**** BEGIN LOGGING AT Mon Sep 05 03:33:57 2005 Sep 05 04:22:01 03hrw 07org.oe.dev * r7ad2dae7... 10/packages/qpf-fonts/ (10 files): upgraded DejaVu fonts in QPF format to 1.13-20050820 snapshot) Sep 05 04:22:18 hrw|work: yipee Sep 05 04:25:47 I had to add "makeqpf suxx" but it is not allowed by commit policy ;) Sep 05 04:26:17 or "'I hate tROLLtECH for their sucky font format' release" Sep 05 04:26:46 my opie-image for collie does not have free space ;( Sep 05 04:27:08 that does not sound too good Sep 05 04:27:33 yep Sep 05 04:28:09 Hi zecke Sep 05 04:28:13 ruediger: hey Sep 05 04:28:15 and I dont have cf writer here ;( Sep 05 04:28:17 ruediger: whats up? Sep 05 04:28:40 zecke: Kurze Frage, muss man den esg-browser auf jeden Fall installieren wenn man das neue mitv installiert? Sep 05 04:28:51 quite on topic ;) Sep 05 04:29:10 ruediger: well, yes we need the updated events.csv Sep 05 04:29:17 ruediger: and esg-browser provides this Sep 05 04:29:52 zecke: ok, säts si onli sing? Sep 05 04:30:02 heh Sep 05 04:30:13 * koen|slug get's flashback to 'Allo 'Allo Sep 05 04:30:36 arh.. Sep 05 04:30:42 ~lart sa1100_bi Sep 05 04:30:42 * ibot_ throws a AN/M-8 smoke grenade at sa1100_bi Sep 05 04:31:38 ruediger: *Timmy* that is the only thing Sep 05 04:31:47 UDC on 2.4.18-sharp-crap is pain Sep 05 04:49:34 hrw|work: only UDC? :) Sep 05 04:50:43 RP: if you dont load modules in right order in less then miliseconds one after other then goodbye keyboard Sep 05 04:50:48 and rest suxx too Sep 05 04:53:19 03rpurdie 07org.oe.dev * rd63c6ca7... 10/packages/linux/linux-openzaurus_2.6.13-rc5-mm1.bb: linux-oz-2.6: Mount the spitz hdd read/write for now until we sort out proper remounting rw after fsck logic into the initscripts Sep 05 04:56:31 almost 3k .bbs in the repo Sep 05 04:56:54 about 5 times as much when I first started Sep 05 04:58:43 I dont remember how much was when I started Sep 05 04:59:03 it was eons ago Sep 05 04:59:59 ChangeSet@1.1351.35.1 2004-05-08 19:54:49+02:00 szczepan@local Sep 05 05:00:03 was my first one Sep 05 05:01:38 first push atleast Sep 05 05:02:38 1.913 my first patch iirc Sep 05 05:15:46 hmm.. we have to remove opie-multikeyapplet from base image Sep 05 05:23:34 can someone give me a pointer as to what I'm doing wrong - first time through all this, and something's not right... got a C3K, trying to get gpe up and running (or opie-image, but that's still compiling). Sep 05 05:24:25 I assume its pivotboot-image as my initrd, the 2.6 kernel as my zImage, and the (gpe|opie)-image as the hddimage1.tgz? Sep 05 05:26:04 it boots ok, but leaves the rootfs as RO, so X can't start. Sep 05 05:26:15 gridzero: correct Sep 05 05:26:29 gridzero: See my last checkin! Sep 05 05:27:20 gridzero: I've just fixed it - you need to update and build a new kernel - the rest of the image is fine Sep 05 05:28:09 ah - cool. I'll give that a go then :) Sep 05 05:28:36 CoreDump|home: ping Sep 05 05:28:52 gridzero: Alternatively run "mount / -o remount,rw" on the device... Sep 05 05:29:15 RP: I would, if I could find the '/' key... :-( Sep 05 05:29:52 gridzero: I see the keymap issue to - need to find out why it doesn't load. Sep 05 05:30:07 gridzero: cd into /etc and loadkeys keymap-2.6 Sep 05 05:30:25 To avoid the / use "cd .." ;-) Sep 05 05:30:37 Can you tell I've done this recently? :) Sep 05 05:31:04 *grin* yip, cheers, thats got the keyboard sorted... Sep 05 06:10:31 hrw|work: busybox-1.01 seems ok to me ... other people may see issues i don't , but it builds fine and works fine too Sep 05 06:10:52 and deploys ok in image Sep 05 06:12:44 ade|desk: great Sep 05 06:12:44 hrw|work: there will be fewer patches needed for the upcoming 1.0.2 release i feel Sep 05 06:17:09 hrw|work: i would warn the distro people before letting it out though, i dont know how fixed the distro confs are... there maybe a few upset folks if busybox got updated without their knowedge Sep 05 06:18:09 ade|desk: thats why I didnt pushed it Sep 05 06:18:27 but I think that I will push it with default_preference=-1 so anyone can test it Sep 05 06:26:58 03hrw 07org.oe.dev * r97de9520... 10/packages/busybox/ (22 files in 7 dirs): Sep 05 06:26:58 added busybox 1.01 Sep 05 06:26:58 - DEFAULT_PREFERENCE = "-1" because it is not tested on all devices Sep 05 06:26:58 - works on my collie with OZ 3.5.4-20050830 Sep 05 06:31:00 hmm. http://bugs.busybox.net/view.php?id=407 looks nice - update-alternatives in busybox.. Sep 05 06:46:36 hrw|work: someone even wrote a patch to put ipkg in busybox Sep 05 06:46:48 ade|desk: wil 1.0.2 support a shared libbb? Sep 05 06:51:25 koen|slug: ipkg in busybox is overkill imho - no libipkg will render gui package managers useless Sep 05 06:56:31 it would save ~200k of flash Sep 05 06:56:45 since ipkg links to a static libbb Sep 05 06:57:06 but teaching busybox to have a shared libb would be even better Sep 05 07:03:52 koen|slug: re shared libbb , not that i know of Sep 05 07:04:17 possible 1.1 feature perhaps Sep 05 07:06:11 mutliple heads again, merged... Sep 05 07:15:18 good morning! Sep 05 07:15:57 hi mickeyl Sep 05 07:16:08 mickeyl: are you clean enough now? Sep 05 07:16:14 quote a shower... *phew* Sep 05 07:16:32 well, i had to prepare for a meeting with my ph.d subvisor Sep 05 07:16:38 supervisor Sep 05 07:16:40 even Sep 05 07:16:54 he wasn't too happy with my performance :/ Sep 05 07:17:03 mickeyl: you survived ;) Sep 05 07:17:16 "Mr. Lauer, you're brilliant when you do something. It's just that you don't." Sep 05 07:17:31 :} Sep 05 07:17:32 <--- Mr.Idle Sep 05 07:17:32 mickeyl: Ah, one of those meetings :) Sep 05 07:17:37 RP: *nod* Sep 05 07:18:10 mickeyl: We were worried you'd disolved ;-) Sep 05 07:18:15 hehehe Sep 05 07:18:16 nah Sep 05 07:18:18 I get poodle user(s) to test 2.4.18 poodle with collie fixes Sep 05 07:18:38 hrw|work: that's good. Sep 05 07:18:42 mickeyl: so, why are you here? Sep 05 07:19:06 zecke: a good question. i guess because i want to have at least one hour of fun per day ;) Sep 05 07:19:18 50 minutes left Sep 05 07:19:44 party ;) Sep 05 07:19:48 *scratches* head Sep 05 07:19:52 mickeyl: my opie-image, meta-opie changes needs more tweaking :( collie opie-image was too full.. Sep 05 07:20:36 hrw|work: in my tree there is no ipsec and openswan. in that case we have 0.9MB left Sep 05 07:20:58 i'm a bit reluctant to do the bootstrap reorganization we talked about before the next set