb 18 18:02:47 gnuradio (824c6092@gateway/web/freenode/ip. has joined #gnuradio*
Feb 18 18:03:03 <AlexCsete> I get "Call/Transaction Does not exist"
Feb 18 18:03:20 <trondeau> AlexCsete, strange. We're seeing you connect and disconnect.
Feb 18 18:03:40 <AlexCsete> ok, probably my end, never tried it before
Feb 18 18:03:40 <jcorgan> might try twinkle instead of ekiga
Feb 18 18:04:14 <voip-gr> Tim (SIP/sip2sip.info-0000006a) has joined the conference.
Feb 18 18:04:33 <voip-gr> Philip Balister (SIP/XXX.XXX.XXX.XXX:5060-0000006b) has joined the conference.
Feb 18 18:04:48 <voip-gr> Thomas Tsou (SIP/vt.edu-0000006c) has joined the conference.
Feb 18 18:04:56 <manu1> voip: connections?
Feb 18 18:04:57 <voip-gr> Conference connections: Josh Blum [SIP/XXX.XXX.XXX.XXX:3813-00000063], Manu Sporny [SIP/6000-00000064], Tom [SIP/], Johnathan Corgan [SIP/corganenterprises.com-0000005e], Tim [SIP/sip2sip.info-0000006a], Thomas Tsou [SIP/vt.edu-0000006c], SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, Philip Balister [SIP/XXX.XXX.XXX.XXX:5060-0000006b], SIP/XXX.XXX.XXX.XXX:5060-00000061
Feb 18 18:05:14 <voip-gr> Philip Balister (SIP/XXX.XXX.XXX.XXX:5060-0000006b) has left the conference.
Feb 18 18:05:24 <voip-gr> Philip Balister (SIP/XXX.XXX.XXX.XXX:5060-0000006d) has joined the conference.
Feb 18 18:05:36 <voip-gr> Tim (SIP/sip2sip.info-0000006a) has left the conference.
Feb 18 18:05:56 <manu1> link to the agenda?
Feb 18 18:06:05 * yootis (~matt@108-65-78-2.lightspeed.sntcca.sbcglobal.net) has joined #gnuradio
Feb 18 18:06:07 <manu1> that's fine
Feb 18 18:06:09 <Crofton> pm your emial
Feb 18 18:07:20 <Crofton> http://pastebin.com/NDBu77iq
Feb 18 18:09:08 <guruofquality> darth vader?
Feb 18 18:10:12 * gnuradio has quit (Quit: Page closed)
Feb 18 18:10:25 <ttsou> much better
Feb 18 18:10:32 <Crofton> voip:connections
Feb 18 18:10:39 <Crofton> voip: connections?
Feb 18 18:10:39 <voip-gr> Conference connections: Josh Blum [SIP/XXX.XXX.XXX.XXX:3813-00000063], Manu Sporny [SIP/6000-00000064], Tom [SIP/], Johnathan Corgan [SIP/corganenterprises.com-0000005e], Thomas Tsou [SIP/vt.edu-0000006c], SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, Philip Balister [SIP/XXX.XXX.XXX.XXX:5060-0000006d], SIP/XXX.XXX.XXX.XXX:5060-00000061
Feb 18 18:10:47 <Crofton> how do I mute myself
Feb 18 18:10:57 <Crofton> can't figure it out here
Feb 18 18:11:01 <ttsou> just mute your mic?
Feb 18 18:11:07 <trondeau> voip-gr: mute 6d
Feb 18 18:11:08 <voip-gr> SIP/XXX.XXX.XXX.XXX:5060-0000006d has been muted.
Feb 18 18:11:08 manu1 try "voip: mute <last 2 digits of your SIP number>"*
Feb 18 18:11:30 <trondeau> you should also be able to tell it to "voip-gr: mute me" (i think).
Feb 18 18:11:41 <Crofton> voip-gr: mute 6d
Feb 18 18:11:47 <Crofton> voip-gr: mute me
Feb 18 18:11:47 <voip-gr> Crofton is not associated with an audio channel. You may specify the first or last name of a person's Caller ID. You may also specify the last 2 or more digits of the audio channel name. You can also map an IRC nickname to an audio channel. Use 'voip-gr: <CHANNEL> is me' to map yourself to an audio channel. Use 'voip-gr: <CHANNEL> is <NICK>' to map someone else to an audio channel.
