| 00:00.02 | joyping | thats all |
| 00:00.13 | kergoth | and what happens wehn you mount? |
| 00:00.17 | kergoth | "refuses to mount" isnt very useful |
| 00:00.58 | joyping | when i try to mount for example: mount /dev/hda1 /mnt/cf |
| 00:01.02 | joyping | bo device |
| 00:01.08 | joyping | err. no device |
| 00:01.18 | kergoth | is syslogd running? |
| 00:01.20 | kergoth | is cardmgr running? |
| 00:01.33 | kergoth | if both are running, run logread and see what cardmgr had to say in the logs |
| 00:01.55 | gb2 | cardctl ident |
| 00:02.16 | gb2 | is ide-cs a module> |
| 00:02.29 | kergoth | and is it ide-cs or ide_cs? |
| 00:02.40 | joyping | cardctl ident works |
| 00:02.41 | kergoth | you know how it renamed in 2.4.somethingorother.. sharp renamed it back in their trees |
| 00:02.45 | kergoth | they just like to fuck with us |
| 00:02.54 | gb2 | heh |
| 00:03.15 | joyping | i get product info, function: 4 (fixed disk) |
| 00:03.44 | joyping | i found no module ide-cs or ide_cs |
| 00:03.50 | gb2 | lsmod |
| 00:04.09 | joyping | isn't it compiled into the kernel? |
| 00:04.14 | joyping | lsmod is empty |
| 00:05.13 | joyping | does ide-cs should come from the kernel? |
| 00:05.23 | gb2 | hmm |
| 00:10.22 | joyping | in sa kernel its called ide-cs |
| 00:11.14 | mjc | how do I get oe not to pull a file that already exists in ~/sources |
| 00:12.27 | joyping | its also compiled inkernel |
| 00:12.50 | joyping | in which package is it packed then? |
| 00:14.28 | kergoth | mjc: dump its md5sum into filename.md5, iirc. |
| 00:14.41 | kergoth | mjc: thats what it uses to know that its been successfully and fully obtained |
| 00:15.26 | mjc | thx. |
| 00:15.30 | kergoth | np |
| 00:15.31 | mjc | I'll try that in a sec |
| 00:18.44 | *** join/#oe KingRobot (~Mathew@dsl093-185-090.sea1.dsl.speakeasy.net) |
| 00:18.53 | mjc | stupid glibc checking out from cvs again |
| 00:19.26 | kergoth | set CVSDATE_glibc and CVSDATE_glibc-initial |
| 00:19.40 | mjc | nah, I'd rather have the newer version ;) |
| 00:19.48 | kergoth | well.. stop bitching then :) |
| 00:20.06 | mjc | kergoth: just means I have to wait to get my results wrt. openzaurus-pxa |
| 00:27.38 | gb2 | i usually lock cvsdate and bump it forward every week or so |
| 00:28.10 | kergoth | invariably as soon as i lock down cvsdate to today, i find out 3 packages were broken upstream today and i have to push it forward or back anyway |
| 00:28.13 | kergoth | damn murphy |
| 00:29.37 | gb2 | i need to find that poster again |
| 00:30.00 | gb2 | "murphy was an optimist" |
| 00:30.05 | kergoth | hehe |
| 00:30.24 | kergoth | http://www.despair.com/changewinds.html |
| 00:30.26 | kergoth | new one |
| 00:31.48 | gb2 | wonder if i have a browser in here |
| 00:32.54 | gb2 | nice |
| 00:37.47 | mjc | <PROTECTED> |
| 00:37.50 | mjc | <PROTECTED> |
| 00:37.53 | mjc | OSError: [Errno 2] No such file or directory: '/home/mjc/packages/tmp/base/glibc-initial-2.3.2+cvs20040519-r1/image/usr/lib/gconv' |
| 00:38.20 | gb2 | how do i copy from a text widget in opie? |
| 00:39.55 | kergoth | select it and click on the clipboard applet and hit copy.. dunno if theres a key combo bound to it |
