**** BEGIN LOGGING AT Sun Jun 18 02:59:58 2006 Jun 18 03:07:17 * myopiate checks his pulse Jun 18 03:07:21 whew! Jun 18 04:47:52 hey people Jun 18 04:47:55 whats up today? Jun 18 06:23:33 morning all ! Jun 18 06:25:24 i need a clue, i have to work with 2 different architectures, what is best way to proced tryng to minimize the duplication disk space needed ? Jun 18 06:25:49 hey gremlin[it] Jun 18 06:26:08 short of using the same source dir, i dont think there is anything Jun 18 06:27:44 at now i 'duplicate' what is usually called the 'build' dir ... Jun 18 06:28:07 so i have the oe repo and sources in common ... Jun 18 06:28:21 hi emte Jun 18 06:30:07 yeah, i dont know taht you can "share" anything else since tehy would differe at the binary levels Jun 18 06:33:15 in fact :) Jun 18 06:34:08 i hoped at least native/cross parts could share but ep93xx and ipaq uses different cross-compilers :( ... Jun 18 06:34:31 reminds me Jun 18 06:34:46 you have the info to access the hh.org kernel tree? Jun 18 06:35:39 i'll get some time starting next week that i can start playing with the h36 battery/sleeve for 2.6 Jun 18 06:36:01 u also with battery ... Jun 18 06:36:19 no i don't have write permission to cvs kernel tree Jun 18 06:37:10 nah i just want the location to pull Jun 18 06:37:27 someoen else is playing with the battery+sleeve? Jun 18 06:38:03 i'm (loosely) working on micro in h3600 ... it control about all Jun 18 06:38:57 yeah i would finally like to get the sleeve battery working atleast with the 2.6 kernel, it's never worked with 2.4 Jun 18 06:39:17 ahhh Jun 18 06:39:28 sleeve ... Jun 18 06:39:30 wait Jun 18 06:39:30 since, i've played alot with the dallas devices it shouldnt bee too bad Jun 18 06:41:28 but i should 'export; battery info easily to othe rkernel part .. and i'm looking how batteries behave on the other part of kernel Jun 18 06:43:06 sleeve is not managed by micro ... except recognising if exist a second battery ... Jun 18 06:43:09 SPI is also used by sleeve ? Jun 18 06:43:16 ... Jun 18 06:43:30 from when i looked the h36 doent use spi Jun 18 06:43:36 does not* Jun 18 06:44:03 was i2c Jun 18 06:44:40 The Battery is controlled/monitored by the Dallas DS2760 chip inside/on the IPAQ itself. is a note i have to myself Jun 18 06:44:54 i just check ... i on the atmel micro i have comamnd to read/write/lock/unlock/ack/... on SPI Jun 18 06:45:00 The DS2760 is an I2C bus device. Jun 18 06:45:11 hrmm Jun 18 06:45:28 h3600 ??? the fisrt color ipaq ? Jun 18 06:45:41 yeah Jun 18 06:46:12 maybe i've misread something Jun 18 06:46:23 mhhhh i get from micro chemistry,voltage and status of battery Jun 18 06:47:18 maybe the dallas is conneted to the micro and not to the SA1110 ... Jun 18 06:47:28 hmm why did i note i2c ... the datasheet says OneWire Jun 18 06:50:05 on the "compaq ipaq h3600 hardware design specification" i found nothing about battery on one-wire Jun 18 06:51:27 http://pdf1.alldatasheet.com/datasheet-pdf/view/58628/DALLAS/DS2760.html Jun 18 06:51:41 aha, the sleeves are spi devices Jun 18 06:51:53 * emte wonders where i got i2c from ... Jun 18 06:55:23 trange cause some analog signal are routed to the micro ... charger current monitor and main battery 2/3 voltage sense input ... Jun 18 06:55:35 s/trange/strange/ Jun 18 06:56:22 seem all is done by hand (software) bythe atmel micro ... Jun 18 06:56:41 which atmel? Jun 18 06:56:52 maybe it is hardcoded for it Jun 18 06:57:32 yes hardcoded ... i don't remember the atmel model ... baybe joshua said me in past ... Jun 18 06:58:16 if it was software upgreadable we could change the bad touchscreen behaviour of early ipaqs :) Jun 18 06:59:40 hahaha Jun 18 07:00:25 AT90LS8535 Jun 18 07:00:47 yes i remember at90 ... Jun 18 07:01:08 have u opened your ipaq ? Jun 18 07:01:39 yes, and the sleeve+battery Jun 18 07:01:55 or you mean this min? Jun 18 07:01:57 no Jun 18 07:02:05 http://openwince.sourceforge.net/jtag/iPAQ-3600/ Jun 18 07:02:19 there is a disassasembly on that page Jun 18 07:04:37 hmm Jun 18 07:04:56 i wonder if i can get a jtag on the atmel chip Jun 18 07:05:13 ahhah Jun 18 07:05:51 seeing as it is exposed it shouldnt be hard Jun 18 07:06:17 need to find an atmel raeder then i guess Jun 18 07:06:22 reader* Jun 18 07:06:24 mhh but usually all devices ona board share the same jtag bus ... or no ? Jun 18 07:07:20 was thinking about just trying to access the atmel Jun 18 07:08:08 but at90 have jtag ??? or they use ISP (?) for program flash ? Jun 18 07:08:11 altho just a scope would work too Jun 18 07:08:33 dunno Jun 18 07:08:45 mhh what is the big HTC chip on the ipaq ? Jun 18 07:08:48 have to look at the board closer Jun 18 07:09:04 ? Jun 18 07:09:06 which one? Jun 18 07:09:31 ah ... dunno Jun 18 07:09:42 my gess without looking is a display controller Jun 18 07:09:47 guess* Jun 18 07:10:16 h3600 don't have display controller ... Jun 18 07:10:33 it's near the rams Jun 18 07:16:41 HTC chip (HTC 30H88028-00 2226560XB101 JAPAN 0320EAI F0103ZAF) - Asic3 Jun 18 07:16:44 is from the h41 Jun 18 07:19:39 h41 ? Jun 18 07:19:53 asic3 on ipaq 3600 ?!?!?! Jun 18 07:20:07 too many surprise today ... Jun 18 07:22:51 almost imposible to find datasheets as well Jun 18 07:25:40 mhhh u should work on battery cause u already know the one-wire right ? Jun 18 07:27:01 i use it on PICs Jun 18 07:27:04 its easy Jun 18 07:27:56 i wonder who it was that i know works for atmel ... Jun 18 07:28:06 er Jun 18 07:28:12 hmm Jun 18 07:31:57 time for more math and then some sleep Jun 18 07:33:15 my roadmap is battery,touch,usb,SPI Jun 18 07:33:32 for me is time to wotk on cirrus chip :) Jun 18 08:15:00 hi, does anyone know a good program to create patches? until now I always did this manually, got fed up with that and tried holger's patcher but that one only gives me gray hairs as it does not do what I understood it would do Jun 18 08:15:03 what do others use? Jun 18 08:17:50 hmm Jun 18 08:17:55 i think JP had a tool Jun 18 08:19:24 eFfeM: I create patches by hand too. Jun 18 08:19:27 jnc thanks for the suggestion Jun 18 08:20:03 patcher looked nice but I had to create the .patches dir manually and then it would not generate the patch Jun 18 08:20:06 nice idea though Jun 18 08:20:28 eFfeM: are you looking for a script that create patches from a 'monotone diff' for example ? Jun 18 08:20:35 cyrilRomain: this is cumbersome especially if patches are a few deep Jun 18 08:20:58 basically I want to create patches I made (e.g. against the head or against unpatched copies) Jun 18 08:21:29 currently I do a cp file file.orig; vi file; diff -u file.orig file > file.patch Jun 18 08:22:06 but if you have to patch multiple files and the dirs are a few deep and the files may be in different dirs then this becomes a little bit cumbersome Jun 18 08:22:42 i could copy the tree and diff it, but for large packages that is also not too nice (although sometimes perhaps the best way) Jun 18 08:23:46 need to go