Feb 18 18:11:57 manu1 only if your IRC name is associated w/ your name.*
Feb 18 18:12:02 <trondeau> ah.
Feb 18 18:12:08 <manu1> voip: 6d is Crofton
Feb 18 18:12:08 <voip-gr> Crofton is now associated with SIP/XXX.XXX.XXX.XXX:5060-0000006d.
Feb 18 18:12:19 <manu1> voip: unmute Crofton
Feb 18 18:12:20 <voip-gr> Crofton has been unmuted.
Feb 18 18:12:23 <manu1> voip: mute Crofton
Feb 18 18:12:24 <voip-gr> Crofton has been muted.
Feb 18 18:14:50 <Crofton> thanks manu1
Feb 18 18:15:17 manu1 np - you can run all of those commands yourself in the future.*
Feb 18 18:15:19 <Crofton> I am sure glad I live in VA where we have real intenet
Feb 18 18:16:26 manu1 when you join, just remember to look for your SIP connection announcement and then do - voip: <LAST_2_DIGITS_OF_MY_SIP_CONNECTION> is me*
Feb 18 18:16:37 manu1 then you can mute/unmute yourself using your IRC nickname.*
Feb 18 18:17:21 <voip-gr> Doug Geiger (SIP/ekiga.net-00000074) has joined the conference.
Feb 18 18:18:30 <ttsou> openbts like api stabilization
Feb 18 18:18:38 * douggeiger (~douggeige@pool-173-73-83-32.washdc.fios.verizon.net) has joined #gnuradio
Feb 18 18:18:48 manu1 has to go - but everything looks good re: the telecon - load average on the telecon bridge is 2%, audio buffers seem to be good for everyone, mixing seems to be happening fine.*
Feb 18 18:18:54 <manu1> good luck folks - you guys are doing some awesome stuff[[File:*

  • Feb 18 18:18:59 <Crofton> thanks*
  • Feb 18 18:18:59 <trondeau> thanks, manu1|]]

Feb 18 18:19:12 <voip-gr> Manu Sporny (SIP/6000-00000064) has left the conference.
Feb 18 18:21:30 <voip-gr> Conference connections: Josh Blum [SIP/XXX.XXX.XXX.XXX:3813-00000063], Tom [SIP/], Johnathan Corgan [SIP/corganenterprises.com-0000005e], Doug Geiger [SIP/ekiga.net-00000074], Thomas Tsou [SIP/vt.edu-0000006c], SIP/XXX.XXX.XXX.XXX:5060-00000053, SIP/XXX.XXX.XXX.XXX:5060-00000041, SIP/XXX.XXX.XXX.XXX:1315-00000059, Philip Balister (Crofton) [SIP/XXX.XXX.XXX.XXX:5060-0000006d], SIP/XXX.XXX.XXX.XXX:5060-00000061
Feb 18 18:22:28 * _ProGammer has quit (Quit: Leaving)
Feb 18 18:22:33 * b7500af1 (~vt@2001:468:c80:4240:21c:bfff:fe8b:9c32) has joined #gnuradio
Feb 18 18:23:03 <voip-gr> b7500af1 (SIP/XXX.XXX.XXX.XXX:5060-00000075) has joined the conference.
Feb 18 18:23:34 <Crofton> yeah
Feb 18 18:23:42 <Crofton> I want to bitch a little
Feb 18 18:23:51 <Crofton> voip-gr: unmute me
Feb 18 18:23:52 <voip-gr> Crofton has been unmuted.
Feb 18 18:25:26 <Crofton> voip-gr: mute me
Feb 18 18:25:26 <voip-gr> Crofton has been muted.
Feb 18 18:26:07 * merdin (~merdin@108-65-78-2.lightspeed.sntcca.sbcglobal.net) has joined #gnuradio
Feb 18 18:26:45 <Crofton> voip-gr: unmute me
Feb 18 18:26:46 <voip-gr> Crofton has been unmuted.