of releases Sep 05 07:21:10 then again, we can't release on the 16mb models without that reorganization :/ Sep 05 07:21:36 mickeyl: did you see my rants about the calibration patch? Sep 05 07:21:50 mickeyl: opie-style-(liquid|metal) will be removed too Sep 05 07:21:51 zecke: sure. i already answered Sep 05 07:22:10 mickeyl: He means the lengthy irc conversation about it... Sep 05 07:22:16 ah Sep 05 07:22:19 no, i didn't read that Sep 05 07:22:21 do i need to? Sep 05 07:22:30 no Sep 05 07:22:38 I think everyone's happy now Sep 05 07:22:42 mickeyl: try http://ewi546.ewi.utwente.nl/tmp/viewmtn/ Sep 05 07:22:43 excellent Sep 05 07:22:49 that's viewmtn 0.04 Sep 05 07:23:07 woot Sep 05 07:23:12 where is graph Sep 05 07:23:31 branch -> more info -> graph Sep 05 07:23:31 ah Sep 05 07:23:33 yep Sep 05 07:23:40 well, that's definitly an improvement Sep 05 07:23:44 koen|slug: http://ewi546.ewi.utwente.nl/tmp/viewmtn/branch.psp?branch=org.openembedded.zecke well Sep 05 07:23:49 yep Sep 05 07:23:58 http://lists.gnu.org/archive/html/monotone-devel/2005-09/msg00029.html <- the announcement Sep 05 07:24:10 mickeyl: I emailed the author about moving the manifest Sep 05 07:24:26 koen|slug: thanks. did he respond yet? Sep 05 07:24:44 no, I suspect he's still sleeping (.au) Sep 05 07:24:56 right Sep 05 07:25:38 Or at least make the manifest a browsable tree... Sep 05 07:25:46 RP: thats in TODO Sep 05 07:26:03 Planned features for the next version are an improved manifest ("file") Sep 05 07:26:04 browser Sep 05 07:26:09 fair enough Sep 05 07:26:45 mickeyl: also notice the 'rss' link in the bottomright corner Sep 05 07:27:21 mickeyl: today/tomorrow evening I plan to 'fix' opie-image/meta-opie to get it working with collie better Sep 05 07:27:42 mickeyl: we have few kernel modules to remove from 2.4.18-sharp devices (registers and devinfo) Sep 05 07:27:51 hrw|work: and other 16MB flash devices Sep 05 07:28:10 koen|slug: ofcourse Sep 05 07:29:45 it will be hard to do it but it should be possible Sep 05 07:30:00 mickeyl: opie-multikeyapplet will get removed from opie-image Sep 05 07:32:54 cu Sep 05 07:40:27 http://pastebin.ca/22145 Sep 05 07:40:33 i guess that should improve bug reports a bit Sep 05 07:40:39 (note the first two new variables) Sep 05 07:41:33 nice Sep 05 07:41:59 now just make sure bitbake svn bumps PV immediately after release Sep 05 07:42:08 yeah, we need to rememeber that now Sep 05 07:42:48 s/MONOTONE/OE/ Sep 05 07:42:56 or put the SVN revision of bitbake into the string Sep 05 07:43:37 that's more complicated Sep 05 07:43:45 better just remember to update bb.__version__ Sep 05 07:43:56 it works also with previous versions and also with released versions now Sep 05 07:44:26 and i didn't have to adjust bb to make it work Sep 05 07:44:39 just tweaked base.bbclass Sep 05 07:44:52 bb.__version__ = "1.3.2 $rev$" Sep 05 07:44:58 good idea Sep 05 07:45:11 will you do that when you add the shell patch ? Sep 05 07:46:28 or shall I? Sep 05 07:47:18 mickeyl: I took the liberty of restarting monotone on vanille last weekend Sep 05 07:47:44 koen|slug: thanks Sep 05 08:05:06 Why not hardcode a version check into OE? Have a BB minimum version in the metadata? Sep 05 08:06:09 good idea Sep 05 08:10:30 I guess it wouldn't catch people using an old OE with a new BB but hopefully we're not going to break things like that often Sep 05 08:10:45 03mickeyl 07org.oe.dev * r3c4ef5d0... 10/packages/tinyscheme/tinyscheme_1.35.bb: add tinyscheme, a lightweight scheme interpreter Sep 05 08:10:50 03mickeyl 07org.oe.dev * r4d61b91c... 10/classes/base.bbclass: classes/base: show BitBake version and Monotone Revision (if working on a valid MT repository) in OE Build Configuration Sep 05 08:14:08 * mickeyl wanders home Sep 05 08:21:31 Good evening Sep 05 08:21:53 hi sirfred Sep 05 08:22:17 Hi RP. Sep 05 08:28:21 is there any faster way to pull changes? monotone pull / monotone update takes (literally) hours on an athlon xp-2500 .. for what is usually a small number of changes.. Sep 05 08:28:37 update doesn't take hours Sep 05 08:29:35 update more often or get monotone.cvssync to work Sep 05 08:29:44 it did when I ran it yesterday, and it did when I ran it friday, and ... etc.. 'hours' may tend to be a low number (2-3) but its still >1 Sep 05 08:30:48 so the answer to my question is "no, that is the current state of things"..? thats fine, was just checking. Sep 05 08:32:44 Disconnect: Most of us run pull at least daily... Sep 05 08:34:23 once konq-e finishes (or fails) I'll rerun it with a typescript, report actual to-the-second numbers. but even running daily it seems to take excessively long. Sep 05 08:36:24 BB_VERSION = "1.3.2" Sep 05 08:36:24 OE_REVISION = "4a842e382979d8da86f6983ae2e0414b5b0b1e14" Sep 05 08:36:30 works great :) Sep 05 08:36:45 hi - got a little hidp problem.root Sep 05 08:37:11 hidp problems are usually distro rele Sep 05 08:37:14 related Sep 05 08:37:27 no- kernel related in that case I think Sep 05 08:38:39 I still bet it isn't OE related Sep 05 08:39:03 nah - just looking for kernel suggestions - bluez is a bit empty Sep 05 08:39:41  /dev/input/event1 report hidp events, while nothing shows up, even if input and input_keybdev are in the kernel... Sep 05 08:41:36 and how does that relate to OE? Sep 05 08:41:43 tl,d Sep 05 08:41:58 told you - just looking for kernel suggestions... Sep 05 08:42:24 and how does that relate to OE? Sep 05 08:43:03 I'd say #kernel or #bluez or #notrelatedtoOE would be better places for such questions Sep 05 08:43:14 the closest relation I can think of is the zipit (zipitwireless) may run oe someday. Sep 05 08:43:53 try #bluez yourself, you'll see how crowded it is. that and I had the very same bug happen to me when I tried a former version of OE on a friend zaurus, so I though you had solved that bug Sep 05 08:49:17 you can't run OE Sep 05 08:49:21 OE isn't a distro Sep 05 08:49:41 is it runs on a device, its a distro Sep 05 08:49:50 which leads me to point to the topic Sep 05 08:52:26 I guess your obstination means you don't know the answer. even if I last found that bug was present in a device running oe, i'll try to find the answer elsewhere. thanks for not helping. Sep 05 08:52:46 um. the device runs "make" and "gcc" too.. go bug them. Sep 05 08:52:47 again Sep 05 08:52:52 a device can't run OE Sep 05 08:52:54 betcha it even runs bash, maybe its a bash bug. Sep 05 