#qi-hardware IRC log for Friday, 2012-03-30

whitequarkanyone here knowledgeable about metalworking/welding?00:01
whitequarkbecause I just tried to fix a metal thingy and failed miserably00:05
whitequarkalso found out that my butane torch has enough power to melt copper, but if it's at least a little bit less thick than the one I need00:06
qi-bot[commit] David Kühling: alpy: some more patching, make exmples python scripts work & package them (master) http://qi-hw.com/p/openwrt-packages/1353ef000:40
kristianpaulergh i could reproduce that VID00003.MP4 file from rejon02:45
kristianpaulcould not*02:45
kristianpauloh you here02:52
kristianpaulVID00002.MP4 seems ok, but the other well02:53
rejonok, I'll try a reupload of that file02:54
kristianpaulthanks !02:55
kristianpaulISO: Unknown File Type Major Brand: MSNV02:55
kristianpaulsaid mplayer02:55
qi-botThe build has FAILED: http://fidelio.qi-hardware.com/~xiangfu/building/Nanonote/Ben/openwrt-xburst.minimal-20120330-0202 06:25
Action: DocScrutinizer moos a little06:38
cladamw(M1r4 KiCad version) schematic symbols/libraries are under creations, see http://projects.qi-hardware.com/index.php/p/kicad-libs/   not finished yet.08:01
qi-bot[commit] Adam Wang: ignore *.bak & *.bck (master) http://qi-hw.com/p/kicad-libs/b3dafeb08:24
kyakbartbes: just a version update (2.0.0 -> 2.0.2)08:59
qi-bot[commit] kyak: nightsky: fix build (master) http://qi-hw.com/p/openwrt-packages/a3d14a909:00
qi-bot[commit] Xiangfu: nanonote-files add more nanonote config files (master) http://qi-hw.com/p/openwrt-packages/1efced809:07
kyakbartbes: and yeah, nlove build is broken -\09:07
qi-bot[commit] Xiangfu: nanonote optimize (master) http://qi-hw.com/p/openwrt-xburst/a808dfa09:12
kyakxiangfu: btw, we don't have to do that.. If we have BUILD_NLS, uClibc is automatically compiled with locale support09:16
xiangfukyak, ok. I don't know that. 09:18
xiangfuso I will revert the code. thanks09:18
kyakyeah, it's some recent change.. thanks to jow09:19
kyakand it also works :)09:19
viricmth: I let to you the fixing of the commit for the kernel, ok?09:22
mthviric: ok09:29
mthI don't understand his second comment though09:29
mth"And why this should be out of order case?"09:29
viricmaybe it is not an English sentence :)09:32
viricout of order means that it does not work, isn't ti?09:32
mthit can mean that ("elevator out of order"), but it can also mean that things are done in a different order ("out of order execution")09:40
mthalso, I don't know if "case" means "situation"/"instance" or refers to upper/lower case letters09:41
mthI mailed asking for a clarification09:48
mthactually, why does the JZ4740 need a different address? is it implied by the hardware somehow or is it just a convention that the existing tooling uses?10:07
mthfor Dingoo, I changed it because the existing boot loader required it, but I don't know if that address was originally picked at random or whether it was the only sensible option10:08
viricI think it's hardcoded in another place of the tree10:29
viricI recall little though.10:29
qi-bot[commit] Adam Wang: added libraries: nxp, diodes_inc; changed outlook same as datasheet as possible. (master) http://qi-hw.com/p/kicad-libs/3a983ff10:39
bartbeskyak: because of physfs? I'm pretty sure we've been building against 2.0.2 for a while now11:35
infobotHey docscrutinizer51, zip 62433 is Hutsonville Crawford IL11:54
infobotDocScrutinizer51: thanks11:54
DocScrutinizer51wpwrak: what's new?11:55
DocScrutinizer51and exciting11:56
DocScrutinizer51hmm, ok. so back to boring ClearCase mess, brunch break declared finished :-S11:59
wpwrakDocScrutinizer: not much new and exciting, i'm afraid. oh, maybe from daily politics - argentina implemented a de facto ban on book and newspaper imports, basically as an overnight surprise. and lifted it a few days later.12:17
viricmth: so we'll have to investigate why I committed such a particular case for jz474012:28
whitequarkDocScrutinizer51: you seen the atheros/qualcomm article?12:32
whitequarknamely http://www.phoronix.com/scan.php?page=article&item=qualcomm_kill_blobs&num=112:32
kristianpaulzrafa_: what you think about that python allegro, and gmenux replacement with a programable script?12:35
kristianpaulsounds powerfull consider no need write your own SDL stuff..12:35
