**** BEGIN LOGGING AT Tue Oct 25 02:59:58 2005 Oct 25 03:07:22 hi guys Oct 25 03:07:34 RP:when build the gpe-bootsplash have a few errors about dependency. Oct 25 03:09:16 RP:how to know their dependency one of all? Oct 25 03:10:27 RP: morning Oct 25 03:11:47 dunky: bitbake should build the dependencies (unless you're using -b?) Oct 25 03:11:50 katossi: hi Oct 25 03:11:58 is any dev reviewing reported bugs, besides the ones found by themselves? Oct 25 03:12:10 if not, somebody should assume that role, I think Oct 25 03:12:12 RP:yes ,I type bitbake -b.. Oct 25 03:12:49 hi Bernardo Oct 25 03:13:07 dunky: Do NOT use "-b" if you DON'T know what it is for Oct 25 03:13:09 otherwise the feedback from users that report bugs is wasted Oct 25 03:13:25 RP: if dont do that ,what shall I do about gpe-bootsplash Oct 25 03:14:06 dunky: just do bitbake gpe-bootsplash Oct 25 03:15:12 AvergerMojo:so,just simply,thanks Oct 25 03:15:18 hi Bernardo - in answer to your question last night, I'm not sure there is much to do - I have someone with a C1000 doing some tests - I'll let you know if I need any more help, thanks Oct 25 03:16:17 RP: ok, if you need extra tests I'll be here Oct 25 03:16:26 hi hrw|work Oct 25 03:16:50 RP: my main problem is that I can't get oe to build on my athlon64 (bug #33) Oct 25 03:16:56 RP: you know I'm here for tests too (unfortunately with no serial lead) Oct 25 03:17:09 Bernardo: I'm actually at a loss as to why my mtd driver still doesn't work on akita :-( Oct 25 03:17:25 katossi: you have a c3000 though? Oct 25 03:17:42 ups sorry, c1000 Oct 25 03:18:10 Bernardo: I'm no expert on 64 bit problems I'm afraid although it would be nice to get OE working around those issues... Oct 25 03:18:52 re Oct 25 03:19:26 it would really be nice, as it would let me contribute again... :) Oct 25 03:19:39 Bernardo: tried linux32 bitbake? Oct 25 03:19:41 back to class... Oct 25 03:19:53 zecke: I have a pure 64 bits machine Oct 25 03:20:11 I'll probably have to set a chrooted 32bits env Oct 25 03:21:04 ah, I had it running 32bits a few months ago, on monotone days, and it built Oct 25 03:25:36 Hello, could some one please give me a DNS server address which has already got the new info about oe.handhelds.org - thank you Oct 25 03:26:21 ~resolve oe.handhelds.org Oct 25 03:26:54 ~dns oe.handhelds.org Oct 25 03:26:57 I can't find oe.handhelds.org in DNS Oct 25 03:28:05 25 12:28 handhelds.org has address 128.31.0.25 Oct 25 03:28:16 that explains why can't I access oe.handhelds.org cvs... Oct 25 03:29:07 thank you, but when following links the dns still tries to resolve the address and fails Oct 25 03:29:40 oho: echo 'oe.handhelds.org 128.31.0.25'>>/etc/hosts Oct 25 03:32:24 <_law_> hey Oct 25 03:33:12 <_law_> hrw|work, have you seen my new script on bugzilla? Oct 25 03:33:32 hrw|work: thank you, but still no luck, my system seems to try the dns server first. I will keep on searching for a working DNS server Oct 25 03:33:37 _law_: nope Oct 25 03:34:18 zecke: stop using aleron.dl.sourceforge.net as SOURCEFORGE_MIRROR - OE switched to heanet as aleron lack too many files Oct 25 03:35:07 hrw|work: ;) did it Oct 25 03:35:30 zecke: this will lower fetch errors number Oct 25 03:36:15 zecke: and could you next time not wordwrap output of test Oct 25 03:36:16 ? Oct 25 03:37:57 hrw|work: sure that it was my mailer and not yours :( Oct 25 03:38:55 zecke: your - when I look into source I see wrapped. KMail does not wordwrap messages on read in ~7x column Oct 25 03:41:03 zecke: list reduced to 60 entries Oct 25 03:41:41 I will try my best next time Oct 25 03:41:55 actually I will 'post' a crappy HTML page Oct 25 03:43:32 what is ewi's full name again? Oct 25 03:45:43 alias mtpulle='mt pull ewi546.ewi.utwente.nl;mt merge;mt update' Oct 25 03:50:04 hrw|work: thanks Oct 25 03:50:30 np Oct 25 03:50:45 It was wrapped here as well Oct 25 03:53:20 03hrw 07org.oe.dev * r0d7e270c... 10/: asterisk: fixed SRC_URI Oct 25 03:55:42 hey reenoo Oct 25 03:55:46 and *sorry* Oct 25 03:56:59 hey zecke Oct 25 03:57:18 no worries. I just haven't got around to responding yet Oct 25 03:57:31 pretty busy with work and university stuff right now Oct 25 03:57:49 03hrw 07org.oe.dev * r6c74e180... 10/: diethotplug: fixed SRC_URI Oct 25 04:00:04 reenoo: same applies here, I totally agree that we should not upgrade to new versions (which I did in the past) Oct 25 04:00:21 reenoo: I see no issue switching the SRC_URI if we can verify that we still build the same source Oct 25 04:01:13 zecke: sure. if the sources have just been moved around that's fine Oct 25 04:01:16 I'm fixing SRC_URI for that pacakges where upstream moved files on server Oct 25 04:02:54 reenoo: and I have finally implemented checking the MD5SUM in the fetcher (with one minor issue) Oct 25 04:03:16 zecke: ah. excellent Oct 25 04:05:40 03hrw 07org.oe.dev * r8cef5b08... 10/: directfb: fixed SRC_URI Oct 25 04:05:44 03hrw 07org.oe.dev * r6c88ebf1... 10/: dfb++: fixed SRC_URI Oct 25 04:05:48 03hrw 07org.oe.dev * r5bb12b69... 