08:53:09 OE is a metadata collection Sep 05 08:53:18 it doesn't even contain binaries Sep 05 08:54:09 when the metadata configure apps, it sometimes misconfigure them. I am experiencing the very same problem- I just wanted to know if someone remembered of any recent changes in bluez utils Sep 05 08:54:54 so you're running an OE generated image? Sep 05 08:56:57 no - custom image. it just happens to share bugs with oe :/ Sep 05 08:58:28 so it's COMPLETELY off-topic here Sep 05 08:58:33 just checking Sep 05 09:07:57 guylhem: I thought you used OE to generate your images? Sep 05 09:08:34 guylhem: I remember reading that on oesf back when you produced the first one Sep 05 09:28:16 | ERROR: Cannot satisfy the following dependencies for opie-button-settings: Sep 05 09:28:18 | opie-button-settings-pics Sep 05 09:28:32 the h3900 opie-image fails on that Sep 05 09:32:02 aha Sep 05 09:32:21 another case of multiple arch per .bb Sep 05 09:32:41 it should prefer the h3900 one shouldn't it? Sep 05 09:33:04 if I build it for h3600 before it thinks it doesn't have to rebuild it Sep 05 09:33:11 ah righto Sep 05 09:33:24 hmm Sep 05 09:33:29 * koen|slug expands the clean list a bit Sep 05 09:34:28 sysvinit, tslib and base-passwd are the other packages I identified doing that Sep 05 09:50:40 any volunteers for adding support fo building OpenSolaris images? Sep 05 09:51:19 for* Sep 05 09:52:19 lardman - was - I try to do my own stuff Sep 05 09:55:25 guylhem: Ah, I see Sep 05 09:55:57 i may be ignorant, but I don't see OpenSolaris being interesting for the devices we mainly support Sep 05 09:56:14 ok.. so today (about 16-18 hours after my last update) it only took 40 minutes. Sep 05 09:57:27 mickeyl: apart from the gimmick value and vmware images I don't see added value either Sep 05 09:57:33 Disconnect: yeah, svn oneway would be useful ... Sep 05 09:58:04 * koen|slug hints at the monotone.cvssync branch of monotone Sep 05 09:58:15 koen|slug: *nod* although, in general, i'd like to have another OS than linux in OE. I bet we would have to remove a lot of Linuxisms then which strikes me as a good thing to enhance the flexibility of OE Sep 05 09:59:29 supporting different inits (sysv, bsd, init-of-the-day) would be nice too Sep 05 10:06:43 RP: which kernel do you recommend for 3.5.4 on c7x0? .13-rc5-mm1 or 2.6.13-mm1 ? Sep 05 10:07:10 RP: btw., you changed spitz defconfig only in rc5-mm1 and changed the backlight driver only in mm1. was that intentional? Sep 05 10:07:56 mickeyl: not that you asked me ;) but fwiw I went back to .11 on my c760 since .13-* seems to have squirrely fb support. (black screen when opie starts up, no fb splash, etc.) Sep 05 10:07:57 I was going to ask that too Sep 05 10:08:38 Disconnect: black screen is due to the now fixed backlight setting bug Sep 05 10:08:56 looked like the bl was on, but ok. I'll retest once I get dev moved over to the mac. Sep 05 10:09:12 Disconnect: splash is a matter of OZ_KERNEL_ENABLE_ELPP = "yes" in local.conf Sep 05 10:09:18 yes, it should be in oz.conf some time... Sep 05 10:09:44 pH5: I noticed the fstab for the hx4700 is wrong, busybox doesn't know about 'user' Sep 05 10:10:44 ~jlime Sep 05 10:10:45 koen|slug: thanks. i think the sd/cf mounting still needs some love, anyways. Sep 05 10:11:37 s/s// Sep 05 10:12:22 I noticed when I wanted to install hostap from a cf card Sep 05 10:13:34 koen|slug: but if we just remove it, all files on the cf belong to root, don't they? Sep 05 10:14:00 03koen 07org.oe.dev * rf4dadd60... 10/packages/base-files/base-files/ipaq-pxa270/fstab: packages/base-files/base-files/ipaq-pxa270/fstab: remove user from mount options Sep 05 10:14:01 I think so Sep 05 10:14:19 belonging to root is a step up from not being mounted at all, though Sep 05 10:14:27 ~seen woglinde Sep 05 10:14:33 woglinde was last seen on IRC in channel #oe, 3d 18h 8m 21s ago, saying: 'night '. Sep 05 10:15:11 koen|slug: definitely Sep 05 10:15:26 * Disconnect got corsair building, but only through brute force. will redo and submit later. (possibly later today, depends on how fast fink builds subversion) Sep 05 10:26:16 RP: ping Sep 05 10:42:05 03koen 07org.oe.dev * r8ff94495... 10/packages/base-files/base-files_3.0.14.bb: packages/base-files/base-files_3.0.14.bb: bump PR Sep 05 10:50:21 *sigh* Sep 05 10:50:24 why does wlan-ng suck so much?= Sep 05 10:52:44 mickeyl: because TT supports it ;) Sep 05 10:54:07 * mickeyl patches the hell out of it Sep 05 10:54:44 NOTE: package wlan-ng-modules-0.2.2: completed Sep 05 10:54:46 wlan-ng? Sep 05 10:54:47 *phew* Sep 05 10:54:51 wlan-ng! Sep 05 10:54:51 working on tosa support? Sep 05 10:55:04 ya Sep 05 10:55:12 tosa deserves 3.5.4 as well Sep 05 10:55:38 mickeyl: you've got another voting for that in #openzaurus Sep 05 10:57:04 aw crap. is monotone's OE.db endian-specific? Sep 05 10:58:12 not that I know off Sep 05 10:58:26 I move it around between x86, powerpc and armeb all the time Sep 05 11:00:07 hmm. found the problem. something ate the db. sigh. Sep 05 11:00:21 bye everyone Sep 05 11:00:25 mickeyl: Werner is on holiday until the 11th Sep 05 11:00:36 ~tosa Sep 05 11:00:38 from memory, tosa is the Sharp SL-6000 PDA, or a dog (see http://www.dogbreedinfo.com/tosa.htm) Sep 05 11:00:44 Disconnect: no it is a sqlite3 database Sep 05 11:01:06 mmh Sep 05 11:01:09 ah I'm getting the Tosa for the next three month Sep 05 11:01:32 03mickeyl 07org.oe.dev * rb3d467de... 10/packages/wlan-ng/ (8 files in 3 dirs): wlan-ng: add modules and userland utilities 0.2.2 (final) Sep 05 11:01:42 mickeyl: I recommend the -mm1 over the -rc-mm1 Sep 05 11:02:03 My commit was from an old build directory and hit the wrong file Sep 05 11:02:11 * RP adds it to the other one Sep 05 11:02:12 can we remove rc then? Sep 05 11:02:18 i think one stable, one devel should be enough Sep 05 11:02:38 good evening pb_ Sep 05 11:02:40 mickeyl: I'm toying with removing 2.6.11 as its rather behind the times Sep 05 11:03:00 but -mm1 and -rc-mm1 are too similar to need two... Sep 05 11:03:21 *nod* Sep 05 11:03:32 good evening mickeyl Sep 05 11:03:55 glad to see you survived the shower. Sep 05 11:03:56 hi pb_ Sep 05 11:03:59 hi rp Sep 05 11:04:50 mickeyl: Should we remove both? I have no intention of working on 2.6.11 any more... Sep 05 11:04:55 go ahead Sep 05 11:05:06 if we need it to compare something, it's still in the repositor Sep 05 11:05:07 y Sep 05 11:05:20 