#qi-hardware IRC log for Friday, 2012-04-06

wolfspraulgood morning everybody01:33
wolfspraulxiangfu: ah, you are already there :-)01:33
xiangfugood morning01:38
xiangfuI saw the message from Mirko. (IRC log)01:38
kristianpaulmorning !01:40
wolfspraulah yes, right01:43
wolfspraulkristianpaul: we are sending your package with goodies today, finally01:43
wolfspraulsorry about the delay, but details details details01:43
kristianpaulgood, atben here to come :)01:44
kristianpaulno problem about delays, anyway i'll be out home next week :)01:45
kristianpaulso i must wait some days to open it... :(01:45
wolfspraulyeah, well.01:46
wolfspraulit was a lot of details01:46
kristianpaulsure no problem, also travel was un-expect01:47
wolfspraulbut for example for the rtl2832u/e4000, you see on the list and forums everywhere now how many people have sourcing problems01:47
wolfspraulwhich is totally to be expected and I told you before01:47
kristianpauli guess next week will arrive the *proably* rtl-sdr dongle as well..01:47
kristianpaulsure sure i understan01:47
wolfspraulso the only way I can deliver something *predictable* is if I do some homework :-)01:47
wolfspraulit's actually all very simple, but amazing how many people like to live with illusions :-)01:48
kristianpaul( btw why nanonote dont ship full vim-full and  vim-help? )01:49
kristianpaulby default01:49
wolfspraulcould be as easy as not having space on the 512mb partition01:50
wolfspraulit's very full, whenever we add something something else has to go01:50
wolfsprauland nobody wants to touch the default partition size :-)01:50
kristianpaulah ok01:50
wolfspraulmaybe a safe way to start would be an option in reflash_ben.sh01:51
xiangfugood. our release build success. now testing...01:51
wolfspraullike --2gb-root01:51
kristianpaulyeah because tha vi from busybox is helish nighmare01:51
wolfspraulwpwrak: do you have any feedback for adam's m1 kicad work so far?01:51
wolfspraulI am super happy that it seems Adam is just leading and pulling forward, even though others like me and xiangfu are struggling to keep up though I very much wanted to participate01:52
wolfspraulI think it's good if adam just charges ahead01:52
wolfsprauleventually the others catch up01:52
wolfspraulbut if you see something wrong, please holler01:52
wolfspraulxiangfu: how about that idea?01:52
wolfspraullet's say we leave the partitioning as-is, but we add a new option to reflash_ben.sh01:53
wolfspraullike --root-2gb01:53
kristianpaulxiangfu: oh it include nanomap? i can reflash right now !01:53
wolfspraulof course I know the numbers are hardcoded here and there01:53
wolfspraulbut by leaving the default we can start to offer a 2gb install process without surprising anyone or causing data losses01:54
wolfsprauljust an idea...01:54
xiangfukristianpaul, yes. it have nanomap01:55
kristianpaulLINK please ! :-)01:56
kristianpaulor just run lastest?01:56
xiangfukristianpaul, http://fidelio.qi-hardware.com/~xiangfu/build-nanonote/openwrt-xburst.full_system-20120404-1633/01:56
xiangfukristianpaul, I am not test it yet. downloading..01:57
xiangfuwolfspraul, --2gb-root that needs different rootfs.ubi and different kernel. mean re-compile.01:57
kristianpaulwait was not 512Mb a plus for fast nanonote boot?01:58
xiangfukristianpaul, you can reflash it by : reflash_ben.sh -d openwrt-xburst.full_system-20120404-163301:58
kristianpauli actually boot pretty fast right, even more when you disable gmenu2x :)01:58
xiangfukristianpaul, after Lars's work on nanonote, 2GB rootfs is not that slow. 01:58
xiangfucompare to 512MB, it increase 1.5 second. if I don't remember wrong.01:59
wolfspraulxiangfu: well, my idea is to keep the --2gb thing a special case01:59
wolfspraulI know it's hardcoded01:59
wolfspraulbut maybe we can find a way to make it patchable from outside, no matter how ugly01:59
whitequarkyou can set up mtd partition sizes from kernel cmdline01:59
xiangfuwolfspraul, yes. needs re-compile.01:59
