00:08 <rjo> mindrunner: all the patches on the underside are ground. imagine you are a voltage pulse and you are injected into one of the patches somewhere and and you want to make your way back to the power supply ground. now estimate where the path would be unnecessary long and where you would rather like to jump over a signal trace that is cutting the ground plane into patches. then scrub of copper on the two patches and place a short bridge connecting the ground patche

00:09 <rjo> something like this: http://i.imgur.com/KqQdJyv.jpg

00:10 <rjo> but this is not a precise science.

00:10 <mindrunner> nice!

00:11 <mindrunner> thanks! I will see what I can do. I also asked gadgetfactory about this and they told me, my board might be faulty. they gonna send a replacement now.

00:11 <mindrunner> and we ordered a pipistrello which might be here next week.

02:07 <sb0> mindrunner, it's a design problem, and all papilio boards I've seen are affected by the issue

02:16 <mindrunner> sb0, rjo , I am really excited if the replacement unit has the same issues. I know that the customer service will never admit that the whole series is faulty. But they gonna send me a new unit now. And hopefully the pipistrello is working better and I can continue with my project.

02:17 <mindrunner> thanks for all the help btw :)

08:39 <GitHub188> [artiq] sbourdeauducq pushed 3 new commits to master: http://git.io/vvDRG

08:39 <GitHub188> artiq/master c1ece33 Sebastien Bourdeauducq: manual: adapt to LED on RTIO

08:39 <GitHub188> artiq/master 91cd79a Sebastien Bourdeauducq: soc/runtime: add lwip (thanks Florent)

08:39 <GitHub188> artiq/master 4c63879 Sebastien Bourdeauducq: runtime: link against lwip, cleanups

10:17 <travis-ci> fallen/artiq#114 (pxi6733 - 0525ed0 : Yann Sionneau): The build has errored.

10:17 <travis-ci> Build details : http://travis-ci.org/fallen/artiq/builds/58886323

10:18 <ysionneau> rjo : after a few email exchange with Katie we figured out the fixes to apply, now it seems to work fine :) (pxi6733)

10:18 <ysionneau> so now let's polish the code and write the controler

10:18 <ysionneau> thanks for finding someone to test the code!

10:55 <acathla> Is there only one component video entry on the milkymist one?

10:59 <ysionneau> only one video channel

10:59 <ysionneau> http://comments.gmane.org/gmane.comp.multimedia.milkymist.devel/451

11:04 <ysionneau> http://m-labs.hk/mixxeo.html < mixxeo has 2 video inputs afaik but they are not analog inputs, only digital

11:08 <sb0> it has 4, but you run out of s6 PLLs to use all them plus the DVI out plus the DDR simultaneously

11:08 <sb0> simultaneously using 2 + switching works

11:10 <acathla> sb0: Il y a un évènement demain au jardin : Bring your own beamer. J'ai trouvé un projecteur vidéo hier (à la poubelle), et donc je me suis dit que j'allais participer vite-fait avec milkymist - projo, camescope et ce que je trouverais d'artistique et fun :)

11:11 <sb0> tu as une m1?

11:39 <GitHub149> [misoc] enjoy-digital pushed 2 new commits to master: http://git.io/vvyOi

11:39 <GitHub149> misoc/master 9666629 Florent Kermarrec: soc/cpuif: add with_access_functions parameter...

11:39 <GitHub149> misoc/master f32cdb1 Florent Kermarrec: soc: add shadow_address parameter...

11:43 <GitHub37> [misoc] enjoy-digital force-pushed master from f32cdb1 to 93de581: http://git.io/LjONPA

11:43 <GitHub37> misoc/master 93de581 Florent Kermarrec: soc: add shadow_address parameter...

12:06 <GitHub146> [misoc] enjoy-digital pushed 3 new commits to master: http://git.io/vvyCP

12:06 <GitHub146> misoc/master d22d58c Florent Kermarrec: add litepcie core

12:06 <GitHub146> misoc/master b4b37fb Florent Kermarrec: litepcie: add linux driver + utilities (sysfs + dma)

12:06 <GitHub146> misoc/master 8a822b9 Florent Kermarrec: litepcie: add litepcie_phy_wrappers to extcores

12:11 <GitHub120> [misoc] enjoy-digital pushed 1 new commit to master: http://git.io/vvyW6

12:11 <GitHub120> misoc/master 602eaf6 Florent Kermarrec: litepcie: fix asciiart in make.py

12:19 <_florent_> https://twitter.com/enjoy_digital/status/589039855356813312

12:19 <_florent_> :)

12:22 <ysionneau> you are talking to MiSoC from x86 host computer, through PCIe connection ?

12:22 <ysionneau> MiSoC being run on the PCIe card?

12:25 <_florent_> yes the x86 control the SoC as a LM32/MOR1Kx would do (the interrupts are also transmitted through PCIe)

12:25 <_florent_> controls

12:27 <ysionneau> ah, there is no softcore running on the MiSoC?