heh, I need a killfile for my email client. Sep 05 11:05:28 mickeyl: right Sep 05 11:05:40 pb_: for what patterns in particular? Sep 05 11:05:55 justin patrin, mostly. Sep 05 11:06:02 oops Sep 05 11:06:05 what did he do? Sep 05 11:06:19 having him ignored on irc is all very well, but his postings to the mailing list are just as irritating. Sep 05 11:06:36 *g* Sep 05 11:06:48 heh, oh well Sep 05 11:07:03 he did some good work on OE, he's just slightly confused sometimes ;) Sep 05 11:07:17 sometimes more often than at other times Sep 05 11:07:17 heh Sep 05 11:07:24 ah, fair enough Sep 05 11:09:50 * RP remembers why he hates forums Sep 05 11:14:06 just did a pull and build of linux-openzaurus-2.6.13-rc5-mm1-r20 - still black screen when opie lights up. (no change other than reflash zimage) .. Sep 05 11:14:17 wait.. scratch that. just need to give it a few minutes (!) Sep 05 11:14:28 ah. nevermind, I see... its inverted. fun. Sep 05 11:14:42 ? Sep 05 11:15:09 if I leave it idle, instead of turning off the bl it comes on low. soon as I 'wake' it it goes black.. Sep 05 11:17:03 Disconnect: Use 2.6.13-mm1 which has the bug fix, not 2.6.13-rc5-mm1 Sep 05 11:17:18 I'm deleting 2.6.13-rc5-mm1 as we speak... Sep 05 11:17:35 ah ok. cool. (just did 'rebuild linux-openzaurus') thx Sep 05 11:19:14 03rpurdie 07org.oe.dev * rbb8dba2f... 10/packages/linux/ (21 files in 3 dirs): Sep 05 11:19:14 linux-oz-2.6: Add the rw option to the correct spitz kernel. Remove the two Sep 05 11:19:14 older versions as they are out of date and 2.6.13-mm1 should be better in all Sep 05 11:19:14 respects. If there are bugs, please report them so we can fix them. Sep 05 11:23:01 03rpurdie 07org.oe.dev * r21490fb9... 10/packages/libxsettings-client/libxsettings-client_0.14.bb: libxsettings-client: Remove an unneeded gpe dependency Sep 05 11:26:19 03koen 07org.oe.dev * r1ba5e5da... 10/packages/gaim/gaim_2.0.0-cvs-mco1.bb: gaim-2.0: use CVSDATE in URI Sep 05 12:02:11 why does there have to be an RD and an RP? Sep 05 12:03:36 it's very confusing :P Sep 05 12:06:58 pb_: hehe, I've considered it as well Sep 05 12:14:21 i have decided that all languages suck :P Sep 05 12:15:22 heh Sep 05 12:15:32 what's wrong with python? Sep 05 12:15:47 yep new rule is directly reprograma a FPGA by telecinesis :) Sep 05 12:17:32 gb2: forced indentation Sep 05 12:17:41 gb2: s/languages/human spoken languages/ Sep 05 12:17:52 mithro: heh Sep 05 12:18:03 koen: i see that as a plus Sep 05 12:18:07 not a burden Sep 05 12:18:22 computer languages rock compared to spoken languages Sep 05 12:19:39 ... CFG rock ... context free grammar ... Sep 05 12:20:00 mithro: try brainfuck as a language ;) Sep 05 12:20:35 gremlin[it]: ? Sep 05 12:21:22 all languaged recognized by computer (or in general aumata) are based on a CFG ... Sep 05 12:21:30 i wish the python profiler was a bit more like oprofile Sep 05 12:22:25 as CFG words say the meaning of a meaninful sentence is the alwais the same ... despite de context ... what absolutelly never happen to a natural language :) Sep 05 12:23:06 i'm just remembering what i studied some years ago :) Sep 05 12:33:46 03koen 07org.oe.dev * r91cea4e0... 10/packages/gpe-bluetooth/gpe-bluetooth_0.45.bb: gpe-bluetooth: add 0.45 Sep 05 12:33:50 03koen 07org.oe.dev * r364d4e2f... 10/conf/distro/preferred-gpe-versions-2.7.inc: conf/distro/preferred-gpe-versions-2.7.inc: update to gpe-bluetooth 0.45 Sep 05 12:36:34 how do i get more information about why a patch failed to apply? Sep 05 12:37:30 in the .rej ? Sep 05 12:37:51 it says Sep 05 12:37:53 No file to patch. Skipping patch. Sep 05 12:37:58 only i can't figure out why Sep 05 12:38:57 negative karma. go pet a dog or something. ;) Sep 05 12:41:27 hrm.... Sep 05 12:41:29 missing header for unified diff at line 3 of patch Sep 05 12:41:29 can't find file to patch at input line 3 Sep 05 12:42:06 gb2: what pisses me off is they didn't use consistent white space in python, they should have made it 1 tab and that was that Sep 05 12:47:07 anyone any ideas? Sep 05 12:47:29 check the patch format..? Sep 05 12:48:36 it looks okay to me Sep 05 12:53:40 hi Sep 05 12:54:47 CoreDump|home: ping Sep 05 12:56:56 hey hrw Sep 05 12:57:47 wow, its really bright out. Sep 05 12:57:52 why c7x0 is so slooow... Sep 05 12:58:19 * Disconnect decided to take the space blanket out of the window and switch to outside-air cooling for a while. not as good an idea as it seemed downstairs. Sep 05 13:00:00 hrw: and slow at doing what? Sep 05 13:00:04 as pda's go its pretty zippy Sep 05 13:00:21 Disconnect: when I'm trying to connect to it via ssh, or when using ipkg Sep 05 13:01:01 yah ipkg is slow as hell. not sure why. seems faster now that I updated the kernel but that could be unrelated (made a bunch of changes on the feed at the same time) Sep 05 13:05:40 hmm.. dont know why qpe take most of cpu time.. Sep 05 13:05:53 PID USER STATUS RSS PPID %CPU %MEM COMMAND Sep 05 13:05:53 4248 root R 10M 1 98.8 16.7 qpe Sep 05 13:06:21 hmm. here too. and its not even active (clamshell closed, bl off, etc) Sep 05 13:06:47 2454 root R 10M 1 91.2 17.2 qpe Sep 05 13:06:50 hrw: it's punishing you for using a gpl qte Sep 05 13:06:56 prolly Sep 05 13:07:04 koen|slug: probably Sep 05 13:07:14 too bad gpe is such a mess right now. Sep 05 13:07:22 * RP wonders if mickey_away's key switch code could do this... Sep 05 13:08:33 the code which pulls input dev? Sep 05 13:09:14 Disconnect: it's working perfectly over here: http://handhelds.org/scap/port.6422.png Sep 05 13:10:16 show me a decent pile of pim apps, and a web browser that is responsive (ie NOT firefox) and vaguely complete (not minimo) and I'll retry. (well.. I'll work harder to get it built again, since thats what stopped me last time) Sep 05 13:10:46 do a real miracle, show syncing ;) but since there isn't a unified sync plan anywhere (including commercial apps) thats more a joke than a comment. Sep 05 13:11:22 * Disconnect can sync his phone to his mac i* apps.. or sync opie to the mac kde apps.. or sync wince to evolution..... whee. Sep 05 13:11:24 Disconnect: add not-so-configurable 'start' menu Sep 05 13:11:38 Disconnect: opensync syncs to gpe Sep 05 13:11:40 I must admit I find opie a lot more polished and usable... Sep 05 13:11:49 and even the kde hippies are using opensync Sep 05 13:11:53 Not that I ever actually use my handhelds mind :) Sep 05 13:11:53 yah. tahts