| 00:40.21 | gb2 | hm |
| 00:40.26 | gb2 | applet? |
| 00:40.40 | kergoth | heh, install the clipboardapplet. could be another way, ask the folks in #opie |
| 00:41.49 | gb2 | hmmm |
| 00:41.51 | gb2 | heh |
| 00:42.01 | kergoth | have you seen http://bylands.dur.ac.uk/~mh/libb.stub.c.txt? its generated by this little app called relaytool. it creates a jumptable that lets you reference symbols from other libs normally, no special work involved, but it actually dlopens them rather than using -l. reduces some of the pain-in-the-ass nature of switching to accessing the libs that way at runtime |
| 00:42.06 | kergoth | kinda cool |
| 00:45.07 | gb2 | there |
| 00:45.33 | gb2 | http://www.art.com/asp/sp.asp?PD=10001840&RFID=042210&FT=Y |
| 00:46.06 | kergoth | ha, nice |
| 00:47.45 | gb2 | neat stub |
| 00:51.48 | gb2 | and makes it easy to do a "fat" binary |
| 00:52.42 | gb2 | do some section magic and i bet you could strip it as appropriate too |
| 00:53.09 | kergoth | yeah, my only oncern with something like that is size. its still nice to disable functionality to keep it small |
| 00:54.36 | gb2 | yeah |
| 00:54.45 | *** join/#oe treke|home (~ggilbert@lsanca2-ar29-4-41-064-194.lsanca2.elnk.dsl.genuity.net) |
| 00:57.13 | kergoth | http://www.art.com/asp/sp-asp/_/ui--762DE495268149718F58EE82576FA759/PD--10001845/Procrastinator'sCreed.asp# |
| 00:57.16 | kergoth | heh |
| 00:57.17 | kergoth | hey treke|home |
| 00:57.22 | chouimat | hi treke|home kergoth_ |
| 00:58.25 | kergoth | "13. I will never put off until tomorrow, what i can forget about forever." |
| 00:58.25 | treke|home | yo. |
| 01:00.37 | gb2 | kergoth: wish the murphy one came smaller |
| 01:00.49 | gb2 | it would be great for work |
| 01:01.37 | kergoth | yeah |
| 01:03.26 | gb2 | it was hanging in the computer lab in my highschool |
| 01:06.20 | gb2 | crap |
| 01:06.35 | gb2 | a window of my screen session is stuck |
| 01:06.41 | gb2 | ctrl-a k |
| 01:06.48 | gb2 | doesnt work |
| 01:07.14 | kergoth | heh, try ctrl-q? i hate it when i accidentally hit ctrl-s in one of my terminals |
| 01:07.20 | kergoth | damn software flow control |
| 01:08.50 | gb2 | it should still respond to kill |
| 01:09.14 | kergoth | heh, good point |
| 01:09.38 | gb2 | there |
| 01:09.45 | gb2 | :kill worked |
| 01:23.44 | heathkit | Is modutils-2.4.27-r1 broken? I'm getting this: "lex.l:429: error: `yytext_ptr' undeclared (first use in this function)" |
| 01:28.40 | *** join/#oe stigger (~jeff@dip-220-235-100-245.nsw.westnet.com.au) |
| 01:32.42 | kergoth_ | heathkit: install flex-old (assuming debian unstable) |
| 01:35.23 | *** join/#oe treke|home (~ggilbert@lsanca2-ar29-4-41-064-194.lsanca2.elnk.dsl.genuity.net) |
| 01:39.23 | heathkit | kergoth: Thanks. |
| 01:39.28 | kergoth | np |
| 01:48.31 | gb2 | crap |
| 01:49.01 | gb2 | hmmm |
| 01:53.58 | *** join/#oe bipolar (~bflong@dsl-66-216-162-196.dejazzd.com) |
| 02:01.46 | *** join/#oe gb2 (~gbritton@fog.sekrit.org) |
| 02:01.49 | gb2 | bah |
| 02:03.23 | *** join/#oe treke|home (~ggilbert@lsanca2-ar29-4-41-064-194.lsanca2.elnk.dsl.genuity.net) |