10/: dnsmasq: fixed SRC_URI for 2.11 version Oct 25 04:06:32 reenoo: it shouldn't be too hard for someone fluent in python to adapt ipkg-upload and upload.cgi for OE Oct 25 04:08:06 file 4.13 need update or mirroring.. upstream keep only 2 releases, debian stable has 4.12, unstable 4.15, current is 4.16 Oct 25 04:08:44 file could also up a seperate libmagic package Oct 25 04:08:53 hrw|work: how about archive.debian.org? Oct 25 04:09:42 reenoo: not checked yet Oct 25 04:09:55 hrw|work: archive.debian.org Oct 25 04:10:08 reenoo: and debian switched from 4.12 to 4.15 directly Oct 25 04:10:46 zecke: perl 5.8.7 is fetchable so 5.8.4 too (server is responding without problem) Oct 25 04:10:52 NOTE: package perl-5.8.7-r14: task do_fetch: completed Oct 25 04:13:34 hrw|work: good, maybe ewi is simply crap ;) Oct 25 04:14:33 reenoo: hi, is there any problem with jamvm 1.3.3 and classpath 0.18 in oe? I mean any technical challenge Oct 25 04:14:38 lftp is not on archive.debian.org at all Oct 25 04:16:39 and is missing on snapshot.debian.net too (listed but 404) Oct 25 04:17:50 reenoo: or they're not there just because the are too new? Oct 25 04:18:56 lecture is starting. bbl Oct 25 04:19:16 snapshot.debian.net had disk failures.. archives lacking Oct 25 04:21:42 NOTE: package libpcl-1.4-r0: task do_fetch: completed Oct 25 04:26:54 03hrw 07org.oe.dev * rce0e128a... 10/: libxml: fixed SRC_URI for 2.6.7, 2.6.9 versions Oct 25 04:30:30 zecke: http://home.hrw.one.pl/oe/fetch-errors.txt Oct 25 04:31:15 03hrw 07org.oe.dev * rfda8cbc2... 10/: cyrus-imapd: fixed SRC_URI for 2.2.5 version Oct 25 04:31:19 03hrw 07org.oe.dev * rd9127ccc... 10/: cyrus-sasl: fixed SRC_URI for 2.1.17, 2.1.18 versions Oct 25 04:31:45 hrw|work: thanks Oct 25 04:32:44 hrw|work: some how bittest fails when getting to the letter 'l' Oct 25 04:32:55 hrw|work: I'm still investigating the cause of this Oct 25 04:33:05 too many errors until that so it gave up ;) Oct 25 04:33:30 I created it, so it never will give up :} Oct 25 04:33:39 even if it would be smart to give up Oct 25 04:34:25 ;) Oct 25 04:49:34 lol my mother used my ibook... Oct 25 04:50:09 my mother loves the zoom feature of the dock Oct 25 04:50:28 but she still doesn't get how the touchpad works Oct 25 04:58:26 I like comment in kernel config about USB_NET_ZAURUS Oct 25 04:59:03 If you install an alternate ROM image, such as the Linux 2.6 based versions of OpenZaurus, you should no longer need to support this protocol. Only the "eth-fd" or "net_fd" drivers in these devices really need this non-conformant variant of CDC Ethernet (or in some cases CDC MDLM) protocol, not "g_ether". Oct 25 04:59:33 ugh Oct 25 04:59:36 retards Oct 25 04:59:39 'ROM' Oct 25 05:00:01 koen: send a patch Oct 25 05:00:16 and dont forget to use word 'moron' ;) Oct 25 05:20:41 hrw|work: mail sent Oct 25 05:23:04 koen: hotplug-linux only RPROVIDES hotplug Oct 25 05:23:19 koen: I added PROVIDES = "hotplug" and now it works Oct 25 05:23:30 nokia770-init is build Oct 25 05:23:51 I'm not sure if that's the correct solution, but I don't know a better one Oct 25 05:24:09 RP or florian_kc would be the 770 OE experts Oct 25 05:24:22 what is the meaning of PROVIDES_${PN} ?? Oct 25 05:24:44 it sets the provides for the package Oct 25 05:25:21 so what is the difference between provides and provides_${pn} ? Oct 25 05:25:43 just two different notations Oct 25 05:26:07 i tried ..._${pn}, but this didn't work Oct 25 05:31:28 koen: do you get warned when ewi abuses too much bandwidth= Oct 25 05:32:28 also why does linux-hotplug (rprovides,rconflicts and rreplaces) hotplug - all at the same time ? Oct 25 05:32:44 I would not call it 'just'.. when recipe generate many packages then provides and provides_${pn} differ Oct 25 05:32:53 zecke: afaik there is no such thing as 'using too much bandwith' :) Oct 25 05:33:36 koen: I guess we will find out soon Oct 25 05:34:19 zecke: use ftp.nl.debian.org as debian mirror Oct 25 05:34:45 that's ~80 meters away from EWI Oct 25 05:34:57 and internal traffic doesn't get counted Oct 25 05:35:31 student connection have a FUP of 250GB/week Oct 25 05:35:43 connections* Oct 25 05:36:56 koen: I already use that mirror (hopefully) Oct 25 05:37:32 well only 11.5mb/s Oct 25 05:37:57 MB/s :) Oct 25 05:38:15 they didn't allow a gbit connection for ewi546 Oct 25 05:38:26 otherwise it would've been more :) Oct 25 05:38:36 morons ;) Oct 25 05:39:22 http://ftp.snt.utwente.nl/pub/linux/kernel/ is the kernel.org mirror Oct 25 05:39:30 same box as ftp.nl.debian.org Oct 25 05:39:44 koen: that is a GNU mirror as well= Oct 25 05:40:23 zecke: http://ftp.snt.utwente.nl/pub/linux/handhelds.org/ :) Oct 25 05:46:09 nice Oct 25 05:47:41 * zecke needs to get a disk for his shark(s) Oct 25 06:08:44 hi Cwiiis Oct 25 06:09:14 Hey hrw|work Oct 25 06:10:10 Cwiiis: long time no see - how things? Oct 25 06:11:34 hrw|work: Pretty good, been very busy with uni and working at OH - How about you? Oct 25 06:11:54 Cwiiis: quite good - wrote 1st application since years Oct 25 06:12:44 hrw|work: Cool :) What app? Oct 25 06:13:23 opie-notes (basing on NoteZ) Oct 25 06:13:25 hi kergoth Oct 25 06:14:55 hrw|work: Any screenshots about? Oct 25 06:15:19 Cwiiis: have to check.. Oct 25 06:15:48 hm. scap does not have any.. Oct 25 06:16:57 argh... Oct 25 06:17:00 ~lart scap on Z Oct 25 06:17:01 * ibot beats scap on Z over the head with a microkernel Oct 25 06:18:37 Cwiiis: http://home.hrw.one.pl/shots/notes.png Oct 25 06:29:33 ~hail Trolltech Oct 25 06:29:36 * ibot bows down to Trolltech and chants, "I'M NOT WORTHY!!" Oct 25 06:29:47 another two hours wasted why key repeats are not working... Oct 25 06:29:59 I guess Qtopia is a better selling product than M$ Word... Oct 25 06:30:18 hail not curse? Oct 25 06:37:16 hrw|work: it is at a point where it is funny and amusing Oct 25 06:37:24 * hrw|work tries to build qt-x11 2.3.1 with gcc4 Oct 25 06:37:52 zecke: keyrepeat is fluff Oct 25 06:39:08 morning Oct 25 06:39:11 hi chouimat Oct 25 06:39:21 koen: sure it is Oct 25 07:06:33 any czech around? Oct 25 07:19:09 zecke: it has to be czech? Oct 25 07:20:34 hrw|work: T-Mobile Czech starts a DVB-H trial Oct 25 07:20:44 o... nice Oct 25 07:21:04 and you have one of those 'not yet released' dvb-h devices? Oct 25 07:21:30 well I have a Opie based one... Oct 25 07:21:48 hrw|work: but that is from a research lab Oct 25 07:21:58 and I have seen the Nokia ones Oct 25 