| 00:00.22 | pb_ | it does sound like that might be the case |
| 00:01.20 | reenoo_ | getting my laptop back might improve things already |
| 00:01.53 | reenoo_ | I'm wasting a lot of time between lectures/tutorials |
| 00:02.43 | pb_ | ah, right |
| 00:03.34 | reenoo_ | with a laptop it's easier to flee from the inevitably "reenoo_, you know UNIX, don't you? can you help me with..." question you get in computer rooms |
| 00:03.43 | pb_ | heheh |
| 00:07.34 | reenoo_ | another thing is that you normally work in groups of three while I mostly work alone after my group has "fallen apart" at the end of last semester |
| 00:07.45 | reenoo_ | anyways |
| 00:07.54 | reenoo_ | 'night all |
| 00:07.55 | pb_ | reenoo_: night |
| 00:08.33 | CIA-5 | 03mickeyl 07 * r1.1385.1.1 10packages/gpe-session-scripts/gpe-session-scripts_0.61.oe: add xtscal to gpe-session-scripts depends - patch courtesy pb_.\ |
| 00:08.33 | CIA-5 | 03mickeyl 07 * r1.1387 10packages/: |
| 00:08.33 | CIA-5 | Merge bk://openembedded@openembedded.bkbits.net/packages\ |
| 00:08.33 | CIA-5 | into gandalf.tm.informatik.uni-frankfurt.de:/usr/local/projects/packages\ |
| 00:28.13 | pb_ | gah. I guess I'm gonna have to give in and fix readline. |
| 00:40.28 | CIA-5 | 03mickeyl 07 * r1.1388 10packages/autoconf/ (2 files in 2 dirs): make autoheader missing-template warnings non fatal - patch courtesy pb_.\ |
| 00:45.18 | CIA-5 | 03mickeyl 07 * r1.1389 10packages/gpe-timesheet/gpe-timesheet_0.14.oe: update gpe-timesheet to 0.14 - patch courtesy pb_.\ |
| 00:53.30 | gb2 | all these rejected suggestsions i had seem to be going in. |
| 00:53.48 | kergoth | ? |
| 00:54.05 | gb2 | pb's COMPATIBLE_HOST check |
| 00:54.38 | gb2 | there was something else too |
| 00:55.19 | gb2 | ah, the base_read_file() function at the top level |
| 00:55.25 | kergoth | ohh that. i disagreed with that being the only methodology for a package to prevent itself from building. as long as a package who so chooses can use an anonymous function to exert more control, i dont see anything wrong with an easier one |
| 00:55.32 | kergoth | whats base_read_file? |
| 00:55.45 | gb2 | http://openembedded.bkbits.net:8080/oe/cset@1.544?nav=index.html|ChangeSet@-3d |
| 00:55.50 | kergoth | ah |
| 00:56.00 | chouimat|tired | kergoth: know Ephel Duath? |
| 00:56.06 | kergoth | nope |
| 00:56.19 | chouimat|tired | kergoth: kind of mix between Jazz and Death Metal |
| 00:56.28 | kergoth | ha, sounds cool |
| 01:01.19 | *** join/#oe Neil (672bf@lsanca1-ar51-4-42-017-130.lsanca1.dsl-verizon.net) |
| 01:03.12 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 01:03.25 | kergoth | fuck. |
| 01:03.28 | kergoth | tomorrow is monday again |
| 01:11.52 | chouimat|tired | kergoth: yup and I need to call the movers, the phone co and the cable co and pick a date next week for moving |
| 01:18.29 | *** join/#oe a1abubba (~rjh@68.113.102.123) |
| 01:36.02 | gb2 | mark: ooooh |
| 01:36.54 | kergoth | gb2: you missed. |
| 01:36.58 | gb2 | yeah |
| 01:36.59 | gb2 | i did |
| 01:37.08 | gb2 | i seem to only do it while trying to talk to mark |
| 01:37.09 | gb2 | weird |
| 01:43.21 | *** join/#oe glennm (~glennm@203-217-29-88.perm.iinet.net.au) |
| 01:44.39 | bug1 | if i press the touchscreen for more htan 3 seconds, Xfbdev will start consuming heaps of CPU resources (>70%) untill i press it down for another 3 seconds. |
| 01:44.56 | bug1 | Does X have some struange button state ? |
| 01:45.17 | bug1 | like if you hold it down it goes into a different mode ? |
| 02:04.06 | *** join/#oe ||ugh (~hugh@216.240.42.42) |
| 02:12.19 | *** join/#oe _stigger (~jeff@203.41.111.227) |
| 02:29.03 | *** join/#oe Ptonyb (~root@DSL135-071.LABridge.com) |
| 02:34.37 | *** part/#oe a1abubba (~rjh@68.113.102.123) |
| 02:56.44 | kergoth | bored. |
| 02:59.46 | Pendalar | drink ;-) |
| 03:00.18 | kergoth | i am, but i'm still bored |
| 03:02.52 | Pendalar | I hope this goes better |
| 03:02.57 | Pendalar | er this build |
| 03:03.37 | Pendalar | last 3 builds I got like 33 builds attempted, 3 failed.. #### dependancies unsatisfied |
| 03:28.55 | chouimat|Zzzzzzz | night |
| 03:51.14 | *** join/#oe Disconnect (~dis@3ffe:bc0:91:0:203:6dff:fe1a:c5b6) |
| 03:58.08 | Pendalar | kergoth: are there 2 OE mailing lists? |
| 03:58.15 | kergoth | no |
| 03:58.28 | Pendalar | alright, because I kept looking for a -devel just in case |
| 03:58.39 | kergoth | nope, oe@ is it |
| 03:58.53 | Pendalar | thanks =-) that's all I was finding at handhelds.org anyways |
| 03:58.59 | kergoth | np |
| 04:00.13 | Pendalar | now if I can get my ups working right with linux, it'll be a good night |
| 04:02.31 | Pendalar | .. I have to know my cable version? wtf |
| 04:02.43 | Pendalar | no wonder I gave up messing with this a while back =-. |