now, back in a few hrs. if anyone has other ideas, feel free to respond, I'll read back when I return Jun 18 08:23:49 eFfeM: I think Justin Patrick script (in contrib/mtnpatch.py) could be a start then. It is intended to do cherrypicking between two tree Jun 18 08:24:10 tnx will look into that (but my perl is not that good) Jun 18 08:24:22 might also try to carve my own c program Jun 18 08:24:22 it is written in python Jun 18 08:24:32 and is quite simple Jun 18 08:24:38 well my python skills match my perl skills quite well Jun 18 08:24:54 really need to go now, later Jun 18 08:26:40 RP: do you have any information on the Poodle sound system? I will have some time available in July and would like to get a look at things to see if I might be able to help out. Jun 18 08:27:15 ~seen RP Jun 18 08:27:19 rp is currently on #gpe (3d 22h 2m 8s) #oe (3d 22h 2m 8s) #openzaurus (3d 22h 2m 8s) #handhelds.org (3d 22h 2m 8s). Has said a total of 107 messages. Is idling for 1d 19h 59m 55s, last said: 'slaine_: Yes, it will try and cross compile'. Jun 18 09:26:20 eFfeM-away: Have you considered quilt? Works here quite nicely. Jun 18 09:29:24 eFfeM-away: a script that basically do what you do by hand http://oe.pastebin.com/716123 As soon as you quit vi the patch is created. Jun 18 09:33:31 hello ! Jun 18 09:33:40 eFfeM-away: otherwise it would be interesting not to create .orig and .patch files; but create to patches on demand from the 'monotone diff' output... Jun 18 09:54:44 03pH5 07org.oe.dev * r48e6836e... 10/packages/libnotify/libnotify_0.4.2.bb: libnotify: add 0.4.2 Jun 18 10:21:22 hi Jun 18 10:27:49 How is the initial ipkg.conf generated for the rootfs? Jun 18 10:29:22 Laibsch: ipkg-colateral Jun 18 10:30:57 from ipkg.conf.comments? Jun 18 10:31:13 check packages/ipkg/ Jun 18 10:31:22 I wrote it over 2 years ago Jun 18 10:32:03 Thanks for writing it. Still important today ;-) Jun 18 10:33:42 rgrep ewi546 /usr/src/openembedded/org.openembedded.oz354x/packages/ipkg/ does not return anything. Jun 18 10:33:47 like many other things Jun 18 10:33:55 http://bugs.openembedded.org/show_bug.cgi?id=1102 Jun 18 10:34:01 is what I am talking about. Jun 18 10:34:05 Laibsch: FEED_URIS are in distro config Jun 18 10:34:14 Laibsch: no www browser here Jun 18 10:34:16 Tried to fix it completely by giving a patch but failed. Jun 18 10:34:41 hrw|husky: There are a couple of places where there are // in the URL. Very minor issue, but.. Jun 18 10:34:49 ah.. Jun 18 10:35:33 thats duplicate bug Jun 18 10:35:40 Laibsch: don't know quilt, will have a look at it Jun 18 10:35:53 eFfeM: quilt is very easy. Jun 18 10:35:54 Laibsch: search for "//" and mark yor as duplicate Jun 18 10:35:54 cyrilRomain: will look at the script and at mt diff Jun 18 10:36:09 eFfeM: I learned it here. google for Laibsch and quilt. Jun 18 10:36:58 cu Jun 18 10:37:34 Laibsch: tnx found the msg Jun 18 10:37:36 * eFfeM is reading Jun 18 10:37:41 hrw|husky: I did before posting it. The other // is about "update-alternatives: Linking //usr/bin/ipkg to /usr/bin/ipkg-cl" also reported by me. Jun 18 10:38:14 eFfeM: I love it. you just have to be sure to unapply all patches before updating via monotone. Jun 18 10:40:26 morning Jun 18 10:41:32 morning all Jun 18 10:42:03 good morning! Jun 18 10:43:09 yo RP & Laibsch Jun 18 10:49:20 morning all Jun 18 10:50:42 morning do13 Jun 18 10:50:45 good morning Jun 18 10:50:49 hey CoreDump|home Jun 18 10:51:04 RP: -extern const struct linux_logo logo_oz320_clut224; Jun 18 10:51:07 +extern const struct linux_logo logo_oz480_clut224; Jun 18 10:51:28 in oz logo patch Jun 18 10:53:07 hmm where did qpf-unifont go Jun 18 10:57:15 do13: Ah, right, sorry :) Jun 18 10:57:44 RP: np :) Jun 18 10:59:06 do13: Its in logo_oz-r2.patch - will push when I add 2.6.17 Jun 18 10:59:54 RP: do you have 2.6.17 ready? Jun 18 11:00:40 do13: The git kernel is near enough 2.6.17 :) Jun 18 11:00:44 (and works) Jun 18 11:01:00 I'll push a specific 2.6.17 .bb file shortly Jun 18 11:01:57 RP: git works here too :) Jun 18 11:02:11 what's about liam latest asoc? Jun 18 11:04:07 do13: He's due to make a new release soon which should see the problems we saw with the last couple of versions Jun 18 11:04:22 fixed Jun 18 11:05:05 RP: ok Jun 18 11:08:03 hrw|gone: ping? Jun 18 11:18:22 CoreDump|home: bug 1070 maybe? Jun 18 11:18:49 It's not a nice font anyway. Jun 18 11:19:44 Laibsch: yeah, it turned out that it was a bad dependency in my images and the font wasn't used at all Jun 18 11:19:53 I just removed it Jun 18 11:20:18 you're gonna save quite some space then Jun 18 11:20:26 that is good. Jun 18 11:20:53 it wasn't ever installed. It was a compile-time dependency because it was probably used years ago Jun 18 11:22:20 * CoreDump|home has another go to get his DVB-T stick to work on his Z Jun 18 11:25:43 Is there a good reason that /tmp is not mounted as tmpfs by default? Jun 18 11:30:32 hi Jun 18 11:30:34 RP: pong Jun 18 11:31:51 bl06 in c760 make it better to use but batterylife suxx Jun 18 11:39:04 hi all Jun 18 11:39:46 ~help Jun 18 11:40:45 ~help botmail Jun 18 11:41:03 mr_nice: /query ibot is nice way of asking ibot Jun 18 11:41:25 ~husky Jun 18 11:41:28 extra, extra, read all about it, husky is Sharp Zaurus SL-C760 or a dog Jun 18 11:41:28 thx Jun 18 11:49:34 * hrw|husky fixed #1100 Jun 18 11:54:27 hrw|husky: Should be be applying the 24-hostap patch or not? Jun 18 11:54:40 not Jun 18 11:54:58 hrw|husky: ok, I'll remove it from the git kernel. Jun 18 11:55:00 RP: it was not applied mainline and will not be Jun 18 11:55:13 hrw|husky: That's fine, I just wondered :) Jun 18 11:55:46 RP: push ckermit changeset into .dev (from .oz345x) - i have to go Jun 18 11:56:20 03hrw 07org.oe.oz354x * r3d924e26... 10/packages/ckermit/ckermit_211.bb: ckermit: ship properly symlinked binaries - close #1100 Jun 18 11:56:27 hrw|husky: We need to decide what we mean by marking patches are merged and pending. I was defining merged to be in mainline and pending "due to be applied to mainline" Jun 18 11:56:56 hrw|husky: ok Jun 18 11:57:18 merged = mainline already Jun 18 11:57:37 pending = sent mainline, accepted, not yet in mainline Jun 18 11:57:54 hrw|husky: The two hostap patches you marked as merged are not in mainline yet, just pending Jun 18 11:58:26 RP: hostap and pcnet? Jun 18 11:58:29 yes Jun 18 11:58:42 They didn't make 2.6.17 as far as I could see Jun 18 11:59:02 ok then i should restatus them Jun 18 11:59:24 ~akita Jun 18 11:59:26 i heard akita is the Sharp SL-C1000, or a dog Jun 18 11:59:33 hrw|husky: I'll take care of it - I'm refactoring the .bb files a bit more Jun 18 11:59:37 both are in "send to linus" branch of linux-netdev-2.6 anyway Jun 18 11:59:57 hrw|husky: So they're pending :) Jun 18 12:00:09 hrw|husky: I just want to make sure we mean the same things :) Jun 18 12:00:23 yes Jun 18 12:00:33 hrw|husky: They'll get merged in the merge window along with the other ones I have marked