**** BEGIN LOGGING AT Thu Dec 06 03:00:00 2007 Dec 06 03:02:51 CoreDump: by chance, does anyone actually make an orinoco wireless card in CF form factor? that driver is the bain of my existance and as far as I know there is no reason to have it installed at all... Dec 06 03:03:08 I guess so =) Dec 06 03:03:46 CoreDump: I doubt it. Dec 06 03:03:51 I just broke my WLAN card :\ Dec 06 03:04:07 CoreDump: airo_cs is built as well and I am almost certain there is no CF cisco aironet card Dec 06 03:04:10 sucks big time to be forced to use USB net Dec 06 03:04:24 well, I dunno Dec 06 03:04:41 CoreDump: yes, I'd imagine it would. How did you break it? Dec 06 03:04:49 take that up with the mentor, but expect a _lot_ of bitching on the MLs if you remove the drivers Dec 06 03:05:25 I guess I sat on it one time too many =) I believe the problem is a broken lead on the PCB Dec 06 03:06:14 CoreDump: I know removing drivers is touchy, but when they don't make cards in that form factor... well... I'll investigate Dec 06 03:06:33 CoreDump: btw, was it you or hrw that fixed the backlight on akita/spitz? Dec 06 03:06:47 me Dec 06 03:06:59 CoreDump: turn off when I close it AND turns on when I open it Dec 06 03:07:06 yes Dec 06 03:07:09 CoreDump: so I suppose, THANKS A LOT :-) Dec 06 03:07:15 np ;) Dec 06 03:07:33 it's more of a work-around to a gpe-conf bug than a fix Dec 06 03:08:05 CoreDump: I couldn't possibly care less. it works :-) Dec 06 03:08:23 ;) Dec 06 03:09:44 CoreDump: now if it doesn't go out to lunch we are all set :-) Dec 06 03:10:21 it still doesn't come back on after the timeout via screen-tap or keyboard click Dec 06 03:10:49 CoreDump: but it does on open/close right? my kind of hack :-) Dec 06 03:10:58 yes Dec 06 03:11:19 CoreDump: it comes back and there is a predictable way to do it, that makes me happy. Dec 06 03:11:24 but it is annoying to be forced to close&open it after it dimms Dec 06 03:13:05 CoreDump: I know, but is that more or less annoying than having to reboot it? Dec 06 03:13:07 :-D Dec 06 03:13:23 true Dec 06 03:58:17 CoreDump: I know kernel isn't your job but by chance do you know: I menuconfiged and added a few things but they don't seem to end up packaged, I forget something? Dec 06 03:59:09 replace defconfig with your new .config and try again Dec 06 04:00:38 CoreDump: really? how rude, it lets me menuconfig then it doesn't use it? Dec 06 04:01:33 well, it should Dec 06 04:01:46 heh, I'll try it, I seem to recall you are correct Dec 06 04:01:48 but I never tried it that way Dec 06 04:24:39 http://www.flickr.com/photos/9706702@N08/2088608880/ lol Dec 06 04:27:29 wow indeed Dec 06 08:14:11 hi Dec 06 12:32:27 hi Dec 06 14:23:30 koen: Angstrom is a crucial part of getting Google Android to run on a Zaurus. http://euedge.com/blog/2007/12/06/google-android-runs-on-sharp-zaurus-sl-c760 Dec 06 14:26:47 rwhitby: I just realised I know benno.id.au :) Dec 06 14:26:58 he was at elc2007 as well Dec 06 14:30:58 koen: I'm keeping a wiki of all the good info about Android at www.android-internals.org Dec 06 14:31:25 benno features significantly there :-) Dec 06 14:31:49 * koen wishes we could legally upload an angstrom-android-image Dec 06 14:33:26 could someone try opening http://www.angstrom-distribution.org/unstable/autobuild/rss.php in an rss reader? Dec 06 14:41:27 842 articles Dec 06 14:42:30 great Dec 06 14:42:42 so it works Dec 06 14:42:53 hi Dec 06 14:43:07 dunno yet, still waiting for thunderbird to download 842 articles Dec 06 14:44:18 koen: yep, it works Dec 06 14:45:12 good Dec 06 14:45:22 so now we have an rss feed for autobuilder uploads Dec 06 14:45:51 * koen adds the script to contrib to scare hrw|gone Dec 06 14:51:39 koen: long time question ... there is something moving for FOSDEM ? Dec 06 14:51:51 gremlin[it]: You'd have to ask mickey|lunch Dec 06 14:52:18 oki Dec 06 14:52:37 i'll plan to be there :) Dec 06 14:52:57 no mistake .. i plan i'll be there ... Dec 06 14:58:02 * rwhitby heads to bed Dec 06 14:58:16 'night rwhitby Dec 06 14:58:56 if everyone in .eu has a party, you think we can keep rwhitby awake by shaking the world? Dec 06 14:59:37 http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2007-December/001075.html Dec 06 16:36:55 hi Dec 06 17:04:19 hello Dec 06 17:04:28 hi Freak_NL Dec 06 17:04:49 The updater.sh.spitz at http://www.angstrom-distribution.org/unstable/images/spitz/20071024/ is inaccessable Dec 06 17:05:03 I get a 403 error Dec 06 17:05:32 i noticed that too Dec 06 17:05:46 The zImage works though Dec 06 17:05:56 i was going to try the updater.sh from cacko, as i imagine it's probably the same Dec 06 17:06:06 but i forgot to bring my charger with me to work Dec 06 17:06:10 Sketch: just use an older one i'd imagine Dec 06 17:06:50 the only other spitz version in the tree has no updater.sh Dec 06 17:06:57 heh Dec 06 17:08:23 the updater.sh is inside all installkits Dec 06 17:08:34 is it the same for all machines? Dec 06 17:08:44 that's not what I said Dec 06 17:08:52 I said that updater.sh is in all installkits Dec 06 17:08:52 that's why I asked :) Dec 06 17:08:59 Freak_NL: all but akita for the time being Dec 06 17:09:13 Freak_NL: but koen is right, if you grab the installkit it has everything you need Dec 06 17:09:45 I'll use that, thanks (I was only updating the kernel, so I didn't use that) Dec 06 17:09:49 does anyone know if the images for akita and spitz are different other than the kernel and layout? Dec 06 17:10:11 Sketch: if I had to guess... very. spitz has a harddrive and akita doesn't Dec 06 17:10:29 Zero_Chaos: yeah, but they are basically the same otherwise Dec 06 17:10:49 at least between a 3100 and 1000 Dec 06 17:11:06 i think i'd rather install on the flash than the HD Dec 06 17:11:06 Sketch: well, aside the from the kernel, most of spitz image goes on the hd, and most of akita goes in flash, I'd imagine a difference Dec 06 17:11:26 Sketch: but I don't really know myself Dec 06 17:11:27 so i was considering trying the akita image, or maybe putting the spitz kernel onto the akita image Dec 06 17:12:10 Sketch: I'm going to guess that won't work, at the very least due to the kernel-modules... Dec 06 17:13:30 akita and spitz have very different hardware Dec 06 17:14:02 i would think the kernel would be the only part of the image that is really hardware specific Dec 06 17:14:10 (other than the hd vs flash layout) Dec 06 17:15:16 XorA|gone: did you get a chance to push the preferred version for kismet? Dec 06 17:15:17 mostly i just don't want to have to repartition my Z ;) Dec 06 17:15:28 but it seems that putting the OS on the HD would reduce battery life Dec 06 17:15:54 Zero_Chaos: yes, closed the bug and everything Dec 06 17:16:03 XorA|gone: sweet, thanks Dec 06 17:32:49 Anyone want to help me troubleshoot the bug just posted by Tim Chick in updater.sh. I'd be happy to troubleshoot if someone can give me something to test... Dec 06 17:42:36 Zero_Chaos: how about checking what the working and non-working script do different when flashing the rootfs? Dec 06 17:42:49 the decoder tool is in OE Dec 06 17:44:21 koen: I have tried, I have no idea what causes this. Dec 06 17:48:20 the flashing routine for the rootfs did not change Dec 06 17:50:05 kernel is flashed ok but rootfs gives errors? Dec 06 17:50:07 CoreDump: something is different because one works and the other doesn't. As a fact, there is so much different between the two that I can barely make heads or tails. Dec 06 17:50:13 hrw: seems to Dec 06 17:50:14 !oebug 3393 Dec 06 17:50:16 * * Bug 3393, Status: NEW, Created: 2007-11-29 05:58 Dec 06 17:50:17 * * autobuild(AT)openembedded.org: gnome-doc-utils-0.12.0-r1-do_compile Dec 06 17:50:18 * * http://bugs.openembedded.org/show_bug.cgi?id=3393 Dec 06 17:50:19 hrw: ML has the specific error Dec 06 17:50:35 !oebug 3339 Dec 06 17:50:36 * * Bug 3339, Status: NEW, Created: 2007-11-20 05:35 Dec 