| 04:34.52 | *** join/#oe treke|laptop (~konversat@lsanca2-ar29-4-43-192-190.lsanca2.elnk.dsl.genuity.net) |
| 04:34.58 | treke|laptop | back in X again |
| 04:35.05 | treke|laptop | using the x.org server |
| 04:35.18 | Pendalar | isn't that a song? |
| 04:35.25 | Pendalar | back in the X agaaaain |
| 04:35.42 | kergoth | ~fishslap Pendalar |
| 04:35.48 | ibot | ACTION slaps Pendalar up side the head with a wet fish. |
| 04:36.11 | Pendalar | mmm fish |
| 04:36.26 | treke|laptop | works pretty well, though the colors seem more washed out |
| 04:37.58 | treke|laptop | also using kde-cvs now |
| 04:43.03 | treke|laptop | hmmm. xgamma -gamma 1.0 fixed it rather nicely :) |
| 04:43.09 | kergoth | heh |
| 04:43.45 | treke|laptop | it was set to something like 2.0 |
| 04:45.18 | gb2 | treke: i ended up with weird fonts in x.org |
| 04:45.48 | treke|laptop | they actually look pretty good here |
| 04:49.29 | treke|laptop | wonder if they are going to keep dri in sync with it |
| 04:51.15 | gb2 | treke: for some reason, gkrellm can't use anything but 'fixed' |
| 04:51.20 | treke|laptop | ah |
| 04:51.30 | gb2 | nomatter what i set for its fonts i get fixed in the actual display |
| 04:51.35 | treke|laptop | odd |
| 04:51.36 | gb2 | could be a gkrellm bug |
| 04:52.04 | gb2 | this happened update from fc1 -> fc2t3 |
| 04:52.39 | treke|laptop | could be a gtk bug |
| 04:54.43 | gb2 | treke: possibly |
| 04:54.57 | gb2 | i only see the issue in gkrellm sofar |
| 04:55.21 | treke|laptop | use many gtk 1.2 apps? |
| 04:56.28 | gb2 | prolly not |
| 04:57.09 | gb2 | but gkrellm is a gtk2 app |
| 04:57.16 | treke|laptop | oh they updated it? |
| 04:57.21 | treke|laptop | thought it was a dead app |
| 04:57.22 | gb2 | maybe not entirely :) |
| 04:57.53 | gb2 | i've stopped running it mostly anyway. it has a few 2.6 issues |
| 04:58.01 | gb2 | mostly cosmetic |
| 04:58.12 | treke|laptop | I gave up on fedora test when the only app I ran all day long was broken :) |
| 04:58.20 | gb2 | what app? |
| 04:58.23 | treke|laptop | rhytmbox |
| 04:58.26 | gb2 | heh |
| 04:58.37 | gb2 | i only listen to music on my iRiver player now |
| 04:59.25 | treke|laptop | the portable mp3/ogg players seem to be catching on |
| 04:59.53 | gb2 | it seems to get 16h on battery |
| 05:00.04 | gb2 | lasts me about a week and a half worth of commutes |
| 05:00.07 | kergoth | went to best buy the other day to see what they had for players. what a joke. the only high capacity players they had were ipods. thats it. |
| 05:01.03 | gb2 | kergoth: expensive for anything too |
| 05:01.07 | gb2 | except ipods |
| 05:01.10 | gb2 | which are price fixed |
| 05:01.11 | kergoth | yeah |
| 05:02.52 | gb2 | http://super-tux.sourceforge.net/screenshots.html |
| 05:02.56 | gb2 | i wonder if this is any fun |
| 05:10.17 | bug1 | kdrive has a press and release state for mouse buttons, anyone know how that relates to touchscreens ? |
| 05:10.56 | *** part/#oe treke|laptop (~konversat@lsanca2-ar29-4-43-192-190.lsanca2.elnk.dsl.genuity.net) |
| 05:11.02 | *** join/#oe treke|laptop (~konversat@lsanca2-ar29-4-43-192-190.lsanca2.elnk.dsl.genuity.net) |
| 05:49.17 | Pendalar | 00:03:41 (150.59 KB/s) - `/home/pendalar/devel/openembedded/oe_downloads/nova-pci216.exe' saved [13581740/13581740] |
| 05:49.17 | Pendalar | warning [/home/pendalar/devel/openembedded/oe_downloads/nova-pci216.exe]: 30038 extra bytes at beginning or within zipfile |
| 05:49.25 | Pendalar | qua? |
| 05:51.47 | Pendalar | you guys are sick =-) |
| 05:51.54 | Pendalar | g'night |
| 05:52.15 | treke|laptop | ? |
| 05:52.32 | treke|laptop | unzip nova-pci216.exe |
| 06:10.09 | *** join/#oe kurre2 (~tomimo@ncircle.nullnet.fi) |
| 06:14.12 | *** part/#oe kurre2 (~tomimo@ncircle.nullnet.fi) |
| 06:16.58 | *** join/#oe Toi|work (~pleemans@oia-gw1.be.jnj.com) |
| 06:18.16 | *** join/#oe kurre2 (~tomimo@ncircle.nullnet.fi) |
| 07:06.30 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 07:58.12 | *** join/#oe a-atwood1home (~xf009976@sdn-ap-030txhousP0245.dialsprint.net) |
| 08:11.39 | *** join/#oe _stigger (~jeff@203.49.17.76) |
| 08:29.41 | CIA-5 | 03mickeyl 07 * r1.1390 10packages/xtscal/xtscal_0.4.oe: depend on libxrandr, not xrandr - patch courtesy pb_.\ |
| 08:29.43 | *** join/#oe pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) |
| 08:30.46 | pb_ | morning |
| 08:37.01 | pb_ | hi mickey |
| 08:39.00 | mickeyl | morning bp |
| 08:39.01 | mickeyl | argh |
| 08:39.02 | mickeyl | sorry |
| 08:39.03 | mickeyl | morning pb_ |
| 08:39.08 | pb_ | heh |
| 08:39.28 | mickeyl | note to self: don't type prior to the first cofee |
| 08:39.32 | mickeyl | coffee even |
| 08:41.33 | CIA-5 | 03mickey 07 * r1.1391 10packages/linux/ (3 files in 2 dirs): upgrade openzaurus kernel to 2.6.5-jl3 |
| 08:44.16 | *** join/#oe zmiq (~chatzilla@45.Red-213-98-141.pooles.rima-tde.net) |