as pending Jun 18 12:04:05 RP: so we have: merged, pending - "do not touch them" Jun 18 12:04:30 RP: then hack as "it work but is too ugly to send mainline" Jun 18 12:04:47 RP: how to mark "we work on it" patches? Jun 18 12:05:02 hrw|husky: I've also added "umergable" for the logos and release versions patches Jun 18 12:05:16 hi all Jun 18 12:05:20 hi florian Jun 18 12:05:21 RP: good point Jun 18 12:05:23 hi florian Jun 18 12:05:45 RP: we also need to add descriptions to patches in patches Jun 18 12:05:46 hrw|husky: I guess the work on it patches are the remaining ones Jun 18 12:06:48 hrw|husky: I've slowly been working towards the descriptions... Jun 18 12:07:26 RP: i know but there are many of them so if someone want to help us with kernel work then it would be nice to know which patch can be taken to improve when you and other kernel hackers are not available Jun 18 12:08:16 hrw|husky: That would be better done as a description somewhere of what work is outstanding on the kernel Jun 18 12:09:23 RP: descriptions should contain infos like "sent to linux-fbdev for discussion" and if they are from mailing lists then some headers so it will be easier to find original thread Jun 18 12:10:18 RP: anyway we need OE PatchPolicy for it probably Jun 18 12:10:20 hrw|husky: If patches are from mailing lists, I just use a saved email Jun 18 12:10:39 there are many patches which are known only for their authors Jun 18 12:10:41 hrw|husky: Agreed. The zaurus kernel .bb file is kind of a leader in patch development :) Jun 18 12:10:47 have to go Jun 18 12:10:49 cu Jun 18 12:10:54 cu hrw|husky Jun 18 12:11:04 later hrw|gone Jun 18 12:12:30 Since I started the kernel, we've gone through over 745 patches taking up 12.6MB of space :) Jun 18 12:12:54 good work Jun 18 12:13:11 and that doesn't include the early ones before I standardised the naming... Jun 18 12:13:32 Any chances of it happening for the collie as well? I heard that SD issues eased a bit recently. Jun 18 12:13:55 Laibsch: SD isn't the problem. We have no docs on the controller on collie Jun 18 12:14:16 so it is not going to happen in the foreseeable future? Jun 18 12:14:28 There are some ways it could happen Jun 18 12:14:36 But don't count on it Jun 18 12:15:19 ok Jun 18 12:15:47 hi koen Jun 18 12:15:59 yo koen Jun 18 12:16:25 hey RP Jun 18 12:16:29 hey CoreDump|home Jun 18 12:17:19 hey koen Jun 18 12:17:37 hey pH5 Jun 18 12:20:51 do13: We should think about merging some of the changes to sharpsl_pm. Do you have anything extra to the patch I have you'd like to see merged? Jun 18 12:32:39 Hi, OE again refuses to build. I have already deleted everything from tmp and tried to restart, but failed. OE and bitbake are up to date. Jun 18 12:33:01 http://pastebin.ca/66785 Jun 18 12:34:10 Laibsch: Strange. Nothing much should have changes to do with that version of glibc... Jun 18 13:04:59 hi all Jun 18 13:08:59 ibat: botmail to mickey|fifa2006: Hi, the compilation is now complete. With MACHINE_KERNEL_VERSION=2.4 everything works well :) - wir 2.6 as setting it run into a Error (http://pastebin.com/716176) but I don't know if this depend on the MACHINE_KERNEL_VERSIN thing, because I have never tried to build Kernel 2.6 for SIMpad. I think it has nothing to do with it. Jun 18 13:09:16 sorry Jun 18 13:10:15 ibot: botmail to mickey|fifa2006: Hi, the compilation is now complete. With MACHINE_KERNEL_VERSION=2.4 everything works well :) - wir 2.6 as setting it run into a Error (http://pastebin.com/716176) but I don't know if this depend on the MACHINE_KERNEL_VERSIN thing, because I have never tried to build Kernel 2.6 for SIMpad.I think it has nothing to do with it. Jun 18 13:13:52 * chouimat is back. Jun 18 13:14:02 morning Jun 18 13:23:30 mr_nice: no one maintains the SIMpad atm., so errors are expected. Jun 18 13:23:37 mr_nice: good to know 2.4 still compiles Jun 18 13:25:44 mickey|: yes this one is realy good! Jun 18 13:27:53 hello Jun 18 13:28:52 yo lardman Jun 18 13:28:57 heh Jun 18 13:29:01 read that oesf post? Jun 18 13:29:09 someone wants to start yet another build system Jun 18 13:29:10 sheesh Jun 18 13:29:46 based on the false hope that it will be simpler to modify Jun 18 13:33:30 hey mickeyl Jun 18 13:33:35 koen! Jun 18 13:33:40 how's the vacation? Jun 18 13:34:01 koen: i started blogging :) added it to planet.linuxtogo.org Jun 18 13:34:18 I noticed :) Jun 18 13:34:22 mickeyl: No, what's the url for that? Jun 18 13:34:53 lardman: http://www.oesf.org/forums/index.php?s=34bef8080c75549c1d9d244e087dff67&showtopic=20021 Jun 18 13:35:09 mickeyl, koen: Will we keep the current filenaming scheme for the files used to flash? I'm going to (try to) update the flashing instructions Jun 18 13:35:27 not the initrd.bin, but the name that the downloads have, etc. Jun 18 13:35:34 btw, I saw an a780 at the duty free shop at the airport Jun 18 13:35:42 lardman: yeah. these are the names as come from bitbake. Jun 18 13:36:02 koen: you'll get yours right after coming home? Jun 18 13:36:13 mickeyl: yep Jun 18 13:36:22 mickeyl: cool Jun 18 13:36:41 fwiw: monotone 0.27 has been released and supports netsync over ssh Jun 18 13:37:55 koen: yay! check oe.org, yesterday we posted the migration scheme. Jun 18 13:38:04 I'll substitue 0.26 w/ 0.27 then Jun 18 13:38:25 I also noticed that :) Jun 18 13:38:51 heh, ok, then you know about all news already :) Jun 18 13:38:59 except one maybe Jun 18 13:39:04 I'm _a bit_ tempting w/ mda pro Jun 18 13:39:13 just a tiny bit until now though Jun 18 13:39:18 * mickeyl smiles innocently Jun 18 13:39:30 i joined #htc-linux and will have an eye on the progress Jun 18 13:40:38 and i have nightmares about EFL based interfaces using the overlay patch for a great phone UI Jun 18 13:42:19 mickeyl: heh, i couldn't resist... :-) Jun 18 13:42:48 well, the problem is like all the ipaqs Jun 18 13:42:49 Is merging changesets across branched easier now? Jun 18 13:42:51 not even 2.4 drivers Jun 18 13:43:00 so everything based on reverse engineering Jun 18 13:43:01 * koen is tempted to try coding up something with efl++ Jun 18 13:43:01 which is a tough job Jun 18 13:43:31 koen: I need to update it to support the new evas module based reorganization Jun 18 13:43:47 Crofton: i don't think there were any changes regarding that functionality. njs ? Jun 18 13:44:54 Crofton: take a look at justinp's mtnpach in org.oe.dev/contrib Jun 18 13:45:25 ok Jun 18 13:46:00 I'm going to be on vacation from July1-10 so I will miss much of the 'action" :) Jun 18 13:46:47 mickeyl: that's true Jun 18 13:47:51 ljp: thanks for taking the time to answer on qtopia-interest. I'm still hoping to see some action soon. Jun 18 13:48:04 has all the mdapro stuff been added to oe yet? Jun 18 13:48:41 as far as i can see, not a bit yet Jun 18 13:50:03 that's what I feared Jun 18 13:51:41 koen: have you seen maemo2 beta? Jun 18 13:51:48 it's kind of nice Jun 18 13:52:02 I've been trying to download it Jun 18 