07:22:33 but in Czech Trial, Software I have written gets used, which makes me to some degree proud Oct 25 07:22:44 ;) Oct 25 07:23:06 * zecke wants a PXA with 800 Mhz now... Oct 25 07:23:36 zecke: not 1.25GHz? Oct 25 07:24:07 dpkg 1.13.4 is not available... snapshot.debian.org give 403, archive.debian.org does not have it Oct 25 07:25:27 hrw|work: I didn't knew they existed Oct 25 07:25:43 hrw|work: I would be happy with a dual, tripple, quad core SMP ARM Oct 25 07:26:07 zecke: Intel had 1.25GHz on some show Oct 25 07:26:25 a portalplayer SoC or an OMAP is much nicer Oct 25 07:28:30 koen: I agree on OMAP2 Oct 25 07:29:02 having a powerVR core integrated is pretty sweet Oct 25 07:30:19 well my other 'boss' was CTO at Convergence Oct 25 07:30:21 iirc you can get an omap2420 with a 330MHz dsp instead of the 220MHz one Oct 25 07:30:51 it is funny to see TI releasing stuff they had canned in 2000 and 2001 Oct 25 07:32:52 ~course TT for what they call multiplexer code Oct 25 07:32:57 ~curse TT for what they call multiplexer code Oct 25 07:32:59 May the fleas of a thousand camels infest your most sensitive regions, TT for what they call multiplexer code ! Oct 25 07:34:21 03hrw 07org.oe.dev * r58b66421... 10/: dash: get debian patch from snapshot.debian.net Oct 25 07:34:28 talking of TT, it's interesting that they'll be making a full Qt4 java binding available Oct 25 07:35:05 zecke: still no news about Qtopia4? Oct 25 07:35:29 hrw|work: sadly none Oct 25 07:35:41 1.25GHz? what is that Oct 25 07:36:04 hrw|work: I believe Qtopia4 will be closed source Oct 25 07:36:17 hrw|work: maybe one should embrace maemo.org Oct 25 07:36:19 because sourcecode is fluff Oct 25 07:36:57 maemo has a huge pile of close source too Oct 25 07:37:01 ~lart symbian Oct 25 07:37:02 * ibot tries to shut symbian up Oct 25 07:37:15 let someone move gumstix support into nonworking.... Oct 25 07:38:00 koen: well we can replace these bits Oct 25 07:38:00 03hrw 07org.oe.dev * rdcbbbac5... 10/: enscript: switched SRC_URI to Debian mirrors Oct 25 07:39:23 let us hope minix does not run on ARM next term... Oct 25 07:39:31 zecke: updated fetch error list Oct 25 07:40:41 hrw|work: still on http://home.hrw.one.pl/oe/fetch-errors.txt? Oct 25 07:40:42 koen: gtk-industrial-engine older then 0.2.4x can be removed? Oct 25 07:40:45 zecke: yes Oct 25 07:42:05 lol my mother uses my ibook again... Oct 25 07:42:07 koen: even Debian 'sarge' have newer versions... :) Oct 25 07:42:24 hrw|work: let me guess, debian sarge is fluff Oct 25 07:42:36 anyway! Oct 25 07:42:50 zecke: ofcourse that it is fluff Oct 25 07:44:12 hrw|work: yes, industrial is in gtk-engines nowadays Oct 25 07:44:58 koen: so could you remove non-needed ones? I dont know which one are they Oct 25 07:51:53 koen: 266.53 B/s wow ewi is fast Oct 25 07:56:17 koen: could you please email me the config files you are using to build packages for the h2200, i.e. bitbake.conf and local.conf (can_man@gmx.de) thank you Oct 25 08:03:18 zecke: check fetch errors Oct 25 08:03:28 zecke: descriptions added Oct 25 08:03:52 zecke: other things from list are fetchable Oct 25 08:04:08 03hrw 07org.oe.dev * r24ea45d9... 10/: libelf: fixed SRC_URI to new location Oct 25 08:15:15 could some one please comment on the reply I got from my sys admin about the handhelds.org DNS server http://pastebin.ca/26612 Oct 25 08:18:19 I could not Oct 25 08:20:26 france: ping!? Oct 25 08:20:43 france: is that http://pastebin.ca/26612 an issue? Oct 25 08:33:43 hey guys, is ther a reason i cant get onto www.openembedded.org? Oct 25 08:33:52 i think i need to update the dns Oct 25 08:33:57 you should be able to get to oe.handhelds.org Oct 25 08:35:01 nope, "oe.handhelds.org could not be found. Please check the name and try again later" Oct 25 08:35:06 odd Oct 25 08:35:10 yup Oct 25 08:35:18 i can get onto handhelds.org Oct 25 08:35:27 and bugs.openembedded.org Oct 25 08:35:29 guess we'll have to pester the hh.o folks. the oe subdomain should be working afaik Oct 25 08:36:02 how do i go about pestering them? Oct 25 08:37:46 france: ping Oct 25 08:39:56 mail to admin@hh.org Oct 25 08:45:24 cu Oct 25 09:07:21 wtf Oct 25 09:07:22 EXT2-fs: loop0: couldn't mount because of unsupported optional features (4). Oct 25 09:46:21 03noodles 07org.oe.dev * r8c6044d6... 10/: radvd: Add build time dependancy on flex. Oct 25 09:53:26 hmmm Oct 25 10:34:48 I'm trying to get handhelds-pxa-2.6_cvs.bb to use a different defconfig for h2200, so I added do_configure_h2200() alongside do_configure(). But when I try to build, it begins with do_compile rather than do_fetch, do_unpack, etc. and of course it fails. What am I doing wrong? Oct 25 10:36:02 after a -c clean? Oct 25 10:36:13 you could try bitbake -b file.bb -c configure -f Oct 25 10:36:29 no, I hadn't tried build it yet. I did try rebuild, and removing any handhelds* stamps Oct 25 10:37:01 is -c the same as 'clean' in bitbake -i? Oct 25 10:37:41 -c tells it what command to run. Oct 25 10:37:49 rtfm Oct 25 10:37:50 if I comment out the do_configure_h2200() and 'reparse' it goes through the proper steps, but when I uncomment those lines it fails Oct 25 10:37:53 thanks kergoth Oct 25 10:39:00 no errors when I do '-c configure -f' Oct 25 10:39:21 but it still doesn't build Oct 25 10:42:12 could you pastebin the .bb? Oct 25 10:43:55 http://pastebin.com/405492 Oct 25 10:44:47 hi Oct 25 10:44:58 hey zecke Oct 25 10:46:36 mreimer: that look ok to me Oct 25 10:46:53 koen|tv: yeah, it seems simple enough Oct 25 10:47:29 it looks like the example in http://bitbake.berlios.de/manual/ch02.html#id2465968 Oct 25 10:48:34 do I need to specify OVERRIDES? Oct 25 10:49:09 $MACHINE should always be in overrides Oct 25 10:49:20 seems like a bug