| 08:44.33 | CIA-5 | 03mickey 07 * r1.1392 10packages/linux/openzaurus-2.6.5-jl3/defconfig-collie: update 2.6.5-jl3 defconfig for collie |
| 08:49.26 | *** join/#oe zmiq (~chatzilla@45.Red-213-98-141.pooles.rima-tde.net) |
| 08:54.19 | *** join/#oe robilad[away] (~topic@mpiat2304.ag2.mpi-sb.mpg.de) |
| 08:59.12 | *** join/#oe sinterklaas (~sinterkla@hoc.xs4all.nl) |
| 09:06.40 | pb_ | hm. so much for my plan to work in the garden today. |
| 09:07.19 | mickeyl | sun here |
| 09:07.37 | pb_ | heh |
| 09:07.37 | mickeyl | no office today for you? |
| 09:07.45 | pb_ | no, it's a public holiday here in England |
| 09:07.49 | mickeyl | ah |
| 09:08.23 | mickeyl | i never visited england. one of these days (probably next year) i'm trying to come over to great britian. in what area are you located? |
| 09:08.31 | pb_ | Cambridge |
| 09:08.45 | pb_ | just about 100km north of London |
| 09:10.15 | *** join/#oe Hrw (~szczepan@c53-11.icpnet.pl) |
| 09:10.18 | Hrw | hi |
| 09:10.58 | mickeyl | hi hrw |
| 09:12.18 | Hrw | I updated my OE repo and found one thing: gcc-cross now depends on libc-for-gcc ;( do I have to rebuild it again? only glibc-intermediate (CVS) provides it |
| 09:13.11 | pb_ | that's odd. regular glibc should provide libc-for-gcc so long as you don't have NPTL selected. |
| 09:13.22 | pb_ | PROVIDES = "virtual/libc ${@['virtual/${TARGET_PREFIX}libc-for-gcc', '']['nptl' in '${LIBC_ADDONS}']}" |
| 09:13.33 | pb_ | I guess you have to figure out why this isn't working for you. |
| 09:13.37 | Hrw | ok |
| 09:14.03 | Hrw | pb_: glibc_cvs has it but I used glibc_2.3.2 |
| 09:14.10 | pb_ | ah, that would be a problem. |
| 09:15.08 | Hrw | actually I removed this dependency from gcc-cross 3.4.0 |
| 09:20.25 | pb_ | that's ok as a workaround, but please do not check that change in. |
| 09:21.01 | pb_ | if you want to make a patch to fix up the PROVIDES in glibc_2.3.2, that would be fine. |
| 09:21.02 | Hrw | anyone tried to compile openobex-apps with gcc 3.4? |
| 09:21.17 | Hrw | OK I'll do |
| 09:22.39 | *** join/#oe robilad (~topic@mpiat2304.ag2.mpi-sb.mpg.de) |
| 09:24.00 | pb_ | mickeyl: I put a workaround in place for that readline autoheader thing. |
| 09:24.34 | pb_ | so I hope the sqlite/ncurses dependency chain will be buildable again now. |
| 09:25.46 | Hrw | pb_: http://www.hrw.one.pl/tmp/glibc.patch |
| 09:26.00 | mickeyl | pb_: ah nice. thanks. |
| 09:26.05 | Hrw | I had sqlite builded few days ago |
| 09:26.25 | mickeyl | yes it was one of the cases where it works on some systems and on some not |
| 09:26.52 | pb_ | oh, and that fetch.py bug that was causing problems with the opie tarballs should be fixed as well. |
| 09:27.37 | Hrw | pb_: I'm thinking about writing script to create all needed opie tarballs from checkouted cvs |
| 09:27.47 | mickeyl | pb_: hey, that's great news. thanks. checked it in yet? |
| 09:27.58 | pb_ | Hrw: how is that different to what OE does already? |
| 09:28.02 | pb_ | mickeyl: yep, last night |
| 09:28.08 | mickeyl | rock |
| 09:28.23 | Hrw | pb_: I didn't tried last fixes yet |
| 09:28.25 | pb_ | http://openembedded.bkbits.net:8080/oe/cset@1.552?nav=index.html|ChangeSet@-1d is the patch you need |
| 09:28.56 | pb_ | Hrw: I mean in conceptual terms. What are you trying to do that's different from OE? |
| 09:31.06 | CIA-5 | 03mickeyl 07 * r1.1390.1.1 10packages/glibc/glibc_2.3.2.oe: add libc-for-gcc to glibc_2.3.2 PROVIDES - patch courtesy pb_.\ |
| 09:31.06 | CIA-5 | 03mickeyl 07 * r1.1393 10packages/: |
| 09:31.06 | CIA-5 | Merge bk://openembedded@openembedded.bkbits.net/packages\ |
| 09:31.06 | CIA-5 | into gandalf.tm.informatik.uni-frankfurt.de:/usr/local/projects/packages\ |
| 09:32.16 | Hrw | pb_: checkout opie from cvs, get cvs:// lines from opie*.oe, create archives in DLDIR |
| 09:32.39 | pb_ | I still don't really understand the difference. |
| 09:33.23 | pb_ | Can you assume I know nothing about opie (which isn't far from the truth), and explain with small words why the fetcher doesn't do the right thing? |
| 09:33.34 | Hrw | pb_: as I said I didn't tried fixed cvs fetcher - old one has problems with fetching all needed subdirs so I had to create some of them by hand |
| 09:33.41 | mickeyl | just try the fixed one |
| 09:33.44 | Hrw | ok |
| 09:33.59 | mickeyl | if it works, you can save the time writing a script to add some opie packages ;) |
| 09:34.13 | pb_ | heh, good plan |
| 09:34.34 | Hrw | mickeyl: btw - "cvs update -A -P" is much faster then "cvs checkout" |
| 09:35.00 | pb_ | that's true, especially for big modules like libc. |
| 09:35.23 | pb_ | maybe you'd like to write a patch for the fetcher to keep the cvs checkout around, and just do incremental updates |
| 09:35.30 | *** join/#oe robilad (~topic@mpiat2304.ag2.mpi-sb.mpg.de) |
| 09:35.45 | mickeyl | yeah that'd be interesting |