| 02:03.33 | treke|home | dunno why my irc client keeps crashing |
| 02:04.26 | *** join/#oe gb2 (~gbritton@fog.sekrit.org) |
| 02:04.28 | gb2 | foo |
| 02:04.33 | gb2 | stupid screen |
| 02:27.09 | *** join/#oe beissk (~beiss@srb9-d9bbf146.pool.mediaWays.net) |
| 03:25.26 | *** join/#oe boss (~boss@adsl-60-232.swiftdsl.com.au) |
| 03:25.40 | treke|home | uh oh |
| 03:25.42 | treke|home | the boss is here |
| 03:25.53 | boss | :) |
| 03:26.01 | boss | why do i always get that |
| 03:26.11 | treke|home | no idea |
| 03:27.41 | kergoth | we need more games with a boss key |
| 03:27.47 | kergoth | never see those anymore |
| 03:28.43 | treke|home | yeah |
| 03:28.47 | treke|home | hopefully larry 8 will have one |
| 03:39.11 | W8TVI | ibot: dumb bot |
| 03:39.13 | ibot | ACTION self-defecates. |
| 03:41.48 | W8TVI | ibot: good bot |
| 03:41.48 | ibot | W8TVI: thanks |
| 03:52.17 | *** join/#oe treke|home (~ggilbert@lsanca2-ar29-4-41-064-194.lsanca2.elnk.dsl.genuity.net) |
| 03:53.45 | kergoth | treke|home: was just checking the usb gadget site.. apparently it has a cdc ethernet driver that windows xp includes a driver for.. |
| 03:53.59 | kergoth | pretty cool, must be new |
| 03:54.19 | heathkit | does DISTRO = "openzaurus" (or other distro if applicable) |
| 03:54.19 | heathkit | INHERIT = "package_ipk" |
| 03:54.19 | heathkit | MACHINE = "collie" (or other device) |
| 03:54.24 | heathkit | whoops |
| 03:54.35 | kergoth | when you set DISTRO = something, it includes the ${DISTRO}.conf |
| 03:54.38 | kergoth | so if you want to know what it does |
| 03:54.41 | treke|home | very cool |
| 03:54.43 | kergoth | read openzaurus.conf |
| 03:55.06 | heathkit | Sorry, bad paste. Does 2.6.5-jl work, or require special workarounds? |
| 03:55.26 | kergoth | heh |
| 03:55.37 | kergoth | it boots. it doesnt support 90% of the peripherals |
| 03:56.01 | kergoth | in other words, dont fuck with it unless you're going to help finish writing the drivers for it |
| 03:56.14 | heathkit | haha, fair enough |
| 04:13.21 | *** join/#oe treke|home (~ggilbert@lsanca2-ar29-4-41-064-194.lsanca2.elnk.dsl.genuity.net) |
| 04:14.10 | treke|home | good choice |
| 04:15.36 | kergoth | hadnt seen it in a while. norm mcdonald is funny |
| 04:28.08 | treke|home | god I love this guy. Keeps insisting that Linux is just as user friendly as OSX for average users |
| 04:36.10 | gb2 | kergoth: which drivers does the 2.6.5 kernel support sofar? |
| 04:36.32 | treke|home | gb2: it's collie only atm |
| 04:36.53 | kergoth | and on collie its only a few. mtd. keyboard. framebuffer. couple others |
| 04:37.24 | gb2 | god i wish i had time |
| 04:38.36 | *** join/#oe treke|laptop (~ggilbert@lsanca2-ar29-4-41-064-194.lsanca2.elnk.dsl.genuity.net) |
| 04:39.16 | heathkit | Is the SD-controller driver still binary only? |
| 04:39.19 | gb2 | a real bootloader would really help that development, *sigh* |
| 04:57.32 | mjc | blarg I still keep having problems with glibc-initial |
| 05:14.12 | kergoth | "ha! you didnt count on my loyal army of prostitutes did you?!" |
| 05:46.13 | *** join/#oe msw (~msw@graft.XCF.Berkeley.EDU) |
