Saturday, 2014-04-26

*** tpb has joined #timvideos00:00
mithroshenki: you about?00:11
*** FeltonChris has joined #timvideos00:40
*** FeltonChris has quit IRC00:48
cancimithro: m( this -V of fxload is arguably somewhat strange...00:53
mithrocanci: It makes sense when you are using fxload from a udev script where you have the devpath00:54
cancimithro: actually I had/have quite an old version, the tarball it's built from is fxload-2008_10_13 (though I am not sure if there are newer releases?)00:54
canci-V shows the build date00:54
mithrocanci: yeah - that is the "old" fxload00:54
canciand as it was compiled when I installed it, it was showing Mar 201400:54
canci(why would anybody show the build date but not the version on -V)00:55
canciwhere is fxload actually maintained? still at sourceforge?00:56
cancihm, but the latest ubuntu version is also "0.0.20081013-1ubuntu1" so it's probably from the same tarball?00:57
canciyea, md5sums match00:58
*** FeltonChris has joined #timvideos01:02
canci... reading helps; the suggested version is at - /me should really go to bed01:03
tpbTitle: pbatard/fxload · GitHub (at
canciThough this -V is really misleading, in combination with:
tpbTitle: Digilent Atlys Board: Loading Firmware · timvideos/HDMI2USB Wiki · GitHub (at
canciAs -V is showing the builddate it's mostly pointless to tell whether the version is the right one or not01:04
mithrocanci: the version at pbatard doesn't support the devpath and considered unmaintained01:04
canciYea, but that is the version suggested at "loading firmware", right?01:04
CarlFKcanci: I think there are 3 versions roaming about...01:07
CarlFKone under github/pbatard01:07
CarlFKone at (which is old, and what is packaged for debian/ubuntu)01:08
tpbTitle: libusb/examples/fxload.c at master · libusb/libusb · GitHub (at
CarlFKand that one is the most current, but not packaged01:08
canciYea, as I read it the version at github/pbatard was intended to end up in the libusb examples01:09
canciSo which one should I use? Probably the one from libusb examples?01:09
CarlFKas far as I can tell, we can use any of them, as long as we only have one device plugged in01:09
cancihm, so why does the "loading firmware" doc state "the version which comes with ubuntu is too old" ?01:10
CarlFKbecause mithro thinks we need to be ready for 2 devices at all times ;)01:11
canciyea, but the _old_ version actually supports that01:12
cancias it uses device path01:12
canciand not vendor-id:product-id01:12
canciwhile the new version doesn't?01:12
CarlFKbut it is hard to build and the author keeps yelling at me01:12
mithrocanci: because that command uses vid:pid01:12
canciyea, but the _old_ version supports device-path. The new version takes vid:pid, right?01:13
CarlFKmithro: if we only have one device, any reason not to use the one in the package repo?01:13
mithroCarlFK: using vid:pid makes it hard to make the process reliably automatic from udev01:14
CarlFKI am not using udev01:15
mithroCarlFK: I know, but if we want "just install the dpkg and plug in the Atlys board" set up then udev is the way to go01:16
mithroCarlFK: you just plug the board in, no programming or other scripts needing to be run01:17
CarlFKuntil things are stable, I doubt that is realistic01:18
canciwhat I don't understand, but probably I am getting something wrong, the ubuntu version/old version uses a device path01:19
canciso it could be used with udev01:19
mithrocanci: what ubuntu version are you on?01:19
canciwhile the new version uses vid:pid which is obviously not so unique01:19
mithrocanci: I was pretty sure the newer ubuntu's have the newer fxload packaged01:20
tpbTitle: Ubuntu – Details of package fxload in trusty (at
cancistill 0.0.20081013-1ubuntu101:20
mithroMaybe Carl has been confusing me01:21
CarlFKit is contagious!01:22
CarlFK  I am pretty sure that is what is packaged01:23
canciyea, it is01:23
tpbTitle: Repository - [linux-hotplug] Index of / Repository - linux-hotplug Index of / (at
canciand that versions usage is: fxload [-vV] [-l] [-t type] [-D devpath]01:23
mithroso that *is* the version which should work with udev01:24