kristianpaulor remove uboot from once :)01:59
wolfspraulyeah we can also try to parameterize some of those values a bit more02:00
wolfspraulxiangfu: no I want to avoid that!02:00
wolfsprauland trust me, no 'need', just depends on what you do :-)02:00
wolfspraulbut anyway, maybe not now. we will surely break something again, which is worst.02:00
whitequarkyou can just dump the mtd partitioning and use UBI volumes02:00
kristianpaulwhitequark: i remenber boot from UBI was terrible slow02:08
kristianpaulor was something else02:10
kristianpaulpiwik stats are interesting 02:11
kristianpaulgood wikipedia comes second after google :)02:12
mthlarsc did some kernel tweaking to speed up mounting of UBI volumes02:13
mthone patch was dropped because it broke non-UBI NAND access, but the rest is still in jz-3.302:14
whitequarkkristianpaul: that was yaffs203:11
whitequarkUBI is very fast compared to yaffs/jffs03:12
Action: kristianpaul happy again with nanomap back03:13
kristianpaulThanks xiangfu  !03:13
qi-bot[commit] Adam Wang: added MIC2550AYTS usb transceiver into micrel.lib (master) http://qi-hw.com/p/kicad-libs/2a2562806:24
qi-bot[commit] Adam Wang: added symbol of AP21X2A dual channel power switch into diodes_inc.lib (master) http://qi-hw.com/p/kicad-libs/5bf49fc08:27
qi-bot[commit] Adam Wang: added symbol of ADV7181C video decoder into analog_devices.lib (master) http://qi-hw.com/p/kicad-libs/f1b67f008:27
kristianpaulzrafa: i guess you are suscribed to MiPrimerRobot mail list, righ?12:08
kristianpauli think there are hardware guys from argentina there12:08
mthlarsc: it seems posts to the ALSA mailinglist by non-members end up in /dev/null (moderation queue but no moderation happened in the last week)12:17
mthis it possible for you to submit a patch for c21d1d90ee7769968c79d5309a2ea4d96641e46a or would you prefer that I subscribe to the mailinglist12:17
mththat's the "replace comma operators by semicolons" patch, so I don't expect any discussion about it12:18
zrafakristianpaul: no. What does that project do? (miprimerrobot)14:40
viricI may be stupid...15:03
viricwhy I don't get any log level KERN_DEBUG output?15:04
viricmaybe they are trimmed out at build time by some setting?15:04
zrafaviric: maybe the messages are going to somewhere you dont check?15:21
viricI'm on x86...15:21
viricI'm hitting this: http://comments.gmane.org/gmane.linux.kernel/126898315:21
larscmth: did you Cc Mark Brown?16:49
mthno, I sent it only to the mailinglist16:50
mthshall I send it again with him on the Cc?16:50
larsche writes at least once a week on the alsa mailinglist "When sending patches always Cc maintainers!" ;)16:51
mthah, but I don't read the mailinglist, that's where the issue started from ;)17:06
mthI'll resend the patch17:07
larscmth: but you should always Cc maintainers anyway17:09
qi-bot[commit] Werner Almesberger: ircstat/: end of March 2012 update (and new script to plot mailing lists) (master) http://qi-hw.com/p/wernermisc/1b9eddf17:16
qi-bot[commit] kyak: gcc-mips: update to 4.6.2; fixed build (master) http://qi-hw.com/p/openwrt-packages/cb9c2cf18:41
whitequarkpabs3: you might be interested20:25
whitequarkI've been contracted to write a Flash ActionScript 3 decompiler20:35
whitequarkhere's what I have so far:20:35
rohwhitequark: my condolences20:35
whitequarkhttp://pastie.org/3741156 (original) -> http://pastie.org/3741164 (decompiled)20:35
whitequarkit can also output a token stream like this: http://pastie.org/374079520:36
whitequarkroh: hehe, I'm almost done20:36
whitequarkI only have control flow analysis left, and it's pretty simple20:36
whitequarkjust dominate the graph and the job is done20:36
whitequark(well, to talk seriously, you really need to find dominating blocks in the control flow graph)20:37
mthlarsc: you were right: Mark applied the patch within 10 minutes of me sending it :)23:59
--- Sat Apr 7 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!