Feb 18 18:29:03 <guruofquality> you mean cross compling volk not orc
Feb 18 18:29:33 <spino327> Hey guys, Are you recording the conference? I had problems
Feb 18 18:30:14 <Crofton> not sure
Feb 18 18:30:27 <jcorgan> don't think so
Feb 18 18:30:56 <Crofton> voip-gr: mute me
Feb 18 18:30:56 <voip-gr> Crofton has been muted.
Feb 18 18:31:33 <guruofquality> volk = good
Feb 18 18:31:48 * ttsou has quit (.net.split)
Feb 18 18:31:49 * larsivi has quit (.net.split)
Feb 18 18:31:49 * borism has quit (.net.split)
Feb 18 18:31:50 <Crofton> volk is annoying cross :)
Feb 18 18:31:53 <trondeau> volk issues: Balister has issues with cross compiling
Feb 18 18:32:00 <trondeau> need to address this before we start integrating volk into GNU Radio
Feb 18 18:32:01 <jcorgan> balister has issues
Feb 18 18:32:08 * larsivi (~quassel@ has joined #gnuradio
Feb 18 18:33:03 <douggeiger> voip-gr: mute me
Feb 18 18:33:04 <voip-gr> douggeiger is not associated with an audio channel. You may specify the first or last name of a person's Caller ID. You may also specify the last 2 or more digits of the audio channel name. You can also map an IRC nickname to an audio channel. Use 'voip-gr: <CHANNEL> is me' to map yourself to an audio channel. Use 'voip-gr: <CHANNEL> is <NICK>' to map someone else to an audio channel.
Feb 18 18:33:11 <guruofquality> the keep_one_in_n_bug
Feb 18 18:33:18 <voip-gr> Doug Geiger (SIP/ekiga.net-00000074) has left the conference.
Feb 18 18:33:18 * ttsou (~ttsou@2001:468:c80:4240:21e:64ff:fe5f:bc08) has joined #gnuradio
Feb 18 18:33:39 <voip-gr> Doug Geiger (SIP/ekiga.net-00000076) has joined the conference.
Feb 18 18:33:47 <voip-gr> Doug Geiger (SIP/ekiga.net-00000076) has left the conference.
Feb 18 18:33:57 <voip-gr> douggeiger is not associated with an audio channel. You may specify the first or last name of a person's Caller ID. You may also specify the last 2 or more digits of the audio channel name. You can also map an IRC nickname to an audio channel. Use 'voip-gr: <CHANNEL> is me' to map yourself to an audio channel. Use 'voip-gr: <CHANNEL> is <NICK>' to map someone else to an audio channel.
Feb 18 18:34:42 <voip-gr> Doug Geiger (SIP/ekiga.net-00000077) has joined the conference.
Feb 18 18:35:09 <voip-gr> douggeiger is not associated with an audio channel. You may specify the first or last name of a person's Caller ID. You may also specify the last 2 or more digits of the audio channel name. You can also map an IRC nickname to an audio channel. Use 'voip-gr: <CHANNEL> is me' to map yourself to an audio channel. Use 'voip-gr: <CHANNEL> is <NICK>' to map someone else to an audio channel.
Feb 18 18:35:11 <trondeau> discussing stream tags
Feb 18 18:35:25 <voip-gr> Failed to associate douggeiger with an audio channel ending in 'Geiger'
Feb 18 18:35:28 <trondeau> voip-gr: 77 is douggeiger
Feb 18 18:35:29 <voip-gr> douggeiger is now associated with SIP/ekiga.net-00000077.
Feb 18 18:35:49 <trondeau> douggeiger, try muting yourself now.
Feb 18 18:35:49 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000061) has left the conference.
Feb 18 18:35:54 <voip-gr> douggeiger has been muted.
Feb 18 18:38:05 * Exadios (~Exadios@XXX.XXX.XXX.XXX) has joined #gnuradio
Feb 18 18:38:44 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000078) has joined the conference.