right.. I got it flashed the other day.. briefly.. I think it had no sd and no cf drivers or something. Sep 05 13:12:10 * Disconnect played with it a while, found it to be a big improvement from last year but .. Sep 05 13:12:44 the opie I have on my h3870 is really polished Sep 05 13:12:49 I have to resize every window.... Sep 05 13:13:05 ? Sep 05 13:13:13 thanks to our great friends at TT Sep 05 13:14:03 Disconnect: ljp broke window placement and sizing Sep 05 13:14:03 koen|slug: bump cvsdate Sep 05 13:14:16 koen|slug: and mickeyl fixed it few days ago Sep 05 13:14:33 ok. Sep 05 13:14:35 hrw: yeah, but it's on my ipaq now Sep 05 13:14:42 and I don't think it's polished Sep 05 13:14:44 * Disconnect waits to hear why this is them being malicious... Sep 05 13:14:50 it's pure crap thanks to that bug Sep 05 13:14:54 ~hail TT Sep 05 13:14:56 * ibot_ bows down to TT and chants, "I'M NOT WORTHY!!" Sep 05 13:15:15 hahahhaaha.. thats awesome. "There was a bug in a development version.. and I haven't upgraded past it.. so it sucks and its all their fault." thats awesome. Sep 05 13:15:39 koen|slug: upgrade one package then Sep 05 13:15:45 Disconnect: no, my point is that opie is broken because some idiot at trolltech broke stuff Sep 05 13:15:48 * Disconnect notes that linux on the zaurus is totally fscked up .. whenever it boots the backlight turns off and won't come back on. (no, wait, that was a bug in a development version and I upgraded past it....) Sep 05 13:16:36 Disconnect: the same idiot who posts to gtk lists that qpe and opie are o so polished Sep 05 13:16:57 koen|slug: see above. and my example is a kernel bug, broken (and fixed) by open-source types. does it make them idiots? no.... Sep 05 13:16:58 just pointing out th irony Sep 05 13:17:32 Disconnect: I'm saying TT is pulling crap Sep 05 13:17:34 koen|slug: i will push one change to ipkg 0.99.153 - can you build any image with that version/ Sep 05 13:17:38 and that opie devs end up fixing it Sep 05 13:17:45 http://computing-dictionary.thefreedictionary.com/bleeding+edge Sep 05 13:18:05 koen|slug: ljp is also opie dev... Sep 05 13:18:12 hrw: heh Sep 05 13:18:17 opie-breaker Sep 05 13:18:20 try comparing release versions some time. Sep 05 13:18:32 or at LEAST rc/prerelease/etc .. not random cvs pulls.. Sep 05 13:18:46 cvs_: show me release tarballs Sep 05 13:19:00 Disconnect: it would be nice if opie will release separate apps not only whole revision Sep 05 13:19:15 (wanna keep comparing random cvs pulls? great.. the other day oz-opie built great for me. and oz-gpe didn't. even after I fixed sylpheed and a couple others. guess it means gpe is in pre-alpha and opie is polished and done, right?) Sep 05 13:19:46 koen|slug: http://opie.handhelds.org/feed/source/stable/opie-source-1.2.0.tar.bz2 Sep 05 13:20:14 hrw: that's a tarball, not tarballs ;) Sep 05 13:20:17 koen|slug: read http://opie.handhelds.org/cgi-bin/moin.cgi/Source (available from left menu on opie homepage) Sep 05 13:20:23 koen|slug: thats the source Sep 05 13:20:27 bbl, tired of watching opie-mediaplayer build Sep 05 13:20:28 Hello, Could anypne show me where I can read about -native builds? I want to make .bb for enca :-/ Sep 05 13:21:20 Disconnect: sylpheed isn't part of gpe Sep 05 13:21:51 Disconnect: and yes, gpe 2.7 is an unstable version in a feature freeze right now Sep 05 13:22:57 anyone here know much about patch? Sep 05 13:24:18 i can't figure out what is wrong with this patch :/ Sep 05 13:25:47 03hrw 07org.oe.dev * re606e184... 10/packages/ipkg/ (ipkg-0.99.153/fix-bug1393.patch ipkg_0.99.153.bb): Sep 05 13:25:47 ipkg: added my fast fix to get ipkg 0.99.153 working Sep 05 13:25:47 patch already submitted to upstream bugzilla: Sep 05 13:25:47 http://handhelds.org/~bugzilla/show_bug.cgi?id=1393 Sep 05 13:26:22 koen|slug: with that patch ipkg 0.99.153 is usable - can you test? Sep 05 13:26:48 hrw: tomorrw evening Sep 05 13:26:52 ok Sep 05 13:27:47 hrw: is that to make it check depends:? Sep 05 13:27:51 NAiLZzZ: yes Sep 05 13:28:03 brilliant Sep 05 13:28:53 I can test it when I get home Sep 05 13:29:01 NAiLZzZ: if it works for you then comment it in bugzilla Sep 05 13:29:14 sure Sep 05 13:30:49 thx Sep 05 13:31:28 Hey, can anyone here explain me how to make correct -navide .bb? :-/ Sep 05 13:31:43 *-native Sep 05 13:31:56 realloc: do normal one, add "inherit native" and proper do-stage90 Sep 05 13:31:59 realloc: it's the same as a normal .bb, just but inherit native on top Sep 05 13:32:09 do_stage() (shift stop working) Sep 05 13:33:13 Well... I want to make bb for enca library. It uses make_hash programm during compilation. How can native build help in this situation? Do I miss something? Sep 05 13:33:41 you need make_hash native Sep 05 13:34:36 YeS :)) pigi will check my patch Sep 05 13:35:34 hrw: What is "proper do-stage()" ? Sep 05 13:35:58 realloc: look into other -native Sep 05 13:36:10 realloc: you need to 'install' make_hash into staging Sep 05 13:37:05 Whaen I bitbake enca, how it will determine that it needs to build native bb? Sep 05 13:38:32 add native into DEPENDS Sep 05 13:39:50 hrw: thnx... Will try it... Sorry, I'm quite new to writing bb files :-/ Sep 05 13:40:24 realloc: each of us had to pass through it Sep 05 13:42:04 hrw: I like OE, but it's so unstable... I can't build qt-x11-free... Erlyer I was able to build gpe-image, but now I get errors... But it's still very interesting :) Sep 05 13:45:53 realloc: never tried to build qt-x11 Sep 05 13:46:47 do we have a paste bin? Sep 05 13:47:00 pastebin.ca Sep 05 13:47:09 or pastebin.com Sep 05 13:47:11 hrw: I wanted to have book-reader. justreader uses QT. Now I want to build fbreader, but it uses enca, so i try to make bb for it Sep 05 13:47:15 or use the pastbin command on bbshell Sep 05 13:47:50 opie-reader is good stuff Sep 05 13:48:09 I prefer justreader Sep 05 13:48:20 Disconnect: I use GPE. Opie doesn't work correctly on SL 6000L Sep 05 13:48:44 ah ok Sep 05 13:49:09 Disconnect: I get strange orange line on the screen and noone knows how to fix it :( Sep 05 13:49:38 Disconnect: Well orange and blue noise around changing objects Sep 05 13:49:59 realloc: under gpe someone noticed them too Sep 05 13:50:37 http://pastebin.com/355554 Sep 05 13:50:49 can anyone see the problem with this patch? Sep 05 13:50:58 hrw: I found modified version of Xfbdev without that bug Sep 05 13:51:37 realloc: so