somewhere Oct 25 12:08:56 hey people Oct 25 12:08:58 anyone alive? Oct 25 12:09:14 * koen makes zombie sounds Oct 25 12:09:28 braaaaains Oct 25 12:09:41 afternoon kergoth Oct 25 12:10:20 * mithro feels like a zombie Oct 25 12:10:41 hey Oct 25 12:38:27 oe.openembedded.org has a corrupted dns entry... Oct 25 12:41:45 "corrupted" in what way? Oct 25 12:42:02 oh, duh, he left Oct 25 12:49:28 hi Oct 25 12:49:42 hi hrw Oct 25 12:50:26 someone have Zaurus with 2.6.14-rc4 and hostap wifi? Oct 25 12:54:03 LoDown: never used irc befor, so did not know about this simple rules :) **** BEGIN LOGGING AT Tue Oct 25 12:55:50 2005 Oct 25 12:56:04 hi all Oct 25 12:56:25 DaniLoveSky, i was not aware of that problem...so I can't say why Oct 25 12:57:04 DaniLoveSky, my OE installation doesn't have any symlinks...so I never met that problem... Oct 25 12:57:11 LoDown: It took me a while to figure out that even my /tmp dir (where I put the compilation results) must be a proper directory, not a link to /var/tmp. Oct 25 12:57:29 :-) Oct 25 12:58:08 use /tmp is weird.. Oct 25 12:58:25 LoDown: WiKi does not state it clearly - it says that the OE directory should satisfy this rule, but tmp dir - it is something different, right? Oct 25 12:58:28 i use ~/devel/build/SUBDIR Oct 25 12:58:42 tmpdir is part of OE Oct 25 12:59:09 hrw: I was forced to use /tmp as I only have 2Gb free space in my home directory where I put OE. Oct 25 12:59:35 and so much in /tmp? nice config Oct 25 13:00:20 hrw: Yes, now I understand. After 3 tryes to install OE. Would be nice to highlite it in the wiki... Oct 25 13:00:49 propose changes to GettingStarted on mailing list Oct 25 13:02:10 hrw: burning DVDs requires this tmp and also compiling differnt stuf, like kde and open office Oct 25 13:02:54 if you say so.. Oct 25 13:03:13 i used any dir as tmp for writing cd/dvd Oct 25 13:03:44 but thats config of machine.. Oct 25 13:04:16 hrw: this is just a matter of habbits Oct 25 13:04:22 exactly Oct 25 13:05:18 anyways, what is the reason (in simple words) for OE not to like sym links? Oct 25 13:06:12 i dont remember Oct 25 13:07:20 it's a bug in autoconf Oct 25 13:07:24 03jbowler 07org.oe.dev * r83e8fa36... 10/: ixp4xx: (comments only) fix bad version number in comment for 2.0 Oct 25 13:07:34 I'm shocked, shocked! Oct 25 13:09:48 pb_: thanks. Oct 25 13:11:01 pb_: actually, autoconf may be borked, but oe really shouldnt be canonicalizing that path Oct 25 13:11:11 theres no reason we cant be using the path that has the symlinks Oct 25 13:12:03 Another question is about building familiar_gpe image using OE. The problem is that in order to compile kernel-2.4.* a gcc <=3.3.3 is required (correct me if I am wrong). Oct 25 13:13:30 OE tryes to compile gcc-3.3.3 using the compiler installed on the host. In my case it is gcc-4.0.1. And fails as gcc-4.0.* cannot complie <=gcc-3.5.* (is it right)? Oct 25 13:14:16 kergoth: iirc, it's autoconf (or rather, the generated configure script) that canonicalises it. I don't think oe is doing anything wrong. Oct 25 13:14:33 cu Oct 25 13:14:39 if test "`cd $srcdir && pwd`" != "`pwd`" && Oct 25 13:14:44 that's the bit that goes wrong Oct 25 13:17:55 I pointed OE to gcc-3.2 insted (gcc32). OE successfully compiled gcc-3.3.3 and stuck on kernel compilation again, exactly the same point as when I used gcc-3.3.4 to compile the kernel. (build/tmp/work/orinoco-modules-0.15rc2-r9/orinoco-0.15rc2/spectrum_cs.c:588) Oct 25 13:19:17 I made mino corrections to the code of this module and it compiled fine. Now I wonder, would the gcc-3.3.4 compile the cernel with these corrections or it would have some major problems later on? Oct 25 13:19:18 That's not actually the kernel. Oct 25 13:19:28 pb_: ahh Oct 25 13:20:07 03rpurdie 07org.oe.dev * rf4799c65... 10/: (log message trimmed) Oct 25 13:20:08 linux-oz-2.6: Various updates to dev kernel: Oct 25 13:20:08 * Add various compile fixes passed to mainline Oct 25 13:20:08 * Update pm code to use pxa_pm_ops as discussed on LAKML Oct 25 13:20:08 * Add ALSA SoC subsystem patches from Liam Girdwood Oct 25 13:20:08 * Add ALSA wm8731 codec driver (as found in c7x0) Oct 25 13:20:10 * Update c7x0 defconfig to make OSS drivers modular and enable ALSA drivers (output only for now) Oct 25 13:20:35 kergoth: it's trying to figure out whether you are building outside the source directory. But, in the presence of a symlink, it (wrongly) decides that $(srcdir) is different from the pwd, and blows up. Oct 25 13:21:06 of course, oe could avoid that problem by always building in a directory outside $(srcdir), which might be good form anyway. Oct 25 13:21:13 * kergoth nods Oct 25 13:21:58 DaniLoveSky: I have a feeling that orinoco 0.15 doesn't work anyway, even if you can compile it. But you should check with mickeyl; he is the orinocomeister. Oct 25 13:22:59 I'm almost positive the word orinocomeister needs some umlats in it Oct 25 13:23:40 pb_, DaniLoveSky : Orinoco 0.15 works but only with 2.6 kernels Oct 25 13:23:50 For 2.4 kernels use 0.13e Oct 25 13:24:32 pb_: Putting orinoco aside, do you think kernel would fail to compile with gcc-3.3.4? Oct 25 13:24:48 DaniLoveSky: I think it should be okay. Oct 25 13:25:02 and, if it has gotten as far as trying to build orinoco, I think it must already have succeeded in building the kernel. Oct 25 13:26:16 RP: do you know how well iPAQh5500 is supported by kernel 2.6? Oct 25 13:26:56 DaniLoveSky: No idea. try #handhelds.org or the handhelds.org website/wiki Oct 25 13:27:47 heh Oct 25 13:27:54 RP: did you see Pavel's little patch on the LKML? :) Oct 25 13:28:16 CosmicPenguin: I was in the cc ;-) Oct 25 13:28:23 it amused me :) Oct 25 13:28:26 oh - I was reading on the gmane - they don't do ccs Oct 25 13:29:42 pb_: I will try to put the things back tomorrow and try to compile everything using gcc-4.0.1 and changing .bb files to make OE compile gcc-3.3.4 insted gcc-3.3.3 (which fails on systems with gcc-4.0.