Feb 18 18:39:13 <douggeiger> pmt == polymorphic type
Feb 18 18:39:24 <douggeiger> Message passing is based on pmt's
Feb 18 18:40:00 <douggeiger> How to SWIG'ize, and/or get into Python
Feb 18 18:40:30 * b7500af1 has quit (Quit: Leaving)
Feb 18 18:41:41 <voip-gr> douggeiger has been unmuted.
Feb 18 18:41:43 <trondeau> we need to get more examples and better documentation about both the stream tags and message passing interfaces
Feb 18 18:41:44 <Crofton> some form of message passing is really critical, with or without python
Feb 18 18:42:36 <trondeau> accessor methods are not thread-safe. Should convert to message passing for this (which is thread safe).
Feb 18 18:42:58 <Crofton> we desperately need examples of this
Feb 18 18:43:04 <voip-gr> douggeiger has been muted.
Feb 18 18:43:37 <trondeau> next -> master discussion
Feb 18 18:43:38 <Crofton> we need to use gobby for minutes
Feb 18 18:43:55 <trondeau> gobby? If it helps keeping the minutes, I'll check it out.
Feb 18 18:44:36 <trondeau> http://gobby.0x539.de/trac/
Feb 18 18:44:40 <Crofton> I am muted correct?
Feb 18 18:44:41 <trondeau> I'll look into it.
Feb 18 18:44:52 <Crofton> yum install gobby :)
Feb 18 18:45:12 <trondeau> yum install gobby
Feb 18 18:45:12 <trondeau> The program 'yum' is currently not installed. You can install it by typing:
Feb 18 18:45:12 <trondeau> sudo apt-get install yum
Feb 18 18:45:16 <trondeau> Ubuntu fail! :)
Feb 18 18:46:41 <Crofton> we should try and make all the pkgs coherenet in how they are split
Feb 18 18:46:47 <Crofton> deb, rpm, opkg
Feb 18 18:46:48 <trondeau> We need to investiage making deb and rpm packages for GNU Radio
Feb 18 18:46:56 <trondeau> Crofton, yes
Feb 18 18:47:09 <Crofton> libusrp and libusrp2 go away with new next
Feb 18 18:47:19 <guruofquality> cmake
Feb 18 18:47:21 <Crofton> this will destroy almost all the examples
Feb 18 18:47:25 <guruofquality> hmm don't know about opkg
Feb 18 18:47:44 <Crofton> voip-gr: unmute me
Feb 18 18:47:44 <voip-gr> Crofton has been unmuted.
Feb 18 18:47:54 <douggeiger> Need to port examples to gr-uhd
Feb 18 18:48:09 <ttsou> would be lots of interest here (Va Tech) on benchmark examples
Feb 18 18:48:17 <guruofquality> soon I hope: a gr-uhd fft and gr-uhd sigggen
Feb 18 18:48:34 <guruofquality> I would do it w/ qtgui
Feb 18 18:49:38 <trondeau> guruofquality, that sounds like a good idea. Make some real sig_analyzer and sig_generator in Qt with UHD would be great.
Feb 18 18:49:39 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000078) has left the conference.
Feb 18 18:49:51 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000079) has joined the conference.
Feb 18 18:49:55 <guruofquality> yea, just need this time thing, where can I find it
Feb 18 18:50:24 <douggeiger> GRC files for all examples? Need more features in GRC first.
Feb 18 18:50:38 <voip-gr> douggeiger (SIP/ekiga.net-00000077) has left the conference.*
Feb 18 18:50:53 <voip-gr> Doug Geiger (SIP/ekiga.net-0000007a) has joined the conference.
Feb 18 18:51:03 <voip-gr> douggeiger is now associated with SIP/ekiga.net-0000007a.
Feb 18 18:51:11 <voip-gr> douggeiger has been muted.
Feb 18 18:51:26 <guruofquality> we need python based examples too
Feb 18 18:52:36 * spino327 has quit (Ping timeout: 255 seconds)
Feb 18 18:53:13 <douggeiger> The difficulty of porting examples to GRC can drive GRC feature development
Feb 18 18:55:01 <Crofton> don't let grc hold up really cool examples
Feb 18 18:55:26 <Crofton> party!