submit a patch Sep 05 13:52:18 hrw: There is no patch :) Only a binary, so I can't fix it Sep 05 13:52:59 use the gpl.. hit up the person you got the bin from for sources Sep 05 13:53:26 realloc: so it is not modified but 'who knows what' Sep 05 13:54:02 Disconnect: My friend sent me it, but he found it on some chinese site and he can't remember where Sep 05 13:54:26 hrw: the patch is rotting away in upstream kdrive bugzilla Sep 05 13:54:38 (iirc) Sep 05 13:55:05 koen|slug: but noone want it too much so it does not appear in OE Sep 05 13:55:19 it seems that way Sep 05 13:55:29 ~lart tosa users Sep 05 13:55:30 * ibot_ pries tosa users's back open with a screwdriver and flashes a new bootldr to tosa users Sep 05 13:55:48 koen|slug: I want it! :) Sep 05 13:56:08 realloc: so dig it, test it and submit to us Sep 05 13:56:19 and pester upstream to get it applied too Sep 05 13:56:42 hrw: I also seem to recall someone thinking oesf is a good place to publish patches for kdrive Sep 05 13:57:06 hrw: I can't build gpe-image now :)) I got errors building it after doing `monotone upadate` yesterday Sep 05 13:57:11 hint: kdrive bugzilla ends in /dev/null Sep 05 13:57:58 koen|slug: dont you know that Invision Power Board got "Best BugTracker" award? Sep 05 13:58:03 realloc: you can use http://ewi546.ewi.utwente.nl/sfeed/ to update (not upgrade!) packages if you have backups Sep 05 13:58:27 koen|slug: award was given by brainless Z users ofcourse Sep 05 13:59:15 cu Sep 05 13:59:24 buying Z lower IQ Sep 05 13:59:38 so its good that my second Z was donated ;) Sep 05 13:59:48 good night and sweet dreams for all Sep 05 14:00:03 so.... can anyone see why that patch would be giving me Sep 05 14:00:05 missing header for unified diff at line 3 of patch Sep 05 14:00:05 can't find file to patch at input line 3 Sep 05 14:03:46 mithro: try to recreate the patch Sep 05 14:08:16 did, still same problem :/ Sep 05 14:11:15 mithro: where is the patch? Sep 05 14:11:21 pb_: Justin just managed to annoy me... Sep 05 14:12:27 http://pastebin.com/355554 Sep 05 14:14:14 mithro: oh Sep 05 14:14:23 mithro: I wish I could remember Sep 05 14:14:52 looks like missing file info.. (patches i'm reading all have tiemstamps and such on the file line.) not sure if thats required, but.. Sep 05 14:14:53 mithro: normally you have times (but that is not mandantory) Sep 05 14:15:09 might be that quilt (or whatever) requires them Sep 05 14:15:23 (yes, it'd be out of spec, but..) Sep 05 14:16:16 docs I'm finding online say timestamp is required. (well.. docs I'm finding say everything is required, so..) Sep 05 14:16:45 mithro: the file numbers look a bit odd Sep 05 14:17:16 yah. this is a patch generated from 'diff -u' right? Sep 05 14:17:56 diff -Nurd is what we advice to use Sep 05 14:20:08 whats the N and d do? Sep 05 14:20:57 for a single file i don't think the -Nrd make any sense Sep 05 14:21:52 so.... again.. did you generate this with diff or did you use some other tool? Sep 05 14:21:57 diff Sep 05 14:22:10 diff -u Makefile.orig Makefile Sep 05 14:22:41 hmm. cuz my diff -always- includes the timestamps. Sep 05 14:23:11 using diff (GNU diffutils) 2.8.1 Sep 05 14:23:11 when i do a "patch -p0 < Makefile.patch" it works fine Sep 05 14:24:01 thats irrelevant since it gets applied by quilt, not patch. and here is also 2.8.1 Sep 05 14:24:53 hrm -- maybe it's because it needs -p0 instead of the normal -p1 Sep 05 14:25:10 yah.. actually.. didn't pick up on that. Sep 05 14:25:20 also.. try http://pastebin.com/355581 if that doesn't work Sep 05 14:28:36 ahh ha! Sep 05 14:28:48 needed Sep 05 14:28:50 file://Makefile.patch;patch=1;pnum=0" Sep 05 14:28:52 not Sep 05 14:28:59 file://Makefile.patch;patch=0" Sep 05 14:33:14 DEPENDS is what is required to build it, RDEPENDS is what is required to run it right? Sep 05 14:33:23 yes Sep 05 14:33:37 use RDEPENDS for stuff that shlibs doesn't pick up Sep 05 14:33:49 (fonts, pictures, etc) Sep 05 14:34:25 kernel modules? Sep 05 14:34:31 that too Sep 05 14:34:38 ehm no Sep 05 14:34:45 use RRECOMMEND for kernel-modules Sep 05 14:34:49 ? Sep 05 14:35:16 why? this software doesn't run without the kernel module? Sep 05 14:36:11 (it's a driver for bluetooth headsets - it has a kernel module and a userspace part) Sep 05 14:36:12 no idea, I was told to use RRECOMMENDS previously Sep 05 14:36:17 btsco? Sep 05 14:36:26 yeah Sep 05 14:36:40 mithro: suppose I staticly compile those modules into the kernel.... Sep 05 14:36:45 koen|slug: don't tell me it's already in OE? Sep 05 14:37:02 afaik it isn't in Sep 05 14:37:15 good because otherwise i'm blind :P Sep 05 14:38:03 mithro: suppose I have a sucky bootloader and I want all modules in the kernel Sep 05 14:38:15 with RDEPENDS btsco refuses to install Sep 05 14:38:39 thinking about it should that kernel provide kernel-module-btsco then? Sep 05 14:38:56 in theory, yes Sep 05 14:39:21 in practice it doesn't work that way Sep 05 14:39:34 so that should be fixed and not worked around? :P Sep 05 14:39:55 technically it isn't a module anymore :) Sep 05 14:40:46 ahh so it should depend on virtual/kernel-feature-btsco or something? Sep 05 14:41:31 which both kernel-module-btsco and kernel with it compiled in should provide? Sep 05 14:43:07 it should RRECOMMEND = "kernel-module-btsco" in the current situation Sep 05 14:43:23 but yes, the nicest way would be the virtual/feature Sep 05 14:43:41 you can't actually compile it as part of the kernel at the moment Sep 05 14:43:43 so... Sep 05 14:43:55 but I don't think you'll find people motivated enough to fix OE for that scenario Sep 05 14:45:38 03koen 07org.oe.dev * r5b49a577... 10/packages/gaim/gaim_2.0.0-cvs-mco1.bb: gaim 2.0: add cvsdate in .bb Sep 05 14:45:44 it doesn't seem to be all that hard to fix - just have a virtul for each config option? Sep 05 14:47:00 and a script to parse the defconfig and/or system.map Sep 05 14:50:59 zecke: hah, so it seems. Sep 05 14:58:14 pb_: is he actually trying to make a point, or just want to pretend he knows better? Sep 05 15:01:41 so can we get opie-embeddedkonsole to be default on c7x0? the default console has font bugs out of the box and in general seems to be the less-desirable option. (at least on this platform) Sep 05 15:02:29 Disconnect: #opie Sep 05 15:02:34 koen|slug: doesn't seem to hard Sep 05 15:02:40 btw whats system.map? Sep 05 15:02:58 mithro: that should contain all symbols Sep 05 15:03:05 it could be a x86 thing though Sep 05 15:07:44 zecke: as far as I can tell, the latter. Sep 05 15:32:09 zecke, if your still awake i found the emacs problem ... has to do with a change in xemacs21 Sep 05 15:32:58 emte: I'm always interested in putting knowledge into my pool of solved problems Sep 05 15:34:01 yeah i started removing emacs packages Sep 05 15:34:16 issue was solved as soon as i removed xemacs21 Sep 05 15:34:43 emte: do you know what makes xemacs21 bad (yes I know it is emacs...) Sep 05 15:35:18 not specifically no Sep 05 15:35:35 it just for some reason hangs on the -batch command Sep 05 15:35:52 hangs/opens shell but doesnt seem to process Sep 05 15:52:52 03koen 07org.oe.dev * r56d53662... 