| 09:36.00 | pb_ | another option would be to teach the fetcher to go look at treke.net to see if there is an already-downloaded tarball with the right CVSDATE. |
| 09:37.46 | *** join/#oe a-atwood_home (~xf009976@sdn-ap-030txhousP0245.dialsprint.net) |
| 09:41.41 | pb_ | mickeyl: btw, do you know how I get a login for bugs.openembedded.org? |
| 09:41.59 | pb_ | there doesn't seem to be any kind of "new account" facility on the page. |
| 09:42.53 | CIA-5 | 03mickeyl 07 * r1.1394 10packages/gpe/meta-gpe.oe: add more stuff to gpe DEPENDS - patch courtesy pb_.\ |
| 09:43.08 | mickeyl | pb_: only treke and kergoth can do it |
| 09:43.14 | pb_ | aha |
| 09:43.35 | pb_ | okay. I guess I'll have to wait for the sun to rise over Texas. |
| 09:43.57 | mickeyl | heh |
| 09:44.04 | mickeyl | we should pester them to bring the bug tracker in shape |
| 09:44.18 | mickeyl | if they want people to use it, make it work :D |
| 09:44.21 | pb_ | yah, quite. |
| 09:45.01 | mickeyl | do users use the gpe bugtracker? |
| 09:45.16 | pb_ | yup |
| 09:45.40 | pb_ | we just use the handhelds.org bugzilla. Some people do grumble about its interface, but most users seem to tolerate it. |
| 09:45.42 | mickeyl | do they use it in a sane way? e.g. distinguishing packaging, gui and base system issues? |
| 09:46.08 | pb_ | not usually. I think that's too much to expect from regular users. |
| 09:46.53 | pb_ | The interactions between different parts of the system are complex enough that it often isn't at all obvious which category a particular problem falls into. |
| 09:47.22 | mickeyl | yeah, i guess there's nothing to do about that |
| 09:47.32 | mickeyl | educating users is hard :) |
| 09:49.47 | pb_ | yup :-) and, to be honest, I think it sends the wrong message if we expect or require users to have intimate knowledge of what goes on under the covers. I'd much rather just let them file bug reports from a high-level perspective ("application XYZ crashes when I do this!") and have the developers or distribution people triage the reports appropriately. |
| 09:52.16 | pb_ | The problem we have really is that there's no easy way to move bugs between the different databases. So, for example, if a user files a bug against Familiar, and it turns out to be an Opie issue, there is no way for me to punt it from bugzilla into Mantis to have the opie dudes look at it. |
| 09:53.03 | mickeyl | yes, i see this also as one of the main problems |
| 09:53.22 | mickeyl | i think i could pester the opie folks to use the OE RT as soon as we have OE based releases out |
| 09:53.56 | pb_ | right, that would help. |
| 09:54.09 | *** join/#oe zinga (~arnaudb@186.80-203-227.nextgentel.com) |
| 09:54.34 | mickeyl | yes, mantis _looks_ more friendly. then again, someone could just change the RT stylesheet to make it look more friendly |
| 09:55.08 | pb_ | personally I would prefer to see everybody using bugzilla, but still :-) |
| 09:56.27 | mickeyl | heh, yeah. i like mantis myself, but i also see the necessary to use a common one to be more important than my personal preference :) |
| 09:57.45 | pb_ | right. my preference for bugzilla mostly comes from the fact that lots of big projects like GNOME, KDE, GCC and Mozilla all use it, so users are likely to be relatively familiar with it already and we can benefit from any changes that the guys in those other projects might make. |
| 09:58.11 | pb_ | Hrw: cool |
| 09:58.29 | mickeyl | Hrw: great. now let the bugfixes roll |
| 09:58.41 | Hrw | nearly all packages from OZ 3.3.6pre1 found their replacement |
| 09:59.25 | Hrw | PATH is wrong, /etc/init.d/networking has wrong permissions |
| 09:59.32 | mickeyl | still? |
| 09:59.36 | mickeyl | i thought someone changed that |
| 09:59.48 | pb_ | me too. maybe we both hallucinated it. |
| 09:59.51 | pb_ | what's wrong with PATH? |
| 09:59.55 | Hrw | no /sbin |
| 10:00.15 | pb_ | for root? I thought that was fixed too. |
| 10:00.24 | pb_ | do you have a .profile in /root? |
| 10:00.30 | mickeyl | odd. you built everything from scratch, Hrw? |
| 10:00.37 | Hrw | mickeyl: YES |
| 10:01.08 | Hrw | http://www.hrw.one.pl/tmp/hrw-image.oe used |
| 10:01.55 | pb_ | I think task-bootstrap should be taking care of most of the stuff you have in BASE_SYSTEM. |
| 10:02.12 | mickeyl | interesting |
| 10:02.17 | pb_ | and meta-opie likewise for OPIE_LIBS and OPIE_BASE. |
| 10:02.22 | pb_ | if that's not happening, we should find out why. |
| 10:03.28 | Hrw | I got all from /usr/lib/ipkg from my Z and started searching for replacements from OE |
| 10:03.28 | pb_ | right |
| 10:03.37 | pb_ | it sounds like you're really trying to produce opie-image. |
| 10:03.56 | pb_ | maybe you'd like to make some patches to add the missing stuff to the appropriate metapackages so that it ends up in there. |