* , e.g. FC4). Oct 25 13:29:54 okay, sounds good Oct 25 13:31:00 does anyone have any ideas why adding a do_configure_h2200() method to a bb file (in this case handhelds-pxa-2.6_cvs.bb) would cause bitbake to not do the fetch, unpack, etc. stages and skip right to do_compile? Oct 25 13:31:32 pb_: For now the easiest solution for users of gcc-4.0 is to make a symlink: ln -s /usr/bin/gcc32 /path_ti_oe/bitbake/bin/gcc Oct 25 13:32:51 tpb_: this will force OE to use gcc-3.2 which works fine for compiling familiar v0.8.3 Oct 25 13:33:56 I wish all this info about kernel/gcc/oe compatability was in wiki. It would save me a couple of days :) Oct 25 13:35:30 mreimer: you could try mailing the list Oct 25 13:35:38 thanks koen Oct 25 13:47:14 DaniLoveSky, H5555 works with familiar, so I would assume H5500 would as well, as they are almost identical... Oct 25 13:48:41 LoDown: Yes, I know. I am having familiar on h5500 for almost 2 years now. All the versions I have seen use kernel 2.4. But I am not sure if kernel 2.6 is working there. Oct 25 13:50:18 DaniLoveSky: uE just got SD working with 2.6.13 Oct 25 13:50:57 koen|tv: good news :) Oct 25 13:51:32 koen|tv: what are the major features missing from kernel 2.6 now? Oct 25 13:51:50 see the wiki Oct 25 13:52:12 http://handhelds.org/moin/moin.cgi/HpIpaqH5400 Oct 25 13:52:37 koen|tv: thanks! Oct 25 14:04:20 good night all Oct 25 16:06:35 * france is back (gone 14:58:41) Oct 25 16:07:00 hi Oct 25 16:07:15 uE said there is a problem with the web stie? Is it work now? Oct 25 16:16:11 france, yes Oct 25 16:16:48 oris_wolfbane: thx, let me know if there are any other issues from the moving of the servers. Oct 25 16:16:49 good night Oct 25 16:16:58 florian_kc: your leaving...?? Oct 25 16:17:04 florian_kc: nite! Oct 25 16:17:45 france: yeah.. almost 1:18am here... Oct 25 16:17:55 florian_kc: heh Oct 25 16:18:08 florian_kc: I will catch you in your morning. I should still be on-line. Oct 25 16:19:01 france: okay - thaks - worst case there is alwas email :-) cu Oct 25 16:24:19 anyone had orinoco .15rc2 fail to compile? Oct 25 16:24:36 why is gpe-image even trying to make this? Oct 25 16:25:53 and the room is quiet... :-D Oct 25 16:40:06 LoDown: I had. Oct 25 16:42:05 LoDown: I do not know why gpe-image target depends on orinoco (may be it is not necessary at all?) Oct 25 17:37:23 heya joshua Oct 25 17:39:51 * Twiun yawns Oct 25 18:36:45 night all Oct 25 18:58:20 kergoth: poke and 1/2 Oct 25 19:07:35 morning Oct 25 19:07:51 hardwire: whats up Oct 25 19:19:02 hardwire is back? Oct 25 21:40:50 kergoth: I am doing some sick shit Oct 25 21:40:56 yay Oct 25 21:41:08 http iptables matching Oct 25 21:41:11 for http audio streams Oct 25 21:41:18 and having it go out over a crappy cable line vs the dsl Oct 25 21:41:29 haha, sick Oct 25 21:41:29 any port Oct 25 21:41:30 anywhere Oct 25 21:41:32 any time Oct 25 21:41:42 using l7-filter Oct 25 21:41:49 that reminds me of that match that can match any string in a packet Oct 25 21:41:53 I was going to hack iptables-p2p Oct 25 21:41:54 whats l7-filter? Oct 25 21:42:04 kergoth: http://l7-filter.sourceforge.net/ Oct 25 21:42:07 finish what you are doing first Oct 25 21:42:12 because what this can do is just too tempting Oct 25 21:42:21 they really really want it to head into patch-o-matic Oct 25 21:42:21 oooh Oct 25 21:42:43 it basically matches the application data.. the first MTU or lower sized packet of a new connection Oct 25 21:42:45 and regexps it Oct 25 21:43:01 I am going to use it to fwmark Oct 25 21:43:19 as well as relocate anything that can be transparently proxied.. into the right places. Oct 25 21:43:19 "In other words, l7-filter is essentially what you'd get if you added the CONNMARK patch to the string patch, used regexec() instead of strstr(), removed the need to set up the connection tracking manually, added the ability to match patterns across several packets and included a collection of protocol signatures that could be identified by name. :-)" Oct 25 21:43:21 damn Oct 25 21:43:26 yeh Oct 25 21:43:30 so I am patching away Oct 25 21:43:35 and adding a few other patches Oct 25 21:43:45 I have dual Via c3 1.2ghz firewalls w/ 6 10/100 in them at work Oct 25 21:43:50 and I am going to make them redundant and happy Oct 25 21:43:58 and vinger out VRRP(sp)? Oct 25 21:44:04 vinger/finger Oct 25 21:44:07 but first.. I must pee Oct 25 21:45:29 MUCHO GRANDE! Oct 25 21:45:53 hmm Oct 25 21:45:54 Q: How can I use l7-filter to redirect some of my traffic (to a proxy, etc)? Oct 25 21:45:55 A: You pretty much can't, at least not in any straightforward way. L7-filter can't possibly identifiy what protocol a connection is using until it sees a packet with data in it. For TCP, this is the third packet, far too late to start redirecting anything. Oct 25 21:46:11 yeh Oct 25 21:46:15 thats one worry I had Oct 25 21:46:32 I figured I could encourage packet spoofing to correct that Oct 25 21:46:45 if match.. log.. hold.. spoof.. replay Oct 25 21:47:06 hmmm Oct 25 21:47:07 because if you catch it at prerouting.. it hasn't been sent yet correct? Oct 25 21:47:43 well, yeah, but some of the packets will have been sent by the time it identifies it Oct 25 21:47:57 can you explain that? Oct 25 21:48:18 because I thought it matched the first packet.. not the first few. Oct 25 21:48:24 syn -> prerouting -> routing tables -> postrouting -> whee Oct 25 21:48:27 etc Oct 25 21:48:31 theres negotiation before the data. Oct 25 21:48:39 it doesnt get any real