13:52:20 but it seems I can only get a 0.7kB/s link to nokia Jun 18 13:52:33 whats new in maemo2? Jun 18 13:52:43 I did see florian's screenshots Jun 18 13:52:52 mithro: eabi Jun 18 13:53:59 mithro: more polished UI, a bit faster, voip client Jun 18 13:54:06 (or something like this) Jun 18 13:54:34 so how is everyone going? Jun 18 13:54:39 got an exam tommorrow :/ Jun 18 13:54:59 good so far Jun 18 13:55:04 *crossing fingers* Jun 18 13:55:09 are you prepared? Jun 18 13:55:59 pH5: did you test the Xorg kdrive yet? Jun 18 13:56:35 mickeyl: it's a bit of a dangerous exam, it's one of those bullshit exam Jun 18 13:57:33 but i'm pretty good at that normally Jun 18 13:57:57 koen: it works, but there are strange crashes if it doesn't find its fonts (without vera installed) Jun 18 13:59:21 pH5: that's pretty good news Jun 18 14:04:22 do you have any screenies of the new maemo? Jun 18 14:05:57 mickeyl: might be worth adding something to your post about the size of the tools and their stability Jun 18 14:06:40 mickeyl: perhaps also mention the option of native toolchains, etc. Jun 18 14:06:43 mithro: http://fl0rian.wordpress.com/ Jun 18 14:07:55 lardman: *nod* Jun 18 14:11:28 it's still a bit ugly Jun 18 14:13:44 03rpurdie 07org.oe.dev * r62300e44... 10/packages/linux/ (28 files in 3 dirs): linux-oz-2.6: Add 2.6.17. Sync changes accross all .bb files. Classify more patches. Fix a logo bug reported by Dirk on Tosa. Rework the poodle patches and remove their hack status. Jun 18 14:20:38 it looks better on the devices? Jun 18 14:22:59 looks good to me Jun 18 14:23:08 don't forget that the resolution is pretty high Jun 18 14:23:17 that doesn't show via screenshots Jun 18 14:30:56 mickeyl: Going back to file naming... Is hrw's correct - gpe-image-initrd.bin for whichever you download? Or *-image-*-tosa.initrd.bin or *-*-tosa.rootfs.jffs2? Jun 18 14:31:16 * lardman hasn't built an initrd.bin in a while Jun 18 14:34:08 CoreDump|home: Are you happy with the position the locomo keyboard is now in? Jun 18 14:34:49 CoreDump|home: I've merged everything into http://www.rpsys.net/openzaurus/patches/locomo_kbd_tweak-r0.patch Jun 18 14:34:50 RP: hmm? I didn't build Poodle in weeks Jun 18 14:35:21 CoreDump|home: We should think about finalising that and getting it into mainline Jun 18 14:35:26 yeah Jun 18 14:35:32 ASAP as the merge window has just opened Jun 18 14:35:41 well, the driver needs numlock support Jun 18 14:36:06 but that's it AFAIR Jun 18 14:36:16 CapsLock works? Jun 18 14:36:29 in the console yes, not in Opie or GPE Jun 18 14:36:59 s/console/VT Jun 18 14:37:48 So Numlock is more of a keymap issue? Jun 18 14:37:53 not really Jun 18 14:38:18 if numlock is on, the driver needs to emmit different keycodes for the "numlock" number-keys Jun 18 14:38:39 hi all Jun 18 14:38:55 CoreDump|home: Mainline will not see it like that Jun 18 14:38:57 hi cedric Jun 18 14:38:59 that's what collies kbd driver in 2.4 does fwiw Jun 18 14:39:19 RP: well, what's the alternative? You can't map something like that Jun 18 14:39:26 is ZKERNEL_VERSION has changes in the local.conf file? set to 2.6 compile a 2.4 kernel... Jun 18 14:39:31 CoreDump|home: All the Zaurus keyboard drivers in 2.4 are utter junk Jun 18 14:39:45 cedric: yes Jun 18 14:39:55 so what the new config then? Jun 18 14:40:05 RP: i know... The problem is that somehow we need to emmit different keycodes for affected keys when numlock is on Jun 18 14:40:18 cedric: Read local.conf.sample Jun 18 14:40:21 even on my notebook that is the case Jun 18 14:40:47 03lenehan 07org.oe.dev * rba21f1fa... 10/packages/ckermit/ckermit_211.bb: ckermit 211: Fix dangling symlink for ckermit-sshusb. Closes bug #1100. Jun 18 14:40:48 CoreDump|home: Is that done in hardware though? Jun 18 14:41:22 I don't think so as the kbd doesn't have a real num block Jun 18 14:41:29 but I can't be certain sorry Jun 18 14:42:16 I tried to make senso of the kbd drivers but failed =) Jun 18 14:42:20 RP: thx Jun 18 14:42:27 CoreDump|home: We can easily hack the kernel driver but mainline won't take it and will say its a userspace problem Jun 18 14:43:09 RP: that's all nice and well but I do not see a way tohandle that in userspace w/o maybe hacking yet-another-deamon Jun 18 14:43:29 capturing the keypresses and stuff. Way ugly Jun 18 14:45:37 CoreDump|home: The numlock mappings are set via the keymap Jun 18 14:45:54 CoreDump|home: We just need to work out the proper keymap settings to make it work Jun 18 14:46:23 RP: nope, you _have_ to emmit different keycodes, trust me. Jun 18 14:46:44 you can't map something like that :( Jun 18 14:46:45 CoreDump|home: That is not correct Jun 18 14:46:57 numlock is a key modifer, just like shift or ctrl but it can be locked on/off Jun 18 14:47:49 not according to my information Jun 18 14:47:56 http://linux.about.com/library/cmd/blcmdl5_keymaps.htm Jun 18 14:48:20 "The keyboard driver supports 8 modifiers. These modifiers are labeled (completely arbitrarily) Shift, AltGr, Control, Alt, ShiftL, ShiftR, CtrlL and CtrlR" Jun 18 14:49:55 CoreDump|home: I was looking at http://bama.ua.edu/cgi-bin/man-cgi?keytables+4 Jun 18 14:51:04 *cough* Sun OS *cough* Jun 18 14:51:16 hey Jun 18 14:51:16 CoreDump|home: I know :-/ Jun 18 14:51:31 RP: would have been to easy, murphys law and all Jun 18 14:51:36 hi zecke Jun 18 14:51:53 CoreDump|home: I knew I'd seen it done in keymap files before... Jun 18 14:52:19 hi zecke Jun 18 14:52:24 hey Jun 18 14:59:18 CoreDump|home: We can workaround this in the keymap by using an unused modifier key and emulating our own numlock functionality on that modifier key Jun 18 14:59:34 Laibsch: did you report #1105? Jun 18 14:59:49 Laibsch: please learn how to write bug reports... Jun 18 15:00:00 Laibsch: saying foo needs moo can only be marked as invalid Jun 18 15:00:05 RP: yes, that might work Jun 18 15:02:49 zecke: Thank you for your "nice" comments. Did you see this is an RFE? Am I responsible for the less than obvious situation regarding bugtracking? If you do not want bug reports, close the tracker. Jun 18 15:03:03 Or make it invitation only. Jun 18 15:03:26 Laibsch: RFE? Jun 18 15:03:27 I do think I do know how to write a bug report. And I write it to the most obvious place. Jun 18 15:03:39 zecke: What is your question? Jun 18 15:03:39 Laibsch: are you talking about E17, GPE, OPie, All?, Qtopia, Maemo? Jun 18 15:03:52 Laibsch: how should one know this... Jun 18 15:04:06 Laibsch: get one thing: Not everyone has the same packages installed as you have! Jun 18 15:04:14 Laibsch: this is why you need to specify meaningfull input Jun 18 15:04:32 Laibsch: I won't use the OpemEmbedded bugtracker to report misc. GNOME bugs Jun 18 15:04:40 Ask me if you need clarification but don't rage like this. Jun 18 15:04:40 Laibsch: this would not make any sense Jun 18 15:04:52 I do not think that elicits bug reports, does it? Jun 18 15:05:08 Laibsch: lol, this is the n-th bug report you fail to