| 10:04.09 | mickeyl | i already added more stuff to opie-image |
| 10:04.20 | mickeyl | no idea if i pushed that |
| 10:04.28 | mickeyl | i don't like specifying the modules there |
| 10:04.35 | mickeyl | where is the appropriate place for the modules? |
| 10:04.45 | mickeyl | since they're devce specific in MACHINE.conf ? |
| 10:04.47 | pb_ | meta-opie, in most cases |
| 10:04.49 | Hrw | pb_: new file doesn't interfere with bitkeeper |
| 10:04.54 | pb_ | oh, sorry, the modules |
| 10:05.02 | pb_ | yeah, EXTRA_BOOTSTRAP in machine.conf |
| 10:05.05 | Hrw | mickeyl: I think that MACHINE.conf will be best |
| 10:05.11 | mickeyl | ok, good. let's make it so |
| 10:05.17 | pb_ | Hrw: pardon? |
| 10:06.28 | Hrw | pb_: I use bitkeeper just like cvs - don't have time to look into docs for more then "pull, co, sccslog, diffs, edit, unedit" |
| 10:07.08 | pb_ | heh. well, I don't use bitkeeper at all. |
| 10:07.28 | Hrw | pb_: so when I realized that making image needs adding more into it that it has I created copy of it |
| 10:08.29 | *** join/#oe robilad (~topic@mpiat2304.ag2.mpi-sb.mpg.de) |
| 10:08.54 | Hrw | hmm.. no depmod in busybox ;( |
| 10:09.13 | mickeyl | didn't we enable it yet? |
| 10:09.33 | pb_ | Hrw: sure, creating a new .oe is fine, but it would be even better if you worked on improving the existing metapackages so that other people could benefit from that too. |
| 10:10.13 | Hrw | pb_: I work in this style: 1st get it working, then check what to fix after it |
| 10:11.04 | Hrw | pb_: I have now 3day long weekend, my girlfriend didn't like when I spend this time with comp instead of her ;) |
| 10:11.18 | pb_ | heh |
| 10:15.59 | mickeyl | how should i decide what goes into RDEPENDS and RRECOMMENDS? |
| 10:16.18 | mickeyl | e.g. e2fsprogs and glibc-gconv-iso8859-1 ? |
| 10:17.33 | Hrw | that's what I had to ask too |
| 10:17.44 | mickeyl | well, ok let's say RDEPENDS is what is needed to get it booting to a command prompt |
| 10:17.56 | mickeyl | RRECOMMENDS all what is needed to have a sane bootstrap image |
| 10:19.13 | Hrw | mickeyl: you are editing conf/collie.conf and meta/*? |
| 10:19.27 | mickeyl | Hrw: can you do it? |
| 10:19.37 | mickeyl | (currently busy with "real" work) |
| 10:19.45 | mickeyl | i can take care about the collie.conf |
| 10:19.56 | Hrw | mickeyl: I started but it would need fixes when I finish |
| 10:20.27 | Hrw | I'm not good in base system building |
| 10:20.31 | mickeyl | no problem |
| 10:20.36 | Hrw | thx |
| 10:20.53 | Hrw | I edited collie.conf (added kernel modules from OZ 3.3.6p1) |
| 10:21.07 | mickeyl | good strategy |
| 10:21.23 | mickeyl | for now we try to recreate oz 3.3.6 to get something like oz 3.5.1 out of the door |
| 10:21.30 | mickeyl | then while moving to oz 3.4 we improve |
| 10:21.36 | mickeyl | s/3.4/3.6 |
| 10:21.39 | Hrw | to task-bootstrap into DEPENDS added all packages which are needed in base system |
| 10:22.09 | pb_ | mickeyl: both RDEPENDS and RRECOMMENDS are installed during image construction. The difference is that ipkg will allow the user to remove a package in RRECOMMENDS, but not one in RDEPENDS. |
| 10:22.24 | Hrw | mickeyl: help to get oz 3.5/3.6 is thing for which I made this work ;) |
| 10:23.25 | pb_ | so, things like drivers for CF cards should probably go in RRECOMMENDS, so that people who don't have that hardware can uninstall the drivers. |
| 10:24.06 | mickeyl | pb_: ah that make sense |
| 10:24.23 | Hrw | apmd is needed or optional? |
| 10:24.40 | pb_ | needed. |
| 10:24.59 | pb_ | I don't think suspend/resume will work without it. |
| 10:25.47 | Hrw | if task-bootstrap DEPENDS on nearly all which is in RDEPENDS can I create one list of packages (NEEDED) and set DEPENDS = "${NEEDED} ${BOOTSTRAP_EXTRA_DEPENDS}"? |
| 10:26.41 | pb_ | yes, though some care may be needed: not all packages have identical naming for the .oe and the .ipk. |
| 10:27.00 | mickeyl | sure. additional semantics is ok, however check the naming differences |
| 10:27.06 | mickeyl | oh |
| 10:27.11 | mickeyl | like pb said |
| 10:27.15 | Hrw | ok |
| 10:27.20 | Hrw | I'll look at it |
| 10:27.38 | pb_ | in particular, remember that all library packages will be named differently under DEBIAN_NAMES (which is automatically set for DISTRO=familiar) |
| 10:29.01 | pb_ | I keep meaning to add a facility to oemake to let you specify just the RDEPENDS for a metapackage, and have it figure out the corresponding DEPENDS for itself. |
| 10:29.30 | pb_ | It won't be able to do that in all cases, but probably 90% of packages will be ok. |
| 10:29.31 | Hrw | few packages I was unable to get build: modutils (busybox doesn't have depmod implemented), openobex-apps (have it on Z) |
| 10:30.24 | mickeyl | pb: should we enable busybox-depmod? |
| 10:30.45 | pb_ | mickeyl: does busybox even include depmod? I didn't think it did. |