12:27 <ysionneau> everything is state machines and the PCIe endpoint controller?

12:29 <_florent_> yes the PCIe endpoint is only done by logic

12:30 <acathla> sb0, oui, celle du loop

12:31 <ysionneau> _florent_: very cool :)

12:31 <ysionneau> so what you show is what? "connecting" a mmapped() user space page on the host to the DDR on the PCIe board?

12:34 <_florent_> You can do 2 things in fact through PCIe: Control the wishbone bus and do DMA

12:35 <_florent_> wishbone is mmapped() in user space

12:35 <ysionneau> very good!

12:35 <ysionneau> nice work

12:36 <_florent_> for the DMA, the FPGA act as a master, the host only have to program where the FPGA shoudl write/read

12:37 <_florent_> thanks

12:39 <sb0> _florent_, careful not to fry your motherboard...

12:40 <whitequark> can you even do that with this setup?

12:42 <_florent_> sb0: why?

12:49 <sb0> _florent_, I suspect the kc705 is resposible for my broken motherboard. but I wasn't using the pcie when that happened...

12:50 <_florent_> ah OK, the KC705 is in this computer for more than 6 months now, it's still alive :)

14:03 <travis-ci> fallen/artiq#115 (pxi6733 - aa0956d : Yann Sionneau): The build has errored.

14:03 <travis-ci> Build details : http://travis-ci.org/fallen/artiq/builds/58912340

14:28 <travis-ci> fallen/artiq#116 (pxi6733 - 82e8ab6 : Yann Sionneau): The build has errored.

14:28 <travis-ci> Build details : http://travis-ci.org/fallen/artiq/builds/58915669

16:23 <travis-ci> fallen/artiq#117 (pxi6733 - fefeb9c : Yann Sionneau): The build has errored.

16:23 <travis-ci> Build details : http://travis-ci.org/fallen/artiq/builds/58933332

16:29 <travis-ci> fallen/artiq#119 (pxi6733 - 2bd24f3 : Yann Sionneau): The build has errored.

16:29 <travis-ci> Build details : http://travis-ci.org/fallen/artiq/builds/58934404

16:33 <GitHub162> [artiq] sbourdeauducq pushed 2 new commits to master: http://git.io/vv9G6

16:33 <GitHub162> artiq/master af6a8f6 Sebastien Bourdeauducq: gitignore: ignore static libs

16:33 <GitHub162> artiq/master 5805240 Yann Sionneau: manual: update runtime flashing instructions

16:41 <travis-ci> m-labs/artiq#125 (master - 5805240 : Yann Sionneau): The build has errored.

16:41 <travis-ci> Build details : http://travis-ci.org/m-labs/artiq/builds/58935554

16:49 <ysionneau> rha, pydaqmx raises an exception when you *import* it on a system where the NI DAQmx driver is not installed

16:49 <ysionneau> which is not helping, when you just want to generate the manual (/doc/manual) ...

16:50 <ysionneau> should I try: import pydaqmx and just print the exception message in the except ? or just silently ignore the exception?

16:54 <sb0> lazy import it?

16:54 <ysionneau> ah ok

16:54 <sb0> how often do you actually need stuff from it?

16:55 <ysionneau> just in the load_sample_values and close methods

16:55 <ysionneau> the code is rather small anyway

16:55 <sb0> if it accesses the driver, I fear it might mess with the card when you are compiling the manual :-)

16:55 <ysionneau> I can import it in __init__

16:55 <ysionneau> I think it just tries to load a bunch of .so/.dll when you import it that's all

16:55 <ysionneau> but that's enough to make the import fail

16:55 <ysionneau> doing lazy import should fix that :)

16:56 <ysionneau> thanks

17:19 <ysionneau> sb0: would something like that be acceptable? https://github.com/fallen/artiq/commit/2f291532b4e6017e18ea5bc71a4819c9db88c9ac

17:25 <travis-ci> fallen/artiq#121 (pxi6733 - 2f29153 : Yann Sionneau): The build has errored.

17:25 <travis-ci> Build details : http://travis-ci.org/fallen/artiq/builds/58940645

17:26 <ysionneau> travis is annoying, even if I cancel the build, it builds anyway

21:05 <whitequark> huh? it cancels the build for me

21:07 <ysionneau> it did for me as well, but today it does not

21:28 <rjo> ysionneau: yes. i know. i helped through a bunch of the problems.

21:29 <rjo> ysionneau: i would recommend spending a little time and pushing the py3k changes in pydaqmx upstream (after making it py2 and py3 compatible).

21:30 <rjo> but it looks good. pydaqmx is not trying too much to outsmart nidaqmx or the user. you should also like it because of linguistic treasures like "char_etoile" ;)

21:31 <rjo> and i would recommend reverting that import failure commit and just add pydaqmx to the mock_modules list in doc/manual/conf.py