application data until a few packets in Oct 25 21:48:42 ah Oct 25 21:48:43 thats what that faq entry is about Oct 25 21:48:50 so port 80 redir is before the syn. Oct 25 21:49:01 you'd have to either hold up those packets until you find out, or.. proxy it or something Oct 25 21:49:19 i was thinking you could offload traffic into userspace via netfilter and proxy that out Oct 25 21:49:20 hmm Oct 25 21:49:24 * kergoth ponders Oct 25 21:49:54 hello all. Oct 25 21:50:06 Tried visiting the #familiar channel but no one was talking. Oct 25 21:50:21 rarely is. for familiar help, try #handhelds.org Oct 25 21:50:21 kergoth: k->u->k->net Oct 25 21:50:25 sounds really really poopy Oct 25 21:50:36 A quick/simple question: what parameters do I specify in "local.conf" to compile Familiar 0.8.2? Oct 25 21:50:47 hardwire: yeah.. i dunno how else you could do it though.. hmm Oct 25 21:51:01 kergoth: hah.. just redirect all ports with a fwmark and log. Oct 25 21:51:05 random fwmark :) Oct 25 21:51:06 cyberchuck: most likely, DISTRO="familiar-0.8.2", but i havent touched oe in months Oct 25 21:51:08 hah Oct 25 21:51:11 hmm Oct 25 21:51:15 and then something to catch the log Oct 25 21:51:28 hmm Oct 25 21:51:33 and notify the daemon that is listening on the redirected port Oct 25 21:51:37 and it dispatches Oct 25 21:51:47 There is no "familiar-0.8.2.conf" file in the "distros" subdirectory. Oct 25 21:51:48 so not transparent happy. Oct 25 21:52:20 There IS a "familiar-0.8.3", but that is so unstable ... we compiled that but the image did some REALLY strange things to the display. It is definitely bleeding edge. Oct 25 21:52:36 hardwire: what protocol are you doing this on again, http? Oct 25 21:53:42 indeed Oct 25 21:54:35 i wonder if you could fake it out. Oct 25 21:54:36 I could just have it get thrown to something that sends a relocate Oct 25 21:54:39 oh Oct 25 21:54:40 wow Oct 25 21:54:42 its hi 5 time Oct 25 21:54:42 pretend to be the other guy at first Oct 25 21:54:45 identify it Oct 25 21:54:50 hey Oct 25 21:54:53 then start a negotation with the destination.. Oct 25 21:54:58 * hardwire hi-5 kergoth Oct 25 21:55:00 basically a man-in-the-middle, actually Oct 25 21:55:04 hehe Oct 25 21:55:15 hmm Oct 25 21:55:22 well.. I wonder how much data I can catch Oct 25 21:55:30 itd be hard to get the sequence numbering happy, probably Oct 25 21:55:31 hmm Oct 25 21:55:33 since I can redir.. I will basically redir it to a netcat listening Oct 25 21:55:39 and see what pops up Oct 25 21:55:42 * kergoth ponders Oct 25 21:55:46 if its enough to send a new location header.. Oct 25 21:55:50 then okay Oct 25 21:55:57 this woudl be a lot easier if you were proxying. Oct 25 21:56:00 squid, anyone? Oct 25 21:56:02 ;) Oct 25 21:56:10 mmmmh Oct 25 21:56:13 dont make me hungry Oct 25 21:56:14 kergoth: think about the fundamental problem I am having :) Oct 25 21:56:17 hehe Oct 25 21:56:20 port != 80 Oct 25 21:56:29 ah, right Oct 25 21:56:30 damnit Oct 25 21:56:37 YEH Oct 25 21:56:38 damnit Oct 25 21:56:48 well Oct 25 21:56:51 think about it this way.. Oct 25 21:57:13 I can add dynamic iptables firewall rules Oct 25 21:57:18 based on log matching :) Oct 25 21:57:26 to make sure it does what I want. Oct 25 21:57:32 they will just have to try it twice :) Oct 25 21:57:34 within an hour Oct 25 21:57:49 i was thinking about that too. store the knowledge for later Oct 25 21:58:00 ugly no matter how you implement :\ Oct 25 21:58:03 You can alter the number of packets at any time through /proc/net/layer7_numpackets. (i.e. "echo 16 > /proc/net/layer7_numpackets".) Oct 25 21:58:04 hah Oct 25 21:58:05 wow Oct 25 21:58:11 change that to like.. the max size of an http header Oct 25 21:58:16 oh Oct 25 21:58:17 right Oct 25 21:58:18 tcp syn Oct 25 21:58:31 well, the tcp session needs to get going to get the data out Oct 25 21:58:37 successful negotiation. it needs replies Oct 25 21:58:42 yeh Oct 25 21:58:45 cant just buffer inbound packets and hold it Oct 25 21:58:46 :\ Oct 25 21:58:47 hmm Oct 25 21:58:50 hmm. Oct 25 21:58:56 layer 7 anyone? Oct 25 21:59:06 we should move this to #iptables -> Oct 25 21:59:07 c'mon help a brother out. Oct 25 21:59:07 :P Oct 25 21:59:12 nobody in there Oct 25 21:59:28 I wanted to put a bee in some bonets.. Oct 25 21:59:28 there's like 60 people in there Oct 25 21:59:32 freenode? Oct 25 21:59:33 but yeah, mostly idle Oct 25 21:59:34 yep Oct 25 21:59:39 i'm in both #iptables and #netfilter Oct 25 21:59:39 I must have typoed Oct 25 21:59:45 the latter is useful for talking to the actual netfilter devs Oct 25 21:59:47 more than setup questions Oct 25 22:00:17 ipp2p? Oct 25 22:05:32 hardwire: its gratifying that everything he's suggested we've already mused on Oct 25 22:05:33 :P Oct 25 22:15:51 http://protocolinfo.org/wiki/Main_Page looks useful Oct 25 22:23:34 kergoth: well Oct 25 22:23:38 he doesn't seem neat. Oct 25 22:24:01 checking out ipp2p as well now Oct 25 22:24:02 damnit Oct 25 22:24:51 ok.. atleast I can proxy udp if the match is quick Oct 25 22:25:08 nod. no session to worry about Oct 25 22:25:16 newp Oct 25 22:25:39 but what the fuck am I supposed to do? Oct 25 22:26:02 http request made.. stream started. Oct 25 22:26:11 scr and dst are locked. Oct 25 22:26:15 src Oct 25 22:26:18 I can't just change src Oct 25 22:26:31 after the request is sent.. waiting for a reply :) Oct 25 22:26:49 dig on that Oct 25 22:26:52 I gotta go be cute Oct 25 22:27:18 hm Oct 25 22:28:09 she is going to want to be watching the dog. Oct 25 22:28:13 watching/walking Oct 25 22:28:19 I should start dinner before she walks in the door Oct 25 22:28:21 get her fat Oct 25 22:28:26 and in need of digesting Oct 25 22:28:46 therefore filling her with guilt of not walking