| 10:31.11 | Hrw | mickeyl: busybox doesn;t have depmod |
| 10:31.14 | pb_ | oh, I remember. it does provide an example depmod, but it's written in Perl. |
| 10:31.34 | mickeyl | oh. ok, let's forget about that then |
| 10:31.39 | pb_ | we need to build depmod from modutils (or module-init-tools for 2.6) |
| 10:31.49 | mickeyl | how about init? busybox contains an init. can we use that instead of sysvinit? |
| 10:32.07 | pb_ | not sure. I think busybox init is too simple-minded, but I haven't tried it myself. |
| 10:32.17 | pb_ | Chris would know about that. I think he played with busybox init in OZ at one time. |
| 10:35.20 | mickeyl | k |
| 10:50.49 | Hrw | mail sent to OE ML |
| 10:51.55 | Hrw | bye guys |
| 10:53.06 | pb_ | "I used OE from Thursday" |
| 10:53.17 | pb_ | doh, I guess that's why he was having all these problems that we thought were fixed. |
| 10:56.51 | CIA-5 | 03mickeyl 07 * r1.553 10oe/conf/collie.conf: apply collie.conf patch from Marcin Juszkiewicz - patch courtesy pb_.\ |
| 10:58.12 | pb_ | mickeyl: do you what sysv-rc is? |
| 10:58.19 | pb_ | do you know, rather |
| 11:03.31 | *** join/#oe sinterklaas (~sinterkla@hoc.xs4all.nl) |
| 11:05.54 | sinterklaas | hi, is oeinstall finctional ? I'm getting strange error's |
| 11:06.34 | pb_ | not sure. I think most people here are not using oeinstall. |
| 11:07.00 | sinterklaas | ok, do you manually install the ipk's or the tarballs ? |
| 11:07.14 | pb_ | familiar-style distros do image construction using image_ipk.oe. |
| 11:08.04 | pb_ | it might be interesting to see if we can refactor image_ipk.oe in terms of oeinstall at some point. |
| 11:08.48 | sinterklaas | oeinstall seems nice, expecially for use in 'custom' embedded devices |
| 11:12.22 | pb_ | yeah |
| 11:12.36 | pb_ | image_ipk.oe does more or less the same thing, if I remember right. |
| 11:13.09 | pb_ | obviously it only works for .ipk, though, so you need a different solution if you are trying to deploy from tarballs. |
| 11:13.42 | sinterklaas | I prefer ipk's |
| 11:14.51 | sinterklaas | I'm trying to 'sell' oe to my boss, we're making more and more linux based (arm)devices, but I'm a little worried about conf. managment, any ideas about that ? |
| 11:15.28 | pb_ | CM in what particular context? |
| 11:15.45 | sinterklaas | reproducability AND migrition |
| 11:16.06 | sinterklaas | migration |
| 11:16.15 | sinterklaas | and I hate bk |
| 11:16.56 | sinterklaas | so, if we make our 'own application oebuild, will they still work in about a year ? |
| 11:18.07 | pb_ | well, nobody has a great deal of real-world experience with deploying OE yet. But I'm reasonably confident that the reproducibility issue is under control. |
| 11:18.53 | sinterklaas | and also, out deliverables (for customers) will also be a toolchain, and we don't want to give them a new toolcain(s) |
| 11:18.53 | pb_ | All the output packages contain a Source: line giving the name of the input .oe file that was used to build them. So, as long as you still have a copy of the metadata and the upstream sources, you should be able to reproduce any build. |
| 11:19.20 | sinterklaas | yeah, but then migration comes to a stop |
| 11:19.35 | pb_ | Just archiving away your downloads/ and packages/ at the end of a build would be an easy way to ensure that you have all the requisite data for reproducibility. |
| 11:19.50 | pb_ | I'm not sure what you mean by migration. Migration from what, to what? |
| 11:20.24 | sinterklaas | 'newer' packages, say openssl update ? |
| 11:20.44 | sinterklaas | then we need to 'backport' it to an old build system |
| 11:21.00 | sinterklaas | and lose the advantage of oe... |
| 11:21.32 | sinterklaas | btw. I can see the solution, but can't really motivate it to my boss. |
| 11:21.43 | pb_ | There shouldn't be any backporting required. You just need to drop the new metadata into your repository, and you should be all set. |
| 11:22.40 | sinterklaas | is oe allready frozen ? I mean the package fromat ? |
| 11:23.33 | pb_ | More or less. kergoth is planning to make a 1.0 release of the OE core system in the fairly near future. |
| 11:24.32 | sinterklaas | till than I have to come up with a nice story to my boss, I really like to use oe for our new platforms |
| 11:24.53 | sinterklaas | and convince him to let me work for oe ! |
| 11:25.49 | pb_ | heh |
| 11:27.16 | sinterklaas | would be great, but the bastard only talk about f*cking configuration managment, bla bla |
| 11:31.18 | pb_ | mickeyl: would you delete packages/xkbd/xkbd-cvs.oe? |
| 11:31.30 | CIA-5 | 03mickeyl 07 * r1.1395 10packages/xkbd/xkbd_0.8.12.oe: add xkbd.desktop.patch - patch courtesy pb_.\ |
| 11:36.34 | *** join/#oe T`|Heli (total@GALILEO.ACT.CS.CMU.EDU) |
| 11:39.02 | CIA-5 | 03mickeyl 07 * r1.1396 10packages/gpe/meta-gpe.oe: remove gpe-conf temporarily - patch courtesy pb_.\ |