10/packages/gnome/gnome-vfs_2.6.2.bb: packages/gnome/gnome-vfs_2.6.2.bb: version dependency on samba Sep 05 16:03:08 wow after three years of using irssi, I start to understand it Sep 05 16:03:44 lol Sep 05 16:07:45 wow Sep 05 16:08:03 my configured irssi even rocks more now! Sep 05 16:15:17 anyone alive out here ? Sep 05 16:15:48 just tried bitbake gpe-image ...ran all night, failed this morning Sep 05 16:16:02 fc-lang.c:(.text+0x2abf): undefined reference to `__ctype_b_loc' Sep 05 16:16:10 used gcc 3.4.3 Sep 05 16:16:27 does the same thing with gcc 4.0.1 Sep 05 16:16:28 OSS542: what does google say about this problem? Sep 05 16:16:53 just a sec Sep 05 16:17:43 http://lists.debian.org/debian-glibc/2003/06/msg00057.html <- my first hit Sep 05 16:19:03 zecke: good call, that puts it squarely in OE's lap ;) Sep 05 16:22:28 hm .....revert to 2.2.5 ? Sep 05 16:22:41 I think that is what I am using Sep 05 16:22:57 and this bug is 2 years old ? Sep 05 16:24:22 I should have thought to check google first Sep 05 16:24:41 but I couldn't imagine such a bug being around so long Sep 05 16:26:19 OSS542: why not use glibc 2.3.5? Sep 05 16:27:06 zecke: looks like I'll have to.....I was trying to avoid it, as I'll be building it from scratch Sep 05 16:27:18 hmm sf cvs still down? Sep 05 16:27:19 OSS542: what version do you build now? Sep 05 16:27:23 I don't care much for rpm you see... Sep 05 16:27:43 2.2.5 Sep 05 16:28:27 I think gcc 3.4.3 can build glibc 2.3.5, or so I'v heard, but I know 4.0.1 cannot....I've tried that.. Sep 05 16:28:49 several bugs the gcc people attribute to glibc 2.3.5 Sep 05 16:29:02 supposed to be fixed in glibc 2.3.6, which is not out yet Sep 05 16:29:09 OSS542: well... I wonder what you're doing... Sep 05 16:29:21 zecke: how so ? Sep 05 16:29:45 trying to make an OZ 3.5.4 for C3000 Sep 05 16:30:07 ...yes...I know it's not out yet...:-) Sep 05 16:31:23 not aout and you expect it to work? Sep 05 16:31:27 out* Sep 05 16:32:01 emte: it hasn't got that far yet.....and this is a glibc issue it seems Sep 05 16:32:31 besides....if no one tries it, how can we know what needs to be fixed ? Sep 05 16:32:54 it beats complaining when something is not out....perhaps I can be of some help instead of complaining about it. Sep 05 16:35:04 does it build on 3.5.3 ? Sep 05 16:35:46 i would think that one would have a static glibc Sep 05 16:37:45 ~botmail for OSS542 trying is good Sep 05 16:37:51 reminds me ... what familiar version did i set ... Sep 05 16:39:33 i suppose i'll run into his problem in a second if it is a host issue Sep 05 16:40:00 sort of i guess Sep 05 17:11:02 other than 'grep foo */*bb' is there any way to search the packages/ tree? Sep 05 19:22:07 evening Sep 05 19:22:14 hey Sep 05 19:22:35 kergoth: we killed onyxia :) Sep 05 19:23:23 nice Sep 05 20:10:12 found the undefined symbol problem I was asking about earlier Sep 05 20:10:42 seems my glibc 2.2.5 is just too old....doesn't define the symbols necessary Sep 05 20:11:01 I'll have to switch to glibc 2.3.x after all Sep 05 20:11:18 drats Sep 05 20:11:25 roarbazzle Sep 05 20:11:28 and all that Sep 05 20:12:03 and gcc 4.0.1 can't build glibc 2.3.x, which means I'll have to ditch that Sep 05 20:25:25 hi Sep 05 20:29:04 apparently no one alive out here Sep 05 20:29:22 hehe Sep 05 20:29:25 that's too bad Sep 05 22:09:31 :o Sep 06 00:11:23 Morning all! Sep 06 00:11:55 hi Sep 06 00:32:33 anyone testing with 2.6 kernel? Sep 06 00:48:02 re Sep 06 00:48:04 morning Sep 06 00:49:11 hi Sep 06 00:50:50 does everybody using gcc 2.95.3 for Xscale cross compile? Sep 06 00:52:59 we use 2.95.3 ONLY for zaurus 2.4.18/2.4.20 kernels Sep 06 00:53:03 ~zauruskernels Sep 06 00:53:04 hmm... zauruskernels is By default Zaurus use 2.4.18 or 2.4.20 and you need GCC 2.95.3 to compile them - all is described in OE wiki: http://openembedded.org/cgi-bin/moin.cgi/ZaurusKernels There is also work on 2.6: http://openembedded.org/cgi-bin/moin.cgi/ZaurusKernels26 Sep 06 00:53:36 hrw|work: thanks Sep 06 00:53:51 but I'm still wondering Sep 06 00:54:13 is anyother combination of gcc and kernel will work ? Sep 06 00:55:15 also ...which version of gcc you are using for build application (crosscompile) Sep 06 00:55:35 I have seen Xscale specific target in gcc 3.3 Sep 06 00:55:47 but gcc 3.4 back to generic Arm Sep 06 00:55:50 good morning all Sep 06 00:55:50 morning Sep 06 00:55:56 hi Sep 06 00:56:23 I'm confused and not sure where to bring up my info on gcc support on XScale Sep 06 00:56:44 does iwmmx instruction set being used in anyway? Sep 06 00:56:51 AvengerMoJo: see conf/machine/tune-xscale.conf to see what we use Sep 06 00:57:04 iwmmx isn't supported by the current toolchains Sep 06 00:57:08 koen|slug: thanks Sep 06 00:57:12 oh ... Sep 06 00:57:13 at least, not in a clean way Sep 06 00:57:53 will there be any effort of adding iwmmx into gcc? Sep 06 00:59:10 AvengerMoJo: take a look at the iwmmx thread in http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2005-July/thread.html Sep 06 01:00:13 koen|slug: thanks alot Sep 06 01:00:25 :) I think I will hang out here more offen :) Sep 06 01:03:22 hello, libsdl-x11 fails for me to compile Sep 06 01:03:24 http://rafb.net/paste/results/WsS1l339.html Sep 06 01:04:17 JiniDog: that doesn't list the actual error Sep 06 01:05:50 that's the funny thing, I don't see any Sep 06 01:06:23 further up in the log? Sep 06 01:06:32 03jbowler 07org.oe.dev * r8d3ed4f5... 