the puppy Oct 25 22:28:53 but at the same time.. I get to relax Oct 25 22:29:05 its not relaxing walking a dog in 10 degrees Oct 25 22:29:09 in the dark Oct 25 22:29:17 looking for moose eyes w/ your flashlight. Oct 25 22:29:32 hah Oct 25 22:29:39 reminds me of invader zim. Oct 25 22:29:44 yeh. Oct 25 22:29:46 a room with a MOOSE! Oct 25 22:29:47 no GCT. :( Oct 25 22:29:49 :( Oct 25 22:30:16 heh. Oct 25 22:30:18 thats not true Oct 25 22:30:22 my palm pilot has a gps Oct 25 22:30:34 .. "yup.. were wakling at 2.1 mph" woot. Oct 25 22:30:41 hmm Oct 25 22:30:49 pasta + pest + alfredo + thick ass noodles Oct 25 22:30:52 sousdn good to me Oct 25 22:31:22 mm Oct 25 22:35:16 yay Oct 25 22:35:17 no walk Oct 25 22:35:22 I convinced her I need to be fatter. Oct 25 22:35:52 http://bogomip.com/~hardwire/mt_yodel_2.jpg Oct 25 22:35:56 our friend is fostering a kitty Oct 25 22:36:00 while the owner is away Oct 25 22:36:05 it took a shine to the dogs. Oct 25 22:36:25 g/f is being loud Oct 25 23:41:44 wow. its hardwire Oct 25 23:46:24 hi Oct 25 23:47:10 yesterday someone talk about the 1.25GHz Xscale is that some new hardware out there already? Oct 25 23:51:01 ljp: wow its you! Oct 25 23:51:04 and off to bed I go Oct 25 23:51:23 kergoth: check out wrr and wipl(d,clients) Oct 25 23:51:25 bbl Oct 25 23:53:04 hey lhp Oct 25 23:53:07 ljp Oct 25 23:53:13 How's california treating you? Oct 25 23:56:07 was only in cali for a week. back in au :) Oct 25 23:56:34 it was nice. had a2005 mustang convert to drive :D Oct 25 23:57:12 ah Oct 25 23:57:13 fun Oct 25 23:57:43 mine isn't a convertible, but it's still fun :) Oct 25 23:57:56 ya, wish they had them here Oct 25 23:58:21 good to be home? Oct 25 23:58:52 ya. but going to munich next week Oct 25 23:58:57 :( Oct 25 23:59:38 need to pack warm clothing, I here Oct 25 23:59:43 hear Oct 26 00:06:37 hi gang, can someone have a look at bug 399, its a bit of show stopper for onboard prism2 wireless at the moment. Oct 26 00:06:44 The erroe is <3>skb_p80211_to_ether: DIXII frame too large (1504 > 1500) Oct 26 00:07:37 Or does someone know of a work around? Oct 26 00:27:17 morning all Oct 26 00:35:54 morning RP Oct 26 00:42:58 heya Oct 26 00:43:03 morning Oct 26 01:07:07 morning Oct 26 01:19:31 morning Oct 26 01:19:45 ops.. frgot to nick change Oct 26 01:20:18 we just assume you work hard :-D Oct 26 01:22:29 The c7x0 dev kernel has preliminary alsa support now btw ;-) Oct 26 01:22:44 Power management isn't good yet but will get there Oct 26 01:24:06 RP: I'm going to build it Oct 26 01:24:40 and I have to fix opie-notes Oct 26 01:25:35 At the moment both the speaker and headphones are enabled at once - I'm working Liam to enable machine specific controls Oct 26 01:26:59 hi alan Oct 26 01:35:37 RP: cool Oct 26 01:41:37 hello all Oct 26 01:45:42 goog morning all Oct 26 01:46:31 morning Oct 26 01:47:51 hi mickeyl Oct 26 01:48:23 is something wrong with the web? I can't get it? dns problem or server down? Oct 26 01:48:37 "the web" ? Oct 26 01:48:50 openembedded.org Oct 26 01:48:55 ah. yes, dns problems. Oct 26 01:48:58 use oe.handhelds.org Oct 26 01:49:01 ok Oct 26 01:49:21 thx, it works Oct 26 01:50:19 mickeyl: how can I use my own bb files in a different directory than the standard location Oct 26 01:50:46 I don't want them to be deleted when I do a update Oct 26 01:50:54 morning mickeyl Oct 26 01:51:03 katossi: you can try using BBCOLLECTIONS Oct 26 01:51:10 hey RP, how's kernel hacking going? Oct 26 01:51:13 aha Oct 26 01:51:15 * mickeyl out of the loop Oct 26 01:51:38 mickeyl: c7x0 has a very alpha alsa driver Oct 26 01:51:52 katossi: never used them myself, but that's how you should have heaps of additional bbfiles that "override" the standard ones. Oct 26 01:52:02 katossi: try asking pb_ or kergoth, they used this feature Oct 26 01:52:05 mickeyl: I also have an akita in front of me on loan. I will get mtd working... :) Oct 26 01:52:08 RP: woot. that's good news Oct 26 01:52:14 RP: ah excellent as well! Oct 26 01:52:19 * mickeyl owner of 770 now Oct 26 01:52:20 katossi: BBFILES = "oneplace/*/*.bb oe/*/*.bb" Oct 26 01:52:39 cxx00 alsa driver is also being worked on... Oct 26 01:52:40 BB>> getvar BBFILES Oct 26 01:52:40 /home/hrw/devel/oe/org.openembedded/packages/*/*.bb /home/hrw/devel/oe/hrw-packages/*/*.bb Oct 26 01:52:46 RP: any news about the nand in the akita? Oct 26 01:53:13 katossi: See above - I have an akita in front of me and am debugging it as we speak... Oct 26 01:53:22 hrw|work: thx, I'll try Oct 26 01:53:49 ~seen CoreDump|home Oct 26 01:53:55 mickeyl: How do you find the 770? Oct 26 01:53:55 coredump|home is currently on #openzaurus #gpe #opie #handhelds.org #oe Oct 26 01:54:53 RP: mixed emotions. Great hardware (except the braindead desk stand and the CPU that could need some boosts), the UI look and feel is very good. Oct 26 01:55:06 for now it just lacks more apps and stability. Oct 26 01:55:41 Pretty much how I feel :). I also miss a keyboard with it... Oct 26 01:55:46 i'm impressed though how connectivity is working. very nice job, i know how hard that is to shield the user from the ugly details Oct 26 01:55:53 ya, a keyboard would be nice Oct 26 01:56:47 connectivity.. I have to go with Z into some wifi zones to check does it has problem with wifi (hostap 0.3.9 or 0.4.5) or with my wrt54gs Oct 26 01:57:35 maemo could success where qtopia failed Oct 26 01:57:43 in providing a common environment Oct 26 01:57:53 that developers like to work with Oct 26 01:57:59 s/common/common and open/ Oct 26 01:58:11 It has a lot going for it... Oct 26 01:59:26 I have an aversion to scatchbox though ;-) Oct 26 01:59:30 heh Oct 26 01:59:33 same here Oct 26 01:59:46 can't wait to build a complete image