06 17:50:37 * * eha(AT)doredevelopment.dk: Upgrade to mpfr-2.3.0 Dec 06 17:50:38 * * http://bugs.openembedded.org/show_bug.cgi?id=3339 Dec 06 17:51:06 !oebug 3397 Dec 06 17:51:07 * * Bug 3397, Status: NEW, Created: 2007-11-29 09:40 Dec 06 17:51:08 * * hvontres(AT)gmail.com: linux-rp-2.6.17: add kexec atag support and make default for poodle Dec 06 17:51:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3397 Dec 06 17:51:47 Zero_Chaos: the diff looks ugly but the actual change is maybe 10 lines Dec 06 17:52:46 CoreDump: considering they two scripts do not even share variable names, i find it difficult to follow. Dec 06 17:52:57 CoreDump: but like I said, something is different, one works, the other doesn't. Dec 06 17:53:25 ah you're diffing the old .sh against the new one Dec 06 17:53:37 there lays the way of madness Dec 06 17:53:42 CoreDump: I'm not diffing anything, I look at them side by side. Dec 06 17:53:50 one only needs to compare teh methods that flash the rootfs, not the complete script Dec 06 17:53:59 the diff in my head works better than the diff on my computer Dec 06 17:54:07 "the flashing routine for the rootfs did not change" Dec 06 17:54:37 CoreDump: well then, it must be magic, because it doesn't work now. Dec 06 17:54:45 CoreDump: there might be a small typo :) Dec 06 17:54:53 true ;) Dec 06 17:55:28 I'm going to make lunch, if no one has fixed it by the time I'm done eating it, I'll look again, maybe I will see something I missed. Dec 06 17:55:49 but by all means, most of you are better equiped than I to fix this, so feel free to give it a shot. Dec 06 17:56:14 most of us don't have an akita to test on Dec 06 17:56:22 so no, we are not better equiped Dec 06 17:56:36 koen: I offered to be test whore, but I don't understand updater.sh so I am not well equiped to find the bug ;-) Dec 06 17:56:43 I already had a shot, but you said the kernel didn't boot, not that the rootfs is hosed Dec 06 17:57:30 CoreDump: jffs2 error I will assume is rootfs problem Dec 06 18:07:14 Zero_Chaos: current rootfs+kernel + old updater.sh works? Dec 06 18:07:35 hrw: if by "old" you mean "akita", yes Dec 06 18:08:17 try this one and write done the printed ROOTFS_SIZE http://hentges.net/tmp/updater.sh-akita03 Dec 06 18:09:20 CoreDump: 5 min please, desire to finish burger is temporarily exceeding desire to play with Z Dec 06 18:09:42 bbiab Dec 06 18:10:22 Zero_Chaos: mmmmm....burger.... Dec 06 18:10:34 * hrw -> bbiab too Dec 06 18:12:31 Zero_Chaos: it seems my fixes to the case-sensitive filesystems got removed somehow... Dec 06 18:12:54 http://linuxtogo.org/gowiki/ZaurusSpitz Dec 06 18:13:13 the oesf wiki says the C1000 uses same circuit board as C3000 but fitted with more flash Dec 06 18:13:45 I trust the oesf site as far as I can throw a piano Dec 06 18:13:52 so i guess the question is do the people writing these wikis just not know what they are talking about, or are the differenecs pretty minor? ;) Dec 06 18:14:57 koen: must have been because you and Coredump (iirc) commited your fixes at the same time. Double headed beast, and apparently a poor merge job Dec 06 18:15:34 koen: if you would be kind enough to recommit your fix, that would be keen. Dec 06 18:16:19 done Dec 06 18:16:49 thanks, I'll test that after we are done with this other issue. Dec 06 18:18:59 Does anyone know where to configure openmoko-terminal2? I think it can use a rc file, but it isn't installed by default Dec 06 18:22:21 Sketch: from user point of view there is nearly no changes Dec 06 18:22:23 Zero_Chaos: this is the diff you are testing now: http://rafb.net/p/6NwNDu87.html Dec 06 18:22:32 just removed second cf slot and added more flash Dec 06 18:22:44 CoreDump: that prints some crazy numbers... Dec 06 18:22:56 Sketch: from kernel view they differ more Dec 06 18:23:03 Zero_Chaos: give them to us Dec 06 18:23:08 the statemnt in the oesf wiki is like "a car is just like a bicycle, it only adds 2 more wheels" Dec 