Feb 18 18:56:01 <Crofton> not the week before labour day
Feb 18 18:56:28 <douggeiger> GNURadio conference: need to lock in dates - UPenn fills up quickly
Feb 18 18:57:12 <douggeiger> PIMRC in Toronto Sept 11-14
Feb 18 18:57:25 <douggeiger> Can attract many SDR people who will be in country
Feb 18 18:57:30 <voip-gr> douggeiger has been unmuted.
Feb 18 18:57:32 <Crofton> that sounds fine to me
Feb 18 18:57:42 <voip-gr> douggeiger has been muted.
Feb 18 18:58:29 <douggeiger> Possible dates: Weds afternoon through Friday (Sept. 14-16)
Feb 18 18:58:39 <douggeiger> Hacking day on Sat. (17)
Feb 18 18:58:44 <voip-gr> douggeiger has been unmuted.
Feb 18 18:58:45 <Crofton> good idea
Feb 18 18:58:53 <voip-gr> douggeiger has been muted.
Feb 18 19:00:42 <douggeiger> Format and scheduling needs to be discussed further
Feb 18 19:00:52 <Crofton> Friday at 6 sucks
Feb 18 19:00:58 <ttsou> agreed
Feb 18 19:01:06 <douggeiger> Next developer call: time/date?
Feb 18 19:02:07 <Crofton> voip-gr: unmute me
Feb 18 19:02:10 <douggeiger> Want a regular, set time every (other?) month, and a more flexible time in between
Feb 18 19:02:57 <douggeiger> Spread the pain to allow east coast, west coast and european participation
Feb 18 19:03:27 <douggeiger> Fridays != ideal
Feb 18 19:04:22 <Crofton> voip-gr: mute me
Feb 18 19:04:23 <voip-gr> Crofton has been muted.
Feb 18 19:04:28 <Crofton> I'll shiut up now
Feb 18 19:04:36 <douggeiger> Other days of the week at the same time (6pm) is fine
Feb 18 19:04:47 <voip-gr> douggeiger has been unmuted.
Feb 18 19:05:38 <Crofton> jcorgan, is doing sip over 3g I thought
Feb 18 19:05:45 <voip-gr> douggeiger has been muted.
Feb 18 19:06:03 <Crofton> thanks guys
Feb 18 19:06:11 <Crofton> heh
Feb 18 19:06:13 <douggeiger> Thanks for putting this together
Feb 18 19:06:19 <Crofton> we do OE meetings via gobby :)
Feb 18 19:06:26 <Crofton> gn
Feb 18 19:06:27 <voip-gr> douggeiger has been unmuted.
Feb 18 19:06:29 <Crofton> woo hoo
Feb 18 19:06:30 <voip-gr> Johnathan Corgan (SIP/corganenterprises.com-0000005e) has left the conference.
Feb 18 19:06:31 <Crofton> later
Feb 18 19:06:32 <voip-gr> douggeiger has been muted.
Feb 18 19:06:35 <voip-gr> Thomas Tsou (SIP/vt.edu-0000006c) has left the conference.
Feb 18 19:06:36 <voip-gr> Josh Blum (SIP/XXX.XXX.XXX.XXX:3813-00000063) has left the conference.
Feb 18 19:06:40 <voip-gr> douggeiger (SIP/ekiga.net-0000007a) has left the conference.
Feb 18 19:06:47 <voip-gr> Crofton (SIP/XXX.XXX.XXX.XXX:5060-0000006d) has left the conference.
Feb 18 19:07:18 <voip-gr> Tom (SIP/ has left the conference.
Feb 18 19:07:26 * ttsou (~ttsou@2001:468:c80:4240:21e:64ff:fe5f:bc08) has left #gnuradio ("Leaving")
Feb 18 19:07:46 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-00000079) has left the conference.
Feb 18 19:08:00 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-0000007c) has joined the conference.
Feb 18 19:08:00 <voip-gr> <unknown> (SIP/XXX.XXX.XXX.XXX:5060-0000007c) has left the conference.