kristianpaultough i may allow us better integration with console based apps12:36
qi-botThe build has FAILED: http://fidelio.qi-hardware.com/~xiangfu/building/Nanonote/Ben/openwrt-xburst.full_system-20120330-0825 12:46
zrafa_kristianpaul: should work. You already has pygame as well I think. python has bindings for everything.. BUt why you would like to replace gmenu2x?13:03
zrafa_kristianpaul: better integration: replace /usr/bin/gmenu2x with 13:04
zrafa_kristianpaul: /bin/bash ;)13:04
mththe plan we had for integration is to add a boolean to the icon info that indicates whether something is a console app or not13:08
mthand switch to a text terminal or not based on that13:09
kristianpaulzrafa_: i dont, i ask you about it :)13:21
kristianpaulzrafa_: i alredy replace by bash ;)13:23
zrafa_kristianpaul: cool :)13:42
mthlarsc (or anyone else who knows the answer): is it safe to assume that if the RTC is taken from an external clock, it will either be 32 kHz or a frequency near that value?13:52
mthnever mind, the issue only starts if RTC is over 40 GHz, which simply won't happen13:53
mthfor some reason I thought it could wrap at somewhat realistic freqs13:54
virica rtc at 40GHz may need einstein relativity correction13:57
kristianpaulzrafa_: ha i just saw you mail about genode-fx ;)14:13
kristianpaulso it compile under linux without minor/zero modifications14:17
zrafa_kristianpaul: yes. Using sdl as canvas. You can build several targets14:34
zrafa_kristianpaul: I just tried on my PC "make sdl"14:35
zrafa_kristianpaul: (and tested of course)14:35
viricmth: for the patch, add also a comment of *why this address*14:42
viricotherwise it's quite a magic number for reasons out of the kernel tree14:43
mthoh, I already sent a new patch14:43
mthand I don't really know why14:43
viricah you looked like having done a research, in your answer14:43
mthI know that booboo's u-boot uses that address and that you adopted it for the NN as well14:46
mthsince the original patch only forced it for the A32014:46
mthhow does a computed load address even work, by the way?14:47
mthI don't think injecting it into the boot loader would be an option, since then you'd always have to replace boot loader and kernel at the same time14:48
mthbut the alternative would be to have the boot loader figure out the load address at run time, which is extra overhead14:48
mthprovided that the vmlinuz.bin image even has the desired load address in it in a way that is discoverable14:50
zrafa_kristianpaul: I am downloading root.tar.gz from your host. Which has some tools for jlime and sie17:31
zrafa_kristianpaul: do you remember if you tried some test with fpga and your tools there?17:31
Jay7hello, ppl17:53
Jay7just quick question17:53
Jay7do NN have 8Gb NAND flash?17:53
Jay7i.e. is it NAND and is it 8Gb? :)17:54
wpwrakJay7: i'm certain it's NAND. but i think it's less (don't have a machine up to check). wikipedia says 2 GB.17:59
Jay7hm.. seems I confised a bit18:00
viricmine, 2GiB18:00
viricmth: the load address is determined at mkimage, isn't it?18:01
viricI mean... there is the kernel start point. Then we make a vmlinuz, which has a specific load address. Then we make mkimage of that vmlinuz, telling to mkimage the same load address.18:02
viricor mkimage takes vmlinux, and it compresses it? Why did I want vmlinuz at all?18:03
DocScrutinizerwhitequark: yes, very interesting18:22
zrafa_Jay7: wpwrak: if we need to check wikipedia to know the size of the nn nand then we are lost :)18:26
wpwrakzrafa_: well, the qi-hw wiki is a bit chaotic, wouldn't you say so ? :)18:42
wpwrakzrafa_: and the nand size hasn't troubled me yet, so it's not the sort if number i'd keep in my head18:42
MabYwenzrafa: my nn has 10cmx7.5cmx1.7cm in size :-D20:48
qi-bot[commit] Werner Almesberger: genex/: generate expanded component view (WIP) (master) http://qi-hw.com/p/eda-tools/490288623:03
qi-bot[commit] David Kühling: liballegro: cleanup and improve keyboard driver fix (master) http://qi-hw.com/p/openwrt-packages/115dfe523:11
qi-bot[commit] David Kühling: liballegro: add png, jpeg image and ogg/vorbis audio file addons (master) http://qi-hw.com/p/openwrt-packages/088d3e723:11
mthviric: I don't know why you wanted vmlinuz, but I wanted it because it loads quicker23:30
--- Sat Mar 31 201200:00

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