06 18:23:37 rootfs: [52953104] > [26214400] Dec 06 18:23:43 Sketch: akita has additional IO extender (which iirc also is in borzoi/terrier which are akita based) Dec 06 18:24:41 koen: that diff looks like it purposely breaks flashing rootfs... Dec 06 18:24:53 Zero_Chaos: "grep root /proc/mtd" - can you? Dec 06 18:25:15 Zero_Chaos: it uses the old hardcoded value instead of determining it from /proc/mtd Dec 06 18:25:50 hrw: mtd2: 03a00000 00020000 "root" Dec 06 18:25:59 koen: that number seems wrong to me... Dec 06 18:26:23 thx Dec 06 18:27:52 hi mimecar Dec 06 18:28:39 evening hr Dec 06 18:30:42 Zero_Chaos: well? Dec 06 18:31:21 CoreDump: rootfs: [52953104] > [26214400] Dec 06 18:31:30 CoreDump: mtd2: 03a00000 00020000 "root" Dec 06 18:31:42 what about the rootfs Dec 06 18:32:04 it's 51 megs Dec 06 18:32:07 it fits Dec 06 18:32:16 do you get the errors?? Dec 06 18:32:39 it didn't flash because you hardcoded the wrong flash size and it thinks my image is too big Dec 06 18:34:02 http://hentges.net/tmp/updater.sh-akita04 <- no size check Dec 06 18:34:09 heh Dec 06 18:38:50 DEBUG: ROOTFS_SIZE:[03a00000 vs 0x1900000] Dec 06 18:39:02 flashing still Dec 06 18:40:07 CoreDump: we have already been through the size thing together you know, the size checks all worked properly from what I saw, well, untill these changes... Dec 06 18:41:08 and the jffs error persists Dec 06 18:45:10 well, I give up Dec 06 18:46:17 CoreDump: it's not a size thing, there is something different about the way the akita updater flashes, or there is a bug (which I doubt because only akita has this issue) Dec 06 18:51:31 what is MVRBLOCK? Dec 06 18:52:40 koen's rss.php is really scary Dec 06 18:56:30 indeed Dec 06 18:59:26 I sent my idea to ML Dec 06 18:59:55 hrw: if borzoi/terrier are akita based, why does the angstrom wiki tell me to use the spitz image? :) Dec 06 19:00:11 Sketch: because from user point of view they are spitz Dec 06 19:00:37 spitz/borzoi/terrier are pxa27x with internal microdrive Dec 06 19:00:45 so they are spitz Dec 06 19:00:59 flash is used only for kernel due to sharp crap called bootloader Dec 06 19:03:46 i like the idea of using the flash since it should use less power Dec 06 19:04:02 http://www.newegg.com/Product/Product.aspx?Item=N82E16820211170 Dec 06 19:04:06 maybe i should just replace the hd with a flash card ;) Dec 06 19:04:10 that's pretty cheap, heh Dec 06 19:04:17 I do not like the idea of having only 53M for rootfs Dec 06 19:05:47 hrw: could you send me the encoder/decoder binaries? I had to wipe my TMPDIR and a rebuild takes forever Dec 06 19:05:52 true, it is a little limiting Dec 06 19:06:13 CoreDump: bitbake -b encoder.bb Dec 06 19:06:21 little? Dec 06 19:06:31 Sketch: when I had spitz I used 1GB rootfs Dec 06 19:06:39 and had ~300M of stuff installed Dec 06 19:06:59 hehe Dec 06 19:07:21 hrw: I can't, I'm missing the toolchain Dec 06 19:07:23 well, i've been using cacko, i don't think it even has that much software available ;) Dec 06 19:07:36 CoreDump: encoder is native stuff Dec 06 19:07:37 a few big things i did install i just manually installed on the hd and did some symlinking Dec 06 19:08:02 hrw: ugh, sorry for that Dec 06 19:08:08 np Dec 06 19:09:52 hrw: BTW, thanks for your blog post about KScope. That tool ROCKS for trying to grok kernel stuff :) Dec 06 19:13:02 ;) Dec 06 19:14:48 * CoreDump tries to flash spitz Dec 06 19:16:41 hello everybody! Dec 06 19:16:46 hey Dec 06 19:17:48 my Z looks every day better Dec 06 21:05:13 hmmm Dec 06 21:05:20 has anythin happened to keylaunch? Dec 06 21:05:48 i seem to have lost /etc/keylaunchrc.matchbox, leaving me with a broken symlink and no keylaunch functionality. Dec 06 21:06:11 its entirely possible i deleted it during a drunken moment, but i honestly don't recall having done so. Dec 06 21:36:56 evening Dec 06 22:04:12 is anyone developing for the ipaq hx2000 (or 2100)? **** ENDING LOGGING AT Fri Dec 07 02:59:57 2007