| 06:09.42 | *** join/#oe ||ugh (~hugh@far.toad.com) |
| 06:14.59 | *** join/#oe emme (monga@195.176.179.81) |
| 06:32.42 | ||ugh | Sigh, SL-5500 was mis-named collie, right? Can we just change the names of the distro directorys to the names printed on the units? |
| 07:03.35 | *** join/#oe pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) |
| 07:05.36 | pb_ | morning |
| 07:20.45 | *** join/#oe kurre2 (~tomimo@ncircle.nullnet.fi) |
| 07:36.44 | calud_ | morning |
| 08:00.41 | *** join/#oe cyn (~cyn@adsl-068-209-224-046.sip.mem.bellsouth.net) |
| 08:20.41 | *** join/#oe CIA-5 (micah@hyperreal.info) |
| 08:21.06 | *** join/#oe CIA-5 (micah@hyperreal.info) |
| 08:38.41 | *** join/#oe ShALLaX|Laptop (~ufifiytt@cpc1-ware1-3-0-cust130.lutn.cable.ntl.com) |
| 08:45.19 | *** join/#oe mithro (~tim@dsl1-83.gw1.adl1.airnet.com.au) |
| 09:08.33 | *** join/#oe Speedy2 (JohnJacobs@cip-71-64.bbs.surfcity.net) |
| 09:08.38 | Speedy2 | re |
| 09:11.24 | mithro | anyone remeber why the cf ejects on suspend on 3.3.6pre1? |
| 09:20.58 | *** join/#oe zinga (~arnaudb@186.80-203-227.nextgentel.com) |
| 10:11.57 | Twiun | hi guys |
| 10:12.37 | Speedy2 | Hey |
| 10:12.43 | Speedy2 | Twiun: Have you ever used win4lin? |
| 10:12.59 | Twiun | nope, never |
| 10:13.08 | Speedy2 | Hrm |
| 10:13.14 | Speedy2 | I'm trying to find out if it's any good |
| 10:13.20 | Speedy2 | So far I've just found propaganda. |
| 10:13.49 | Twiun | lately I've seen wine mentioned a lot (and have used it in the past) |
| 10:14.00 | Twiun | I've also used vmware successfully |
| 10:14.20 | Speedy2 | I've used vmware. I've yet to try WINE but I know that some of the programs I want don't work under it. |
| 10:15.22 | Twiun | that's a shame, what programs out of interest?> |
| 10:16.08 | Speedy2 | M$ MapPoint 2004 |
| 10:16.18 | Speedy2 | One of the few (only?) M$ programs I use |
| 10:17.58 | Twiun | hmmm |
| 10:18.43 | Speedy2 | It's actually well done |
| 10:19.03 | Speedy2 | It has flaws, but 90% of it's functionality is usable. |
| 10:19.31 | Twiun | I wasn't putting it down ;) haven't used mappoint at all |
| 10:19.40 | Speedy2 | Oh I know |
| 10:19.45 | Speedy2 | I'm just saying it takes my breath away |
| 10:19.54 | Speedy2 | 90% functionality is good from an M$ product |
| 10:20.01 | Speedy2 | That's like a one a billion event. |
| 10:22.54 | Speedy2 | Twiun: What products have you used WINE with? |
| 10:23.14 | *** join/#oe pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) |
| 10:23.20 | Speedy2 | Heya pb_. |
| 10:25.16 | pb_ | Speedy2: Yo. |
| 10:26.12 | Speedy2 | pb_: Ever used win4lin? |
| 10:26.38 | Speedy2 | Just trying to find out how it compares against VMWare and WINE (though it's more like VMWare than WINE...) |
| 10:28.25 | pb_ | Speedy2: Nope, never tried it. |
| 10:35.09 | chouimat | morning |
| 10:35.15 | pb_ | hi chouimat |
| 10:45.26 | *** join/#oe ShALLaX|Laptop (~ufifiytt@cpc1-ware1-3-0-cust130.lutn.cable.ntl.com) |
| 10:45.47 | calud_ | hi all, do you have the following error too? In file included from sharpsl_apm.c:84: |
| 10:45.49 | calud_ | /home/oe/devel/build/tmp/base/openzaurus-pxa-2.4.18-rmk7-pxa3-embedix-r0/linux/include/asm/arch/irqs.h:51: warning: `NR_IRQS' redefined and so on.... |