10/conf/distro/ (ucslugc-packages.conf ucslugc.conf): ucslugc: Add upslug2-native to the UcSlugC build Sep 06 01:06:48 that's the full log: http://rafb.net/paste/results/i34oMV72.html Sep 06 01:06:58 nothing in there Sep 06 01:08:40 should wifi work on h5550 with the latest gpe snapshot? Sep 06 01:10:07 herren: it does over here Sep 06 01:10:32 do you have to do any modprobes, or did it just work out of the box? Sep 06 01:10:36 <_law_> http://www.rpsys.net/openzaurus/patches/pxa_pcmcia_init-r1.patch isnt available so i cant build a new kernel Sep 06 01:10:57 herren: out of the box, although I do have to start udhcpc by hand sometimes Sep 06 01:11:57 how do you do that? Sep 06 01:12:25 udhcpc -i wlan0 Sep 06 01:13:03 _law_: hmm.. RP changed something and forgot to update Sep 06 01:13:19 03koen 07org.oe.dev * rc0ba0c80... 10/packages/gnome/ (gnome-vfs-dbus_2.8.4.4.bb gnome-vfs_2.6.2.bb): gnome-vfs: fix typo in depends Sep 06 01:13:21 herren: see http://localhost.ruhr.de/~stefan/h5450/#nodhcp Sep 06 01:13:35 cool, thanks Sep 06 01:16:37 does mplayer really need libsdl? Sep 06 01:22:04 morning all Sep 06 01:22:52 JiniDog: should not Sep 06 01:23:00 JiniDog: yes Sep 06 01:23:28 JiniDog: well it doesnt technically need it, but its nice to offer that option Sep 06 01:24:22 I can live without, because libsdl-x11 fails Sep 06 01:24:46 JiniDog: works here Sep 06 01:24:52 I changed the configureline and took libsdl out of the DEPEND Sep 06 01:25:09 compiling mplayer fails with a segfault Sep 06 01:25:11 http://rafb.net/paste/results/slWkQk61.html Sep 06 01:25:33 JiniDog: your machine or compiler is b0rked Sep 06 01:26:02 this time it's all oe Sep 06 01:26:08 updated some days ago Sep 06 01:26:40 JiniDog: so its too old Sep 06 01:28:17 updating took hours Sep 06 01:28:34 why is it too old? did you bugfix that stuff? Sep 06 01:28:54 Success: 10372256 opie-image-collie-20050906074406.rootfs.jffs2 :) Sep 06 01:30:10 hrw|work: how big is the image? Sep 06 01:30:41 koen|slug: 10M as you see Sep 06 01:30:51 koen|slug: without padding Sep 06 01:31:01 tmp/rootfs is 23M Sep 06 01:32:04 nice Sep 06 01:32:10 koen|slug: .... still reading the iwmmx thread Sep 06 01:32:14 you removed ipsec and themes? Sep 06 01:32:20 koen|slug: start having a head ache ... Sep 06 01:32:42 AvengerMoJo: sound familiar Sep 06 01:32:45 koen|slug: get a question about the jff2 Sep 06 01:33:03 koen|slug: http://home.hrw.one.pl/tmp/ipk-status.txt Sep 06 01:33:07 koen|slug: is there any tool able to let me mount the jff2 on normal kernel system? Sep 06 01:35:47 AvengerMoJo: see http://handhelds.org/moin/moin.cgi/FamiliarBackupHowto?action=highlight&value=mtd on how to mount a .jffs2 on your desktop Sep 06 01:38:52 koen|slug: thank you sir Sep 06 01:45:52 koen|slug: 12M with openswan Sep 06 01:48:26 * AvengerMoJo still reading the thread .... Sep 06 02:01:39 mornin' Sep 06 02:01:43 hey AvengerMoJo :) Sep 06 02:04:22 hi Twiun Sep 06 02:04:26 Twiun: how are you doing Sep 06 02:04:42 long time no see Twiun ... didn't know you hang out here also :) Sep 06 02:05:03 Twiun: my project is getting into the system level... so I'm start working in this room now hehe Sep 06 02:05:31 koen|slug: ... I'm still reading the thread... Sep 06 02:05:57 :) Sep 06 02:06:13 koen|slug: is there an answer to the so call EABI and the old one ? if infact I'm going to use the new one ... where can I read up? Sep 06 02:06:57 glibc is still missing some EABI bits Sep 06 02:09:23 koen|slug: I see ... so me trying to make iwmmx support in glibc should be out of the pictures then? Sep 06 02:09:23 AvengerMoJo: Yeah, this is where I started just under 4 years ago ;) Sep 06 02:09:31 Twiun: hehe :P Sep 06 02:09:38 Twiun: that long already? Sep 06 02:09:49 Twiun: did you get a new pda? Sep 06 02:09:52 koen|slug: is there a replacement in the works for hotplug-ng ? Sep 06 02:09:55 AvengerMoJo: doesn't feel that long at all ;) Sep 06 02:10:05 AvengerMoJo: no idea, you'd have to talk to the glibc/arm people Sep 06 02:10:07 AvengerMoJo: no, couldn't justify the expense :( Sep 06 02:10:16 rwhitby: udev or hotplug Sep 06 02:10:20 Twiun: true :P time go fast ... Sep 06 02:11:25 Twiun: ... after my project .. may be I can send you one :P Sep 06 02:11:35 hehe :P Sep 06 02:12:01 koen|slug: I have heard ... the glibc is move arm out ... Where to ? Sep 06 02:12:03 AvengerMoJo: I'd be happy to relieve you of such a heavy burden :) Sep 06 02:12:15 AvengerMoJo: to the ports/ subdirectory Sep 06 02:12:49 koen|slug: so the arm ports is not completely stop right .. just being move out of the main tree? Sep 06 02:12:54 yes Sep 06 02:14:22 Twiun: hehe that may take longer then you want ... my project is only 2 months in the development...expecting 6 months long... so :P you may be waiting for a New Years present here hehe Sep 06 02:14:42 AvengerMoJo: It's allllllll good :) Sep 06 02:14:49 Twiun: :) Sep 06 02:15:00 _law_: Sorry, it got moved to archive accidently - its back now Sep 06 02:16:16 koen|slug: I just want to clear this out .... if Kernel and glibc not compile with iwmmx ... I can still use iwmmx instruction set in my application right? Sep 06 02:16:39 I've seen apps using it Sep 06 02:16:47 maybe they where statically linked Sep 06 02:17:32 hum Sep 06 02:18:46 I updated oe but mplayer still fails Sep 06 02:19:39 try rebuilding from scratch Sep 06 02:20:14 * XorA doesnt test with CSL gcc as thats known bad Sep 06 02:20:33 * gridzero tries desperately to stop his OE build from forkbombing.... Sep 06 02:22:02 when I rebuild all the packages from scratch, how do I make ipkg updating every package that is new? Sep 06 02:22:33 ipkg takes care of that of it's own Sep 06 02:22:47 morning Sep 06 02:24:26 wow Sep 06 02:24:38 it seems we have ~1000 revisions/month Sep 06 02:24:54 koen|slug: ~40-60 per day Sep 06 02:27:52 '9890 messages since the first one, 1.66 years ago, for an average of 1.47 hours between messages' according to cia Sep 06 02:28:45 rwhitby: is nslu2-linux also feeding cia? Sep 06 02:32:25 anyone had any trouble with edb-native-1.0.5.004-r1? looks like thats what I was compiling (well, configuring) when it forkbombed? Sep 06 02:34:42 koen|slug: "4689 messages since the first one, 0.99 years ago, for an average of 1.85 hours between messages" for nslu2-linux Sep 06 02:35:10 but of course each OE-related checkin is now reflected in both CIA project logs. Sep 06 02:35:22 rwhitby: it seems my stats get pimped by nslu2 and openembedded feeding the same csets :) Sep 06 02:39:25 nslu2 feeds the csets (all org.openembedded.dev csets) to the nslu2-linux cia project :-) **** ENDING LOGGING AT Tue Sep 06 02:59:57 2005