| 11:39.29 | CIA-5 | 03mickeyl 07 * r1.1397 10packages/gpe-question/gpe-question_0.03.oe: add gpe-question - patch courtesy pb_.\ |
| 11:40.07 | CIA-5 | 03mickeyl 07 * r1.1398 10packages/gpe-session-scripts/gpe-session-scripts_0.61.oe: add gpe-question - patch courtesy pb_.\ |
| 11:43.42 | CIA-5 | 03mickeyl 07 * r1.554 10oe/bin/oemake: avoid python crash if latest version of a package becomes ineligible - patch courtesy pb_.\ |
| 12:02.25 | mickeyl | pb_: deleted. |
| 12:02.28 | pb_ | thanks |
| 12:02.42 | CIA-5 | 03mickey 07 * r1.1393.1.1 10packages/BitKeeper/deleted/.del-xkbd-cvs.oe~beb11d126ad4c5e3: remove xkbd-cvs |
| 12:02.43 | mickeyl | hmm, no idea about sysv-rc |
| 12:03.01 | mickeyl | seems not to be in our repository |
| 12:03.12 | pb_ | odd. I wonder what Hrw was smoking. |
| 12:03.15 | mickeyl | heh |
| 12:03.35 | mickeyl | NOTE: package opie-image: build completed |
| 12:03.35 | mickeyl | NOTE: build 200405031134 completed |
| 12:03.35 | mickeyl | Build statistics: |
| 12:03.35 | mickeyl | <PROTECTED> |
| 12:03.38 | mickeyl | great. |
| 12:03.40 | mickeyl | from scratch |
| 12:03.48 | mickeyl | now lets see what is in rootfs |
| 12:04.13 | pb_ | cool |
| 12:05.03 | mickeyl | hmm |
| 12:05.09 | mickeyl | there are too much errors in the log |
| 12:05.16 | pb_ | not so cool |
| 12:05.16 | mickeyl | could you glance over it, please? |
| 12:05.19 | mickeyl | http://www.vanille.de/temp/log.do_images.22838 |
| 12:05.20 | pb_ | sure |
| 12:05.57 | mickeyl | most of it looks ok |
| 12:06.10 | mickeyl | but those ipkg-native errors |
| 12:06.10 | pb_ | Cannot find package qte-for-opie. |
| 12:06.10 | pb_ | Check the spelling or perhaps run 'ipkg update' |
| 12:06.10 | pb_ | Cannot find package qpf-bitstream-vera. |
| 12:06.10 | pb_ | Check the spelling or perhaps run 'ipkg update' |
| 12:06.10 | pb_ | Cannot find package libqpe-opie. |
| 12:06.12 | pb_ | Check the spelling or perhaps run 'ipkg update' |
| 12:06.14 | pb_ | that doesn't look so good. |
| 12:06.21 | pb_ | qte-for-opie sounds like DEPENDS vs RDEPENDS confusion |
| 12:06.23 | mickeyl | right. that's something i can fix |
| 12:06.32 | mickeyl | what about the fakeroot-native? |
| 12:06.43 | mickeyl | and the other -native packages it doesn't find |
| 12:06.48 | mickeyl | why is it attempting to install these... hmmm |
| 12:06.50 | pb_ | not sure. I've seen that too, but it seems to be harmless. |
| 12:06.56 | pb_ | I'll investigate. |
| 12:06.58 | mickeyl | ok thanks. |
| 12:07.15 | mickeyl | I'm going to look into the other things |
| 12:07.53 | pb_ | it sounds like image_ipk.oe's own DEPENDS are somehow leaking into IPKG_INSTALL. |
| 12:08.19 | mickeyl | yeah, that's what i was thinking |
| 12:08.27 | mickeyl | hmm |
| 12:08.29 | mickeyl | ERROR: Cannot satisfy the following dependencies for opie-systemtime: |
| 12:08.29 | mickeyl | <PROTECTED> |
| 12:08.33 | mickeyl | ah |
| 12:08.35 | mickeyl | we call it ntp |
| 12:08.45 | pb_ | oh, I see the problem. opie-image has: |
| 12:08.46 | pb_ | DEPENDS = "task-bootstrap ${OPIE_LIBS} ${OPIE_BASE}" |
| 12:08.46 | pb_ | export IPKG_INSTALL = "${DEPENDS}" |
| 12:09.03 | mickeyl | ah right |
| 12:09.08 | mickeyl | false package name |
| 12:09.16 | mickeyl | thanks |
| 12:09.29 | pb_ | you need to do that the other way around. write out the list of packages in IPKG_INSTALL, and then say "DEPENDS = ${IPKG_INSTALL}". |
| 12:10.46 | mickeyl | ah yes |
| 12:12.16 | CIA-5 | 03mickeyl 07 * r1.1398.1.1 10packages/ipkg/ipkg_0.99.120.oe: merge libipkg into ipkg.oe - patch courtesy pb_.\ |
| 12:12.17 | CIA-5 | 03mickeyl 07 * r1.1400 10packages/: |
| 12:12.17 | CIA-5 | Merge bk://openembedded@openembedded.bkbits.net/packages\ |
| 12:12.17 | CIA-5 | into gandalf.tm.informatik.uni-frankfurt.de:/usr/local/projects/packages\ |
| 12:14.27 | mickeyl | hmm |
| 12:15.13 | mickeyl | don't we need to say RPROVIDES="libipkg" in ipkg.oe now? |
| 12:15.27 | pb_ | no, it ships a separate libipkg.ipk |
| 12:15.33 | mickeyl | ah right |
| 12:16.32 | pb_ | for some weird reason it really is coming out as libipkg.ipk, not libipkg0.ipk, but I'm not too bothered about that right now. |
| 12:17.53 | mickeyl | to make sure DEPENDS = ${IPKG_INSTALL} works, i have to add more provides right? |
| 12:18.18 | pb_ | shouldn't do. what are you thinking of, specifically? |
| 12:18.32 | mickeyl | example is qte |
| 12:18.44 | mickeyl | we need the package libqte2 |
| 12:18.47 | pb_ | oh, right, yeah. I guess IPKG_INSTALL is really RDEPENDS. |
| 12:18.54 | pb_ | maybe you should just write out the list of packages twice. |
| 12:19.11 | pb_ | I don't think it's a good idea to make qte PROVIDE libqte2. |
| 12:19.11 | mickeyl | yeah, makes more sense than adding too much PROVIDES |
| 12:19.18 | mickeyl | k |