| 10:47.16 | pb_ | calud_: that is not an error. |
| 10:48.28 | calud_ | but I think this is where all begins... |
| 10:49.15 | calud_ | compilation aborts with sharpsl_apm.c: At top level: |
| 10:49.15 | calud_ | sharpsl_apm.c:282: warning: `clock_slowed' defined but not used |
| 10:49.16 | calud_ | sharpsl_apm.c:316: warning: `on_press_time' defined but not used |
| 10:49.18 | calud_ | sharpsl_apm.c:1517: warning: `apm_get_bp_status' defined but not used |
| 10:49.19 | calud_ | make[2]: *** [sharpsl_apm.o] Error 1 |
| 10:49.50 | pb_ | None of those three lines are errors either. The problem must be elsewhere. |
| 10:56.10 | calud_ | the warnings in the log file starts in "In file included from sharpsl_apm.c:84:" and ends with make[2]: *** [sharpsl_apm.o] Error 1 |
| 10:59.06 | pb_ | calud_: any line that starts with 'warning:' is not an error. |
| 11:05.14 | calud_ | pb_: I know but it seems that those warnings resulted in the error... I could'n find any significant warnings/msg which may lead the error... |
| 11:06.31 | pb_ | calud_: I guess you have to debug this a bit further then. |
| 11:09.19 | calud_ | I guess so too I just want to ask if someone already had something similar :) |
| 11:15.59 | calud_ | how can I tell oe which kernel to build? |
| 11:16.58 | calud_ | I think when I do a oemake bootstrap-image it tries to compile all available kernels, is this right? |
| 11:17.49 | pb_ | it'll compile them all until it finds one that works. |
| 11:19.39 | calud_ | ok then it is normal :) |
| 11:21.34 | calud_ | and one more thing, how oe decides what packages to include in the image? is it all packages in the tmp/deploy/ipkg ? |
| 11:22.20 | pb_ | no, just the ones listed in IPKG_INSTALL. |
| 11:26.04 | calud_ | is there any predefined values for IPKG_INSTALL? if not this is probably tha cause of having ~100k images |
| 11:29.35 | calud_ | btw sorry for bothering but I know no sources to find answers except the mailing list and here... |
| 11:32.38 | pb_ | calud_: it's set in the appropriate .oe |
| 11:32.53 | pb_ | so task-bootstrap.oe, if that's what you're building |
| 11:48.02 | *** join/#oe CosmicPenguin (~nobody@aus-ext-proxy02.amd.com) |
| 12:12.54 | *** join/#oe schurig (~schurig@80.128.160.183) |
| 12:19.38 | pb_ | hi schurig |
| 12:19.44 | schurig | pb_: hi! |
| 12:20.10 | schurig | I've got a tip from Dimitry Andric on software floating point :-) |
| 12:20.42 | pb_ | aha |
| 12:46.37 | *** join/#oe sinterklaas (~sinterkla@hoc.xs4all.nl) |
| 12:47.57 | sinterklaas | hi, I send a ppp.oe and iptables.oe fix to oe@handhelds.org, who is responseble for adding/commenting on this ? |
| 12:49.42 | schurig | sinterklaas: I'm not sure if someone is formally responsible for this |
| 12:50.03 | schurig | sinterklaas: but kergoth, mickeyl, pb_ and /me can commit |
| 12:50.16 | sinterklaas | hmm, who's reading the oe@handhelds.org mail ? or is it a mailinglist ? |
| 12:50.49 | schurig | I think that mail to oe@hh.org goes to kergoth, in which case you'll have to wait until tonight. He's based in Texas and works when we Europeans sleep :-) |