CarlFK  home:
tpbTitle: Debian -- Details of package fxload in wheezy (at
canciyea, I also compared the tarballs from ubuntu/debian to those on the sourceforge page, they are identical01:25
cancithey add a patch so they will also read DEVNAME as environment variable instead of only reading DEVICE as the original version does01:26
cancibut other than that there are no more functional changes that I can see01:27
canciokay, I'll be off for now, there's a long day ahead tomorrow ;)01:39
CarlFKcanci: thanks for poking into the fxload thing.  it was driving me bonkers.01:39
mithrookay, I just managed to flash the Atlys with the latest Cypress+FPGA firmware using only open source tools01:47
CarlFKyay indeed01:47
mithroand the capture works fine01:50
mithroI'm going to upload new files to the prebuild repository for you to test CarlFK01:51
CarlFKmithro: uploaded?04:22
*** tvCommitBot has joined #timvideos04:35
tvCommitBot[HDMI2USB] mithro pushed 2 new commits to master:
tvCommitBotHDMI2USB/master 96710a0 Tim 'mithro' Ansell: Adding Makefile for Cypress and updating the .gitignore04:35
tvCommitBotHDMI2USB/master c827c8a Tim 'mithro' Ansell: Adding xsvf file production (needed for libFPGAlink programming).04:35
*** tvCommitBot has left #timvideos04:35
*** mparuszewski has joined #timvideos05:28
mithroCarlFK: there you go05:47
CarlFKk - you here for another 10 min?05:47
mithroJust writing a script to make it easier to publish new binaries05:47
mithroCarlFK: I'm still around now06:08
CarlFKI am fighting with reading data from a csv file.. how do I turn byte strings into unicode strings?06:09
CarlFK'\xe2'.decode()  UnicodeDecodeError: 'ascii' codec can't decode byte 0xe2 in position 0: ordinal not in range(128)06:09
CarlFK'\xe2' is nifty double back qutoes06:09
CarlFKI need to convert it to u'\u2018' because that is what comes back when I store it into postgresql, and I am looking for changes in the text file sense the last time I stored it in the db, but I am detecting the quotes being encoded? differently06:11
mithroyou need 'utf-8' in there06:36
mithroand \xe2 isn't valid utf-806:36
CarlFKmithro: yeah, just figured that out06:38
CarlFKmy print statements confused me06:38
mithrodecode means "take bytes and convert it to text/characters"06:38
mithroencode means "take text/characters and convert them to bytes"06:38
*** Niharika has joined #timvideos07:33
CarlFKmithro: gah, I used up my last bit of energy on importing csv and providing json ..  whacky backwards land today.07:35
*** tvCommitBot has joined #timvideos07:47
tvCommitBot[HDMI2USB] mithro tagged v0.0.0 at 879ed26:
*** tvCommitBot has left #timvideos07:47
*** tvCommitBot has joined #timvideos08:04
tvCommitBot[HDMI2USB] mithro pushed 1 new commit to master:
tvCommitBotHDMI2USB/master eaee2d1 Tim 'mithro' Ansell: Adding missing hdmi2usb.batch file needed for xsvf generation.08:04
*** tvCommitBot has left #timvideos08:04
*** Niharika has left #timvideos08:34
*** FeltonChris has left #timvideos10:26
*** Niharika has joined #timvideos11:34
*** mparuszewski has quit IRC11:54
*** CarlFK has quit IRC13:55
*** Niharika is now known as Niharika|afk14:11
*** CarlFK has joined #timvideos15:42
*** ChanServ sets mode: +v CarlFK15:42
*** Niharika|afk is now known as Niharika15:43
*** skay is now known as nottherealskay16:18
*** nottherealskay is now known as skay16:18
*** Niharika has left #timvideos17:08
*** puck has quit IRC17:26
*** puck has joined #timvideos17:29
tariq786anyone have an idea how to create progress blog for GSOC?17:48
aps-sidshey tariq786 , you can choose from various blogging platforms like Blogger or Wordpress.17:52
tariq786ok thanks aps-sids18:06
*** mparuszewski has joined #timvideos19:18
*** CarlFK has quit IRC21:56
*** mparuszewski has quit IRC22:26
*** CarlFK has joined #timvideos22:43
*** ChanServ sets mode: +v CarlFK22:43
CarlFKmithro: let me know if there is something I can test.  but nap time now, I'll bb in an hour or so.23:04
mithroYou can test the new .hex/.bit files23:49

Generated by 2.12.1 by Marius Gedminas - find it at!