write properly Jun 18 15:05:09 zecke: Then just leave it, OK? Jun 18 15:05:22 Then just kick my account, fuck it. Jun 18 15:05:46 Laibsch: why are you not able to write the complete application name? Jun 18 15:05:54 The bug report system does not seem to be written properly, I assume if it fails to ask for relevant info Jun 18 15:05:58 Laibsch: why are you not able to write the gui environment Jun 18 15:06:15 Because it does not have an app name in the GUI?! Jun 18 15:06:40 I will not spend any more time on arguing with you. Jun 18 15:06:41 Laibsch: well, to start: What GUI environment do you use? Jun 18 15:06:44 Laibsch: E17 or maemo? Jun 18 15:06:48 Just close them if you do not like them. Jun 18 15:06:52 lol Jun 18 15:07:03 Laibsch: I did not mark it as invalid but... Jun 18 15:07:09 PLEASE, make the BTS invitation-only. Jun 18 15:07:31 Laibsch: wtf? it is open to everyone, just write meaningfull bug reports Jun 18 15:07:49 Laibsch: How should anyone know which app you mean? Jun 18 15:08:04 Laibsch: just ask yourself this question, and you will write meaningfull bug reports Jun 18 15:08:13 crystal ball Jun 18 15:08:20 telepathy Jun 18 15:08:22 take your pick Jun 18 15:08:32 yeah, the bug tracker sucks because it has no telepathy support Jun 18 15:08:40 zecke: Laibsch is a droid, he can't think properly... Jun 18 15:09:21 yes, I can't think like a person doing nothing but OE, indeed. Jun 18 15:09:36 And I don't know all of OE. Jun 18 15:09:39 Laibsch: e.g. "I'm using GPE 2.7 and the application called 'Moo' which is seen in the folder 'Foo' should additionally allow me to do the ...' Jun 18 15:09:39 Indeed. Jun 18 15:09:42 Laibsch: lol Jun 18 15:09:54 Laibsch: you fail to see that what you are asking for doesn't even belong to OE Jun 18 15:10:02 And I DO write bug reports only to annoy the devs and keep them from doing meaningful stuff. Jun 18 15:10:06 Laibsch: it belongs to either KDE, GNOME, GPE, MAEMO, Opie, Qtopia Jun 18 15:10:08 And I do work for MS. Jun 18 15:10:12 And I do.... Jun 18 15:10:35 Laibsch: last thing before I mark this bug report as invalid. What GUI do you file this bug report against Jun 18 15:11:16 Laibsch: once we know the GUI, one can ask you to use the right place to report the bug Jun 18 15:11:27 Laibsch: So what GUI do you file this bug against? E17 or GNOME? Jun 18 15:12:50 Laibsch: Rule of Thumb: If you write a bug report. Always ask yourself: If I say 'foo' will the guys reading this bug report know what 'foo' is? Jun 18 15:13:11 Laibsch: it is not more than obeying this simple and straight forward rule, otherwise we can just skip using bug trackers Jun 18 15:13:54 rule of thumb: not all users will know all info necessary when they write a report. Some will not even know what info is necessary Jun 18 15:13:54 rule of thumb: some ppl with good intentions will be pissed off by reactions like this. Jun 18 15:14:11 Laibsch: please don't be pissed off Jun 18 15:14:32 Laibsch: well: "Timer app" Jun 18 15:14:36 zecke: As I said it seems you should skip it or better make it invitation-only. Jun 18 15:14:54 Laibsch: what is that? Should everyone involved with OE know what "Timer app" is? Jun 18 15:15:08 THEN ASK. Jun 18 15:15:13 I gave an example Jun 18 15:15:18 two in fact Jun 18 15:15:31 Laibsch: Ageed, users don't always realise what info is required. Someone is trying to help you understand what info is needed so in future your bug reports are more useful Jun 18 15:15:56 RP, yes that is OK. I am pissed by the "you are dumb"-attitude Jun 18 15:16:07 Laibsch: did I use 'dumb' Jun 18 15:16:17 Laibsch: I have asked you now 10 times to specify the GUI Jun 18 15:16:19 and by the "you seem to doing this on purpose"-attitude Jun 18 15:16:43 I am sure this has limited OZ's and thus OE's potential quite a many times in the past Jun 18 15:16:45 Laibsch: lol, Jun 18 15:17:18 Laibsch: Keep in mind English isn't the native language of some here and things do sometimes translate badly... Jun 18 15:17:30 Both ways :) Jun 18 15:19:05 infering attitude from non-native speakers utterances is never accurate Jun 18 15:22:58 Laibsch: BTW: 'pissed' in the UK means drunk Jun 18 15:23:58 zecke: It has at least two other meanings depending on context :) Jun 18 15:24:15 RP: I just saw you said something? Jun 18 15:24:29 I said: zecke: It has at least two other meanings depending on context :) Jun 18 15:24:51 I was PISSED could mean drunk? Jun 18 15:25:00 Laibsch: a bug tracker for opie can be found on the opie home page: http://opie.handhelds.org/cgi-bin/moin.cgi/BugReport Jun 18 15:26:08 too late he left Jun 18 15:26:13 * mickeyl shakes head Jun 18 15:26:31 heh Jun 18 15:27:33 zecke: It that case it was clear he meant angry (although usually you'd write that as "pissed off"). Jun 18 15:27:39 back to work... one needs to fix OE= OZ = OPIE = GPE thinking... Jun 18 15:27:52 It can also mean drunk or refer to someone who's lost bladder control... Jun 18 15:27:58 RP: Al was complains when I omit the 'off' Jun 18 15:28:07 yeah := Jun 18 15:30:03 * mickeyl prepares the #oe diplomatic bat and threatens to strike on the next occurance Jun 18 15:31:08 ~fear mickeyl Jun 18 15:31:29 Someone needs to teach ibot that one :) Jun 18 15:31:33 hehe Jun 18 15:32:40 mickeyl: I will just quit using irc, and subscribe to microsoft messenger Jun 18 15:33:12 that won't improve your communication skills :D Jun 18 15:33:44 mickeyl: I could talk with Al :} Jun 18 15:34:13 zecke: good idea. talking with women always makes you calm... since you're losing anyway Jun 18 15:34:14 heh Jun 18 15:34:27 zecke: adium Jun 18 15:35:05 koen: Well, I do not want a hotmail account Jun 18 15:35:37 you don't need to Jun 18 15:35:49 you only need to associate a passport to an email account Jun 18 15:35:51 koen: oh well, only MSN without hot? Jun 18 15:36:09 yes Jun 18 15:36:27 koen: I know MSN is big in NL, I will not use it! Jun 18 15:36:44 :) Jun 18 15:36:56 koen: and I will continue to say "Al shows anti-social behaviour by using MSN" Jun 18 15:36:57 my hotmail account predates the MS takeover Jun 18 15:39:00 Is there a way to compile Opie with mouse support? I just got my Bluetooth mouse to work with 'od -x /dev/input/mice' Jun 18 15:39:31 goxboxlive: QtE needs to be compiled with cursor support Jun 18 15:40:09 zecke: Hi , but how to i do that with OE? Do i have to patch something? Jun 18 15:40:15 goxboxlive: not easily. some guys at oesf.org did it with an additional launcher input method that reads from mice input, shows a cursor and feeds stuff back into the qws input system. Jun 18 15:40:20 goxboxlive: change the configure line Jun 18 15:40:32 zecke: you mean "leverage" gcc to exploit the full capabilities of "award winning" software created by "paid developers"? Jun 18 15:40:32 goxboxlive: or qconfig.h Jun 18 15:40:59 koen: I have ranted enough for today *guess* Jun 18 15:41:55 * koen waits for the museums to open Jun 18 15:42:28 koen: remind me where