| 12:19.44 | mickeyl | actually most of the DEPENDS are not necessary because they are implicitly |
| 12:19.54 | mickeyl | if I say opie-taskbar, then the rest will be sucked in |
| 12:20.04 | pb_ | but you shouldn't really be needing stuff like libqte2 in IPKG_INSTALL anyway, because meta-opie and its dependencies should be taking care of that. |
| 12:20.04 | pb_ | right |
| 12:20.09 | mickeyl | here we are |
| 12:22.13 | mickeyl | a bit annoying in OE is when i built a complete image and changed some things in apps i must oebuild clean those individual apps to make sure they are rebuilt. |
| 12:22.20 | mickeyl | it'd be nice if that could happen automatically |
| 12:22.28 | pb_ | if you bump the PR, it'll happen automatically |
| 12:22.34 | mickeyl | heh |
| 12:22.41 | mickeyl | bumping the PR = manual work :D |
| 12:23.00 | pb_ | heh, right, but you're editing the .oe file anyway. |
| 12:23.00 | mickeyl | you're right though, i'm editing the files anyway |
| 12:23.05 | mickeyl | :D |
| 12:23.14 | pb_ | just changing "PR = 1" to "PR = 2" doesn't seem like it should be too taxing :-) |
| 12:23.24 | mickeyl | i think i can do that ;) |
| 12:23.31 | pb_ | hehe |
| 12:27.17 | *** join/#oe cmartin (~cmartin@gw.krypnet.de) |
| 12:29.16 | pb_ | bbl |
| 12:29.20 | mickeyl | cu have fun |
| 12:32.15 | cmartin | hi mickeyl |
| 12:32.45 | mickeyl | hi cmartin |
| 12:32.47 | mickeyl | how are things |
| 12:33.21 | cmartin | mickeyl, busy, busy, busy - been traveling all over Germany the last few weeks |
| 12:33.38 | cmartin | i hardly have time time check my mails |
| 12:34.23 | cmartin | mickeyl, but tonight my wife is out and i could start getting aquainted with oe - where should i start? |
| 12:35.37 | mickeyl | good question |
| 12:35.57 | mickeyl | major problem: OE progresses _FAST_ and docs are out of date |
| 12:36.05 | mickeyl | so you could give us a hand in updating docs and wiki |
| 12:36.07 | mickeyl | while you learn |
| 12:36.15 | mickeyl | my suggestion to start: |
| 12:36.26 | mickeyl | 1.) download oe, packages, oe-doc from openembedded.bkbits.net |
| 12:36.38 | mickeyl | 2.) read packages/local.conf.sample and adjust |
| 12:36.39 | CIA-5 | 03mickey 07 * r1.1401 10packages/ (meta/opie-image.oe ntp/ntp_4.1.2.oe): need to distinguish RDEPENDS and DEPENDS, however only in OPIE_LIBS |
| 12:36.40 | CIA-5 | 03mickey 07 * r1.1400 10packages/ntp/ntp_4.1.2.oe: ntp provides ntpdate |
| 12:36.51 | mickeyl | 3.) try oemake opie-image and see what happens |
| 12:37.05 | mickeyl | 4.) install and boot this image |
| 12:37.11 | mickeyl | 5.) fix bugs, add packages, etc. :D |
| 12:37.24 | cmartin | wow |
| 12:37.27 | mickeyl | correction to step 2 |
| 12:37.37 | mickeyl | 2.) read packages/local.conf.sample, copy to local.conf and adjust |
| 12:37.52 | cmartin | does a simpad image build?? |
| 12:38.01 | mickeyl | no |
| 12:38.07 | cmartin | :) |
| 12:38.20 | mickeyl | the jffs2 at least will be generated |
| 12:38.31 | mickeyl | you can then use make-simpad-flashimg manually |
| 12:38.46 | cmartin | k, so i could use the simpad utils to build the image |
| 12:38.50 | mickeyl | yeah |
| 12:38.56 | mickeyl | no one tried to build a simpad image yet |
| 12:39.01 | cmartin | sounds good |
| 12:39.23 | mickeyl | I no longer need the simpad-utils thankfully |
| 12:39.29 | mickeyl | since I use hh.org bootldr now |
| 12:39.39 | mickeyl | which is nearly finished |
| 12:39.45 | cmartin | yes, that's what I would work with also |
| 12:39.59 | mickeyl | good. then you can directly flash the .jffs2 |
| 12:40.08 | cmartin | k |
| 12:40.12 | mickeyl | and we no longer have a dedicated kernel partition *yay* |
| 12:41.07 | cmartin | since i don't want the next release to use the Siemens loader i wouldn't want to waiste any more time testing those images |
| 12:43.01 | cmartin | do you know if anybody working on the h2210 is involved with oe? since i now have one i want linux on it as soon as possible - i have been testing the rootfs they have out |
| 12:43.22 | mickeyl | not offhand |
| 12:43.24 | mickeyl | zecke works on kernel |
| 12:43.40 | cmartin | yes, i noticed recently in the ML |
| 12:43.55 | mickeyl | i think the others will follow |
| 12:44.04 | mickeyl | since OE seems the way to go for the familiar folks |
| 12:44.10 | mickeyl | it's just that not all have noticed yet :D |
| 12:44.31 | mickeyl | pb_ does _HUGE_ and good work for OE |
| 12:44.37 | mickeyl | we are glad that he joined |
| 12:44.57 | mickeyl | so all work we do benefits both OpenZaurus, OpenSIMpad and familiar now |
| 12:44.59 | mickeyl | which is a good thing |
| 12:45.02 | mickeyl | less duplicate work |
| 12:45.04 | mickeyl | more eyes on it |
| 12:45.06 | mickeyl | etc. |
| 12:46.30 | cmartin | yes, it would be nice to see all those projects work on a common base system |