for it with OE Oct 26 01:59:58 OE can build valid images for it... Oct 26 02:00:20 cool. how many things are missing? Oct 26 02:00:59 OT: people, are any of your ****@handhelds.org forwards working? Oct 26 02:01:05 RP: by power management problems do you mean wont suspend, or just that sound device wont power down? Oct 26 02:01:06 mine is dead since the move Oct 26 02:01:25 RP: is it me or is the 770 not suspending but shutting down and rebooting? Oct 26 02:01:29 mickeyl: I'm not tried maemo-image. Some custom xserver images worked fine though Oct 26 02:01:31 mickeyl: I dont use hrw@hh so cant tell Oct 26 02:01:41 RP: do you have udev automatically bringing up usb0? Oct 26 02:02:06 mreimer_: Not yet. We need to sort out udev and network interface handling in OE Oct 26 02:02:20 XorA: I've not acutally tried sound suspend/resume :) Oct 26 02:02:42 XorA: I just mean the sound chip doesn't power up/down its internal domains depending on usage at the moment Oct 26 02:02:49 anything new from Dirk lately? 2.4.18 on my tosa is really killing me... Oct 26 02:03:00 RP: ok, just checking what I should look out for Oct 26 02:03:07 mickeyl: He hasn't been in IRC for 29 days :-/ Oct 26 02:03:16 bummer. Oct 26 02:03:27 RP: thanks, Ill let you know if I get anywhere with it Oct 26 02:03:54 mreimer_: We're going to hit this when we remove pcmcia-cs so we need to fix it Oct 26 02:04:20 RP: with pcmciautils we dont have module autoload Oct 26 02:04:31 hrw|work: Yes we do Oct 26 02:04:42 hrw|work: It worked for me anyway... Oct 26 02:04:54 RP: hmm. I'll look do I have up-to-date env on Z Oct 26 02:05:19 ARGh... Oct 26 02:05:30 my home machine is getting more and more unstable ;( Oct 26 02:05:31 hrw|work: There may have been something I had to tweak - if certainly seen it work though... Oct 26 02:08:24 XorA: I'm particuarly interested in the apps that work or don't work with the alsa driver. I know madplayer works in oss mode but not alsa mode for example (or was it the other way around?) Oct 26 02:08:41 These are all bugs in the SoC audio code so I can pass them back to Liam Oct 26 02:08:57 RP: I will try and get that kernel on today and test Oct 26 02:10:26 XorA: There's no rush. I know people like features to play with though :) (hint: try alsamixer to see what you can fiddle with) Oct 26 02:10:55 There are a lot more buttons than the oss driver had as everything in the codec is configurable... Oct 26 02:11:30 mickeyl: Under alsa we can enable the speaker and the headphones at once btw. The share the volume control though :-(. Under cxx00, the two are independent Oct 26 02:17:53 RP: do you enabled oss pcm emulation? Oct 26 02:18:05 hrw|work: Its available as a module Oct 26 02:18:26 ok Oct 26 02:18:50 fscking reiserfs oopsed on home machine so I'm even unable to reboot ;( Oct 26 02:21:25 hrw|work: magic sysreq Oct 26 02:21:47 XorA: remotely? Oct 26 02:21:55 hrw|work: yeah there is a /proc entry for it Oct 26 02:22:09 hrw|work: saved my bacon when ext3 oopsed on me Oct 26 02:22:15 mickeyl: it seems the 770 doesn't do suspend, but relies on dyntick Oct 26 02:23:10 /proc/sys/kernel/sysrq Oct 26 02:24:52 XorA: I cannot use 'su' or sudo ;( Oct 26 02:25:24 hrw|work: hmmm, and its your / that has gone? Oct 26 02:25:58 XorA: looks like it is rather my build partition but each su or sudo == D process Oct 26 02:26:13 28753 ? D 0:00 svn update Oct 26 02:26:13 28770 ? D 0:00 sudo bash Oct 26 02:26:13 28790 ? D 0:00 su Oct 26 02:26:37 Dead and Zombies is something which I dont like in Linux Oct 26 02:28:27 looks like I need to have suided copy of bash somewhere in homedir.. Oct 26 02:28:46 hrw|work: your not running amd64 are you? Oct 26 02:29:15 RP: excellent! it has always bugged me that speaker was getting muted with headphones plugged in Oct 26 02:29:32 koen: dyntick? Oct 26 02:30:00 mickeyl: http://www.muru.com/linux/dyntick/ Oct 26 02:30:14 XorA: normal x86 athlon Oct 26 02:30:36 and http://www.free-it.de/archiv/talks_2005/paper-11017/paper-11017.html Oct 26 02:30:56 hrw|work: your problems lately remind me of amd64 before kernel and bios was fixed Oct 26 02:30:57 very interesting Oct 26 02:31:12 so on/off button really triggers shutdown/reboot Oct 26 02:31:22 XorA: here it is probably dying PSU which I cannot replace due to lack of any spare cash Oct 26 02:31:29 sometimes it shuts down here, sometimes it automatically reboots. a bit odd. Oct 26 02:31:54 the powerbutton behaves the same as a nokia phone Oct 26 02:32:05 my last nokia phone was a 2110 Oct 26 02:32:06 short press = menu, long press = shutdown Oct 26 02:32:27 hmm in what is this dyntick approach different from cpufreq? Oct 26 02:32:38 my current nokia phone is my last nokia phone Oct 26 02:32:46 hrw|work: have you hoovered the PSU? Oct 26 02:32:55 cpufreq lowers the frequency, dyntick just sleeps between cycles Oct 26 02:33:20 XorA: ke? Oct 26 02:33:42 XorA: my dictionary cannot help me with translation.. Oct 26 02:33:46 hrw|work: get all the dust and crap out of it, get it cooling properly again Oct 26 02:33:56 hoover, vacumn, suck stuff up Oct 26 02:34:11 XorA: its clean, it has fan replaced Oct 26 02:36:27 ok. it's / which died.. (or even /usr/lib) Oct 26 02:36:37 fsck it Oct 26 02:37:03 it sucks when you cant even ssh root@ Oct 26 02:37:35 XorA: on my system only one account is allowed to ssh in Oct 26 02:38:51 hrw|work: load the machine to the max and hope for spontaneous reboot :-) Oct 26 02:39:11 XorA: or hang.. Oct 26 02:39:35 ~hail opie-sh Oct 26 02:39:39 * ibot bows down to opie-sh and chants, "I'M NOT WORTHY!!" Oct 26 02:50:41 back to alsa: iirc ipaq which used alsa used also oss-pcm emulation to get all software working **** ENDING LOGGING AT Wed Oct 26 02:59:58 2005