you are Jun 18 15:42:57 florian: barbados Jun 18 15:43:09 13 degrees north, 59 degrees west Jun 18 15:43:23 zecke: What about this line, do i have to comment it out to? : #define QT_NO_QWS_MOUSE_AUTO Jun 18 15:43:28 koen: cool Jun 18 15:44:11 ~seen prpplague Jun 18 15:44:16 prpplague is currently on #edev (2h 29m 47s) #uclibc (2h 29m 47s) #handhelds.org (2h 29m 47s). Has said a total of 48 messages. Is idling for 52m, last said: 'hehe'. Jun 18 15:44:25 goxboxlive: not the only line Jun 18 15:44:31 goxboxlive: MOUSE and CURSOR are good keywords Jun 18 15:45:04 yes, i also found cursor, so i comment out both of them then. Thx again zecke :-) Jun 18 15:45:48 goxboxlive: have a look at http://www.oesf.org/forums/index.php?showtopic=8611&hl=mouse Jun 18 15:45:50 goxboxlive: and then we need a way in Opie to switch the mouse prototype on the fly :) Jun 18 15:46:02 (which is possible, we just lack a GUI) Jun 18 15:46:13 thx both Jun 18 15:47:42 mickeyl: It would be good to get that into OE as qemuarm would then work with opie Jun 18 15:48:03 RP: ok, I'll note that onto my magic TODO list. Jun 18 15:48:26 mickeyl: Its also on mine but low down as I don't have a clue what I'm doing Jun 18 15:48:32 heh Jun 18 15:48:46 mickeyl: for qemuarm a different qconfig.h and different 'opie' script would be enough?! Jun 18 15:49:06 mickeyl: we would only change QWS_MOUSE_PROTO and QWS_CURSOR=1 Jun 18 15:49:17 one way, yes. I'm leaning more towards the japan input helper because that doesn't require changes in qte config Jun 18 15:49:46 mickeyl: yeah, in a perfect world (Qtopia4) we would have hot switchable input devices Jun 18 15:49:56 mickeyl: I made qemuarm run as RGB so that's no longer an issue for opie Jun 18 15:50:13 RP: read that. excellent work! Jun 18 15:50:22 zecke: kdrive doesn't manage that yet... Jun 18 15:50:38 this qemuarm stuff brings me back to the question is there a thing like bitbake "gcc-image" for qemuarm? Jun 18 15:50:51 RP: I hope to get next month a week off :( Jun 18 15:51:03 mr_nice: There is in poky Jun 18 15:52:05 * mickeyl wanders off to see Brazilian vs. Australia Jun 18 15:52:21 RP:set poky as machine + make bitbake gcc-image? or having qemuarm as machine and doing bitbake poky? Jun 18 15:53:16 mr_nice: http://projects.o-hand.com/poky Jun 18 15:53:27 thx Jun 18 15:55:21 cool I will download now and hope it will fit my needs Jun 18 16:00:36 is bitbake able to use more than 1 pc for compiling? Jun 18 16:01:24 mr_nice: I think there is some distcc support Jun 18 16:01:30 I've never used it though Jun 18 16:01:39 compiling isn't the bottleneck Jun 18 16:02:26 RP: well on my pc it is ;) or what do you think is the bottleneck? Jun 18 16:02:54 mr_nice: It usually spends more time in configure and its the single thread nature of bitbake that slows things down Jun 18 16:03:24 RP: so there is no hope to getting it faster? Jun 18 16:03:39 mr_nice: There are plans for a multithreaded bitbake Jun 18 16:03:53 RP: :) Jun 18 16:04:33 mr_nice: icecream! Jun 18 16:05:09 zecke: where do you live? Jun 18 16:05:35 hehehe Jun 18 16:05:49 mr_nice: I have icecream in my fridge Jun 18 16:06:07 äh this one shuld goes to zecke *lol Jun 18 16:06:51 mr_nice: hehe, icecream is a special version of distcc Jun 18 16:06:59 mr_nice: it will distribute your toolchain as well :) Jun 18 16:07:12 zecke: lol Jun 18 16:07:19 mr_nice: http://en.opensuse.org/Icecream Jun 18 16:07:24 thx Jun 18 16:07:45 mr_nice: and we had a working icecream integration Jun 18 16:08:04 zecke: and I have iceccd running on my slug ;) Jun 18 16:08:39 ok I must read the page first Jun 18 16:20:15 *time* :} Jun 18 16:20:18 cya later Jun 18 16:53:04 lardman: the files are called .img, not .bin Jun 18 16:57:05 |/cr Jun 18 16:57:14 hi all Jun 18 16:57:33 hey hrw Jun 18 16:57:47 wb hrw Jun 18 16:58:02 RP: 2.6.17... Jun 18 17:00:06 mithro: do you watch soccer ATM? Jun 18 17:03:56 does icecream actually requier kde? Jun 18 17:04:00 require* Jun 18 17:04:45 emte a question for u about one-wire Jun 18 17:04:47 wb hrw Jun 18 17:04:54 hrw: 2.6.17? Jun 18 17:05:08 hey gremlin[it], your still awake :) Jun 18 17:05:20 here is 7:20 pm :) Jun 18 17:05:22 what do you want ti know? Jun 18 17:05:24 not late :) Jun 18 17:05:37 i'm looking on hx4700_battery code. Jun 18 17:05:38 RP: I saw 2.6.17 commit Jun 18 17:05:44 10:24:38 AM Jun 18 17:06:12 if i understand correct periodically all info about batteries are read also if no one from user space ask for them, right ? Jun 18 17:06:59 yes, onewire is timming sensitive DURRING a command Jun 18 17:07:22 but you can ignore it for periods of time until a command is sent Jun 18 17:07:38 mhh so is a constraint of one-wire ... cause i'm looking what i can do with h3600 ... Jun 18 17:07:52 the dallas chips are 440ms -> 720ms for a conversion Jun 18 17:07:57 RP: question is: 2.6.16 for 3.5.4.1 or 2.6.17? Jun 18 17:08:00 atleast in the temp sensors Jun 18 17:08:20 hrw: Since we've had 2.6.16 for so long, I'd say 2.6.16 Jun 18 17:08:23 mhh quite long ... i thing instead micro will reply me about imemdiatelly ... Jun 18 17:08:28 hrw: 2.6.17 hasn't been used much Jun 18 17:08:38 that's all :) Jun 18 17:08:52 (during the -rc series, it was broken and hence not used much) Jun 18 17:09:01 you/i would have to read the datasheet more on conversion/aquisition times Jun 18 17:09:27 RP: ok wanted to know Jun 18 17:09:48 but for a battery monitor there is no need to request info often Jun 18 17:10:55 emte: icecream daemon does not require Qt or KDE Jun 18 17:11:14 emte: there is a GUI to show how many clients you have, what gets built and such Jun 18 17:11:19 emte: and this GUI is a KDE application Jun 18 17:12:17 ~lart me.... for his first uint underflow... Jun 18 17:12:17 * ibot executes killall -KILL me.... for his first uint underflow... Jun 18 17:15:16 zecke, ah, the suse site mentions it for use with kde Jun 18 17:15:28 and not much else Jun 18 17:15:50 emte, yes ... thanks Jun 18 17:16:15 gremlin[it], looking at the data sheet, the chip varries for times Jun 18 17:16:27 depending on eth command you want Jun 18 17:16:53 to disable PIO you need to hold the line for 2+ seconds Jun 18 17:17:01 hrw: We could always do a quick followup of 3.5.4.2 with 2.6.17 :) Jun 18 17:17:34 emte i don't have one-wire ... Jun 18 17:18:25 i think micro convert continuosly/periodically data and give me (when i ask) the last value converted Jun 18 17:18:25 gremlin[it], i know, i expect you need to send the propper commands to the atmel which controls the onewire interface Jun 18 17:18:52 RP: :) Jun 18 17:18:58 well the device has 5 banks you would want to access Jun 18 17:18:58 emte: it used to be hosted in kdenonbeta/icecream Jun 18 17:19:10 RP: 3.5.4.2 is planed for 2.4 powered devices currently Jun 18 17:19:18 emte: and it is a SUSE 'internal' project Jun 18 17:19:30 emte: SUSE has a huge, huge icecream cluster Jun 18 17:19:43 Voltage, current, Accumulated Current, temperature, and the eeprom block Jun 18 17:19:56 as well as it's controll registers Jun 18 17:19:59 emte: basicly, every time one would use distcc, use icecream (it is based on the distcc source) Jun 18 17:20:15 heh.. chuckr thread about 'new fantastic buildsystem' is funny. fork of buildsystem used by pdaX... heh.. people never learn Jun 18 17:20:36 hrw: did people fork it? Jun 18 17:20:40 hmm Jun 18 17:20:49 well i use distcc crrently ... Jun 18 17:21:00 emte i can read only voltage and temperature ... not current or accumulated current ;( Jun 18 17:21:04 hrw: oh man, that is tough for sash... not opening it, then opening and then have people running away with it Jun 18 17:21:21 emte: icecream has a scheduler, which schedules your job to the best/fastest node Jun 18 17:21:25 gremlin[it], you should be able to access all Jun 18 17:21:35 emte: and it can distribute toolchains Jun 18 17:21:36 zecke: people could fork it before too - pdaXrom builder was released after each release Jun 18 17:21:56 hrw: right, I remember Jun 18 17:21:58 nope ... protocol SA1110<->micro give me only these 2 info Jun 18 17:22:12 I hope sash enjoys his new job (whatever that is) Jun 18 17:22:26 and chemistry of battery and AC status ... Jun 18 17:22:35 gremlin[it], i am sure that i could view that info in wince ... Jun 18 17:22:58 but you need access to the protection register at any rate Jun 18 17:23:33 zecke: I know what this job is - got offer Jun 18 17:23:51 maybe i'll have to reinstall wince this week and look again Jun 18 17:24:01 emte ... don't know :( ... have to talk with joshua maybe ... Jun 18 17:24:33 yar Jun 18 17:26:46 time for shower and call my dad Jun 18 17:27:20 c u Jun 18 17:32:10 hrw: how's the setting up of your business going? Jun 18 17:32:13 hrw: I didn't know that. We should see if we can get things into place for a 2.6 release for poodle instead :) Jun 18 17:33:16 RP: thats unofficial rather because I do not want to give any more infos so can get some time after 3.5.4.1 Jun 18 17:33:42 hrw: np, I can understand that Jun 18 17:34:29 cya later, al is here... I need to quit now Jun 18 17:34:52 someone know who is 'al'? Jun 18 17:35:26 dunno but he hangs out in #opie.de from time to time Jun 18 17:35:30 bye Jun 18 17:41:08 ~lart gstreamer and libtool Jun 18 17:41:08 * ibot DoSes gstreamer and libtool Jun 18 18:02:32 "with reference to our yesterday discussion" - how to write it in proper English? Jun 18 18:03:26 I have only libqte in my Opie image ( build with OE dev branch, htc universal), but when i try to install in example zroadmap (also OE) it ask for libqte2 and i have only libqte . How do i solve this? This is also happening with konqueror-embedded. Jun 18 18:04:01 goxboxlive: OE 'qte' build 'libqte2' package Jun 18 18:04:02 libqte2 (>= 2.3.10) Jun 18 18:04:35 hrw: So i have libqte2 then? Jun 18 18:04:44 yes Jun 18 18:05:56 hmm, but why is this happening? Jun 18 18:08:08 you got libqte_ not libqte2_ package? Jun 18 18:08:58 Where can i find that? Jun 18 18:10:06 tmp/deploy/ipk/ Jun 18 18:11:11 libqte-mt2_2.3.10-r2_htcuniversal.ipk Jun 18 18:11:54 myabe that is why the, Maybe i should edit local.conf and get rid of mt Jun 18 18:12:36 hrw: "with reference to our discussion yesterday, " Jun 18 18:12:52 koen: thx Jun 18 18:13:04 goxboxlive: so you got libqte-mt2 instead Jun 18 18:13:50 hrw: I guess i solved it now, it wasent build. I have build it now. I dont now the diffrents of it, i guess it's best to build without *mt. I seem easier. Jun 18 18:14:47 goxboxlive: hard to tell for me - I do not use .dev Jun 18 18:14:52 for qt/e stuff Jun 18 18:31:19 hi all Jun 18 18:32:42 how to find the file which defines this mystic task-bootstrap things? Jun 18 18:38:53 find . -exec grep "task-bootstrap" '{}' \; -print din't brought that much Jun 18 18:39:10 is it a bitbake specific thing? Jun 18 18:40:05 mr_nice: MACHINE config, DISTRO config Jun 18 18:40:07 task-bootstrap is rather ugly Jun 18 18:41:15 someone good at autotools? Jun 18 18:41:22 | /home/hrw/devel/build/3541/tmp/work/armv5te-linux/zbedic-1.1-r0/configure: line 1478: syntax error near unexpected token `;' Jun 18 18:41:25 | /home/hrw/devel/build/3541/tmp/work/armv5te-linux/zbedic-1.1-r0/configure: line 1478: `;' Jun 18 18:41:48 configure.ac is quite simple Jun 18 18:43:38 CoreDump|home:I sow this task- things in the files in packages/meta Jun 18 18:44:10 hrw: I cant figure out a devinition of bootstrap-image in any of the machine configs? Jun 18 18:44:28 hrw: sorry tast-bootstrap Jun 18 18:44:44 machine.conf sets variables used by task-bootstrap Jun 18 18:44:46 mr_nice: machine config define BOOTSTRAP_* stuff which is used by task-bootstrap. bootstrap-image contain only task-bootstrap (and all its dependencies) Jun 18 18:45:10 but how to find tast-bootstrap? Jun 18 18:45:35 hmm? Jun 18 18:46:09 how to find the file which parses those BOOTSTRP_* things Jun 18 18:46:25 and makes a task-bootstrap out of them Jun 18 18:47:06 task-bootstrap.bb Jun 18 18:47:09 in meta Jun 18 18:47:20 thx Jun 18 18:47:23 np Jun 18 18:48:12 I din't saw the tree cause of that wood Jun 18 18:48:24 or so Jun 18 18:48:47 the other way around ;) Jun 18 18:49:25 :) Jun 18 18:54:19 http://www.therawfeed.com/2006/06/worst-usb-gadget-yet-decapitated-teddy.html Jun 18 18:54:27 uf.. solved configure problem Jun 18 18:56:10 hrw: what was the fix? Jun 18 18:56:21 koen: removal of two semicolons Jun 18 18:56:37 9 - PKG_CHECK_MODULES(SQLITE3, sqlite3 >= 3.2 ); Jun 18 18:56:41 11 + PKG_CHECK_MODULES(SQLITE3, sqlite3 >= 3.2 ) Jun 18 19:12:40 how do I add a "-D" to CXXFLAGS in a bb file? is it possible? Jun 18 19:13:40 CXXFLAGS += "-D" ? Jun 18 19:27:26 Is there anything as a wikipedia offline reader for OE? Jun 18 19:28:32 Laibsch: OE is buildsystem not distro Jun 18 19:28:35 ~mantra Jun 18 19:28:37 mantra is probably 1) Think fubar, foobar, foo, bar, baz, boo. 2) Say out lound what you think. 3) Repeat. Jun 18 19:29:48 Laibsch: What I meant is "any bb file in OE that builds something similar to an offline reader"? I thought that was obvious. Jun 18 19:30:04 Laibsch: What I meant is "is there any bb file in OE that builds something similar to an offline reader"? I thought that was obvious. Jun 18 19:30:38 apparently not Jun 18 19:31:04 hey guys, i have a patch you might be interested in, i don't know how you handle it now in OE but, this changes timezone/Makefile so that the timezone info is installed Jun 18 19:31:11 in glibc Jun 18 19:31:46 just doing a few tests now to see if anyting gets screwed up Jun 18 19:32:13 JoeSchmo: OE bugtracker Jun 18 19:32:59 hrw, would that patch be useful to OE? Jun 18 19:33:02 koen: OK, unfortunate. Jun 18 19:35:09 JoeSchmo: can you package timezone info like it is done in packages/timezones/ recipes? Jun 18 19:37:07 i'm not using OE for this development, but i don't see why not, I don't know if you're aware that when you're cross-compiling glibc it won't install anything in /usr/share/zoneinfo, this makes a few changes so that it does, if there was a way to take /usr/share/zoneinfo on the glibc install, the yea Jun 18 19:37:42 JoeSchmo: Your post was truncated after "the yea" Jun 18 19:37:53 that was the end Jun 18 19:38:01 s/the yea/then yea/ Jun 18 19:39:02 JoeSchmo: timezones takes too much space thats why they are not installed by default Jun 18 19:39:33 koen: http://www.oesf.org/forums/index.php?showtopic=20021#entry131704 Jun 18 19:39:37 glibc in general doesn't do it when cross-compiling, becase it can't run the zic it compiles Jun 18 19:40:38 JoeSchmo: we use qemu to build locale so we can use qemu to build timezones too Jun 18 19:40:54 hrw: yeah, yet another waste of time Jun 18 19:40:59 ah, i'm trying todo both without qemu Jun 18 19:41:10 so far, i got timezones, locales are next Jun 18 19:41:40 koen: yep Jun 18 19:48:48 the patch is at, http://trac.cross-lfs.org/browser/trunk/patches/2.0/glibc-2.4-crosscompile_timezone-fix-1.patch?rev=1838&format=txt Jun 18 19:49:42 re zecke Jun 18 19:53:42 "./configure --enable-openembedded-build" looks good? Jun 18 20:05:15 "recent" bitbake tries to calculate DEPEND from RDEPEND, right? how exactly does that work? (i have a package here which seems to try building gst-plugins, probably because of gst-plugin-playbin (which comes from gst-plugins-base, or something, not gst-plugins) is in RDEPENDS. not too sure, though.) Jun 18 20:11:18 tmbinc_: IIRC. the -image specify a BUILD_ALL_DEPS=1 Jun 18 20:11:36 tmbinc_: which results BitBake in trying to build runtime deps Jun 18 20:13:25 it's set in classes/image_ipk.bbclass, yes.. Jun 18 20:13:26 but Jun 18 20:13:46 i've now removed gst-plugin-0.8 (which i don't use. yes, i should probably rename them to gstreamer0.10 but that's antoher thing) Jun 18 20:14:26 and now i get "Nothing provides runtime dependency gst-plugin-decodebin" Jun 18 20:14:51 however, that package exists - gst-plugin-something has already built it Jun 18 20:15:06 so how do i tell bitbake that not gst-plugin but gst-plugin-something builds gst-plugin-decodebin? Jun 18 20:16:37 | obj/moc_editor.o(.gnu.linkonce.r._ZTI12ZBEDicEditor+0x8): undefined reference to `typeinfo for QWidget' Jun 18 20:16:40 ;( Jun 18 20:16:56 zecke: ideas? Jun 18 20:17:15 problem is i don't know in advance which packages are built by a specific provider. is that a problem? Jun 18 20:18:32 where is opensimpad+64 definded? Jun 18 20:20:13 oh it is called opensimpad-64+0 sorry Jun 18 20:20:53 mr_nice: its kernel so packages/linux/ Jun 18 20:21:33 thx Jun 18 20:25:39 include opensimpad_${PV}.bb where is PV defined? - getting into a never ending storry. So I will take a long storry short. I will figure out what the bitbake is doing with the kernel how it patches and compiles it etc. Jun 18 20:27:22 mr_nice: opensimpad-64+0_2.4.27-vrs1-pxa1-jpm1.bb has PV="2.4.27-vrs1-pxa1-jpm1" Jun 18 20:27:33 mr_nice: recipe is: PN_PV.bb Jun 18 20:27:51 mr_nice: do "bitbake -e virtual/kernel|less" Jun 18 20:28:11 thx Jun 18 20:33:26 np Jun 18 20:36:42 mr_nice: PN and PV are read from the filename, but can be set from within the bb file Jun 18 20:38:55 zecke PN is constant and PV is varable? Jun 18 20:39:07 mr_nice: all are variable Jun 18 20:39:18 mr_nice: just for convience they are read from the filename Jun 18 20:39:18 ah, ok Jun 18 20:39:27 mr_nice: we could have 1.bb, 2.bb, 3.bb Jun 18 20:41:40 zecke: good evening Jun 18 20:42:17 has PV something to do with PREFERRED_VERSION ? i Jun 18 20:43:35 mr_nice: yes Jun 18 20:43:44 mr_nice: IIRC PV+PR Jun 18 20:45:12 ok Jun 18 20:52:25 tmbinc_: That looks like the dynamic package naming code is causing problems. We probably need to rethink those specific package named Jun 18 21:08:50 heh.. zbedic 1.1 contain 8K patch to configure/makefile.... Jun 18 21:08:57 and thats not finished Jun 18 21:09:23 =D Jun 18 21:11:32 I just built it (first time) Jun 18 21:11:40 whole recipe is rewritten Jun 18 21:12:28 patches == 27,401 bytes in 6 files Jun 18 21:12:51 but 17K is 1.1-1.diff from upstream which should help to build it Jun 18 21:14:32 Its a sad state of affairs when I'm surprised a given patch is in -mm as I can't remember submitting it :) Jun 18 21:14:47 ;P Jun 18 21:16:10 Of the remaining patches, only 4 are ready for unstream. Not a good position :-/ Jun 18 21:24:37 how much left? Jun 18 21:26:52 * hrw goes sleep Jun 18 21:26:56 good night guys Jun 18 21:27:38 n8 hrw|gone Jun 18 21:30:18 'night hrw|gone Jun 18 21:30:46 n8 doesn't work... Jun 18 21:30:55 * RP shakes head sadly ;-) Jun 18 21:32:25 good nite Jun 18 21:42:07 'night Jun 18 22:02:12 RP: I am trying to get OE to work on my Fedora 4 Machine. I added ENABLE_BINARY_LOCALE_GENERATION = "0" to my local.conf, but glibc is still trying to build the locale data. Any Ideas? Jun 18 22:03:30 hvontres: Which DISTRO? Jun 18 22:04:37 openzaurus-3.5.4.1 Jun 18 22:06:01 that flag has littel to do with glibc Jun 18 22:06:10 its for qemu Jun 18 22:06:22 hvontres: You need to edit conf/distro/openzaurus-3.5.4.1.conf to set it due to the way its been setup for oz3541 Jun 18 22:06:38 thanks. Jun 18 22:07:13 belatedly, there is in fact no added cherry-picker support in 0.27. maybe 0.28. Jun 18 22:09:33 yay! Jun 18 22:09:36 0.28! Jun 18 22:09:45 err? Jun 18 22:09:58 cherry picker in 0.28! Jun 18 22:10:11 man, you guys are going to crucify me if I don't get it in, aren't you :-) Jun 18 22:11:32 you should just talk to JustinP and use his script :P Jun 18 22:14:38 njs: yes! Jun 18 22:15:36 njs: and I'd like a "this patch is from Johny Foo, but I committed it for him" support in monotone Jun 18 22:20:31 g'night Jun 18 22:24:17 03mickeyl 07org.oe.dev * r82f4e07e... 10/packages/sysfsutils/ (sysfsutils_1.1.0.bb sysfsutils_1.2.0.bb sysfsutils_2.0.0.bb): sysfsutils: remove default preference for v2.0.0, remove some obsolete ones Jun 18 22:24:22 03mickeyl 07org.oe.dev * r7312b630... 10/packages/lm_sensors/ (3 files in 2 dirs): lm_sensors: add 2.10.0, reformat 2.9.2 per style guideline demands Jun 18 22:52:46 mtn sync xmpp://njs@jabber.org <---- sound cool Jun 18 22:59:50 03rpurdie 07org.oe.dev * r013572ec... 10/packages/linux/ (10 files in 3 dirs): linux-oz-2.6.16+git/2.6.17: Further improve poodle patches. Fix hx2750 configure/compile failures. Further comment patches to reflect status. Jun 18 23:10:55 * chouimat is away: stuff and Zzzz Jun 18 23:12:02 "this patch is from Johny Foo, but I committed it for him" = "commit --author johny.foo@fooworld.com"? Jun 18 23:12:14 chouimat: send a patch :-) Jun 18 23:15:28 njs: :) Jun 19 00:36:10 is there something I can put in a .bb to keep the binaries/libs from getting stripped as part of the package process Jun 19 00:36:12 ? Jun 19 00:41:32 only in local.conf Jun 19 00:41:45 INHIBIT_PACKAGE_STRIP = "1" or something like that Jun 19 00:41:58 local.conf.sample should have an example Jun 19 00:42:33 thanks **** ENDING LOGGING AT Mon Jun 19 02:59:56 2006