| 12:51.25 | sinterklaas | lazy ;-) |
| 12:51.26 | schurig | sinterklaas: do you plan to provide more *.oe files? |
| 12:51.45 | schurig | in which case you can get a BK account and commit/maintain yourself |
| 12:51.59 | sinterklaas | think so, but don't feel familiar enough with oe jet... |
| 12:52.34 | sinterklaas | schurig: I really need to learn more about $%^% bk first.. |
| 12:52.48 | pb_ | oe@handhelds.org is a mailing list. Anybody can subscribe and read it. |
| 12:53.00 | schurig | sinterklaas: you don't need much BK knowledge. And you can get it all here in the channel. |
| 12:53.26 | schurig | sinterklaas: I only use "bk pull", "bk push", "bk citool", "bk rm", "bk del", "bk -r edit -q" and "bk -r get -q". That's all. |
| 12:53.31 | sinterklaas | schurig: I'm really waiting for my boss to let me spend more time on oe |
| 12:53.47 | schurig | sinterklaas: ahhh. I'm not in this mailing list ... |
| 12:54.01 | sinterklaas | schurig: me niether |
| 12:54.08 | schurig | sinterklaas: anyway, when YOU want you can get your bk account either from kergoth or from me |
| 12:54.35 | schurig | sinterklaas: for what are you using OE? |
| 12:54.58 | sinterklaas | schurig: 'custom' embedded arm platform(s) |
| 12:55.29 | schurig | sinterklaas: yeah, the interesing part is the "platform(s)" :-) Which ones? |
| 12:55.46 | sinterklaas | thre's allready a parking meter (partially) running oe ! |
| 12:56.11 | sinterklaas | abd a coffee machine coming up... |
| 12:56.14 | sinterklaas | really... |
| 12:56.36 | schurig | sinterklaas: hmm, I can't see a mail from you at http://handhelds.org/hypermail/oe/current/ |
| 12:57.02 | sinterklaas | schurig: I send it monday, I think |
| 12:57.23 | schurig | hmm, maybe it's stuck in some "has to be moderated" queue because you aren't subscribed |
| 12:57.24 | sinterklaas | schurig: from: jeroen@hoc.xs4all.nl |
| 12:57.32 | *** join/#oe lumi (~dleifer@netblock-66-159-210-39.dslextreme.com) |
| 12:58.01 | sinterklaas | I'll subscribe |
| 12:58.12 | schurig | sinterklaas: send you OE files to hs4233@mail.mn-solutions.de, I'll add them |
| 12:58.37 | sinterklaas | schurig: I hope it's worth it... |
| 12:59.48 | schurig | sinterklaas: I'll need a pppd at some later time anyway, so your OE files is a good contribution (for me, that is, BIG GRIN) |
| 13:03.46 | sinterklaas | schurig: send it |
| 13:04.27 | schurig | sinterklaas: ahh, you've got a baby. Me too. She's now 10 months old |
| 13:04.55 | sinterklaas | 8 months, that's why I need time to work for oe from my boss... |
| 13:06.30 | schurig | sinterklaas: Upps, which ancient e-mail program do you use? Your files end up as uuencoded text. |
| 13:07.34 | schurig | sinterklaas: in which RPM is uudecode? |
| 13:07.41 | sinterklaas | schurig: uhm, just using unix mail and uuencode... which modern mailclient can't read uuencoded files ? |
| 13:08.19 | pb_ | heh, I don't think _any_ modern email client can directly read uuencoded files. |
| 13:08.28 | sinterklaas | mozilla does |
| 13:08.30 | pb_ | obviously you can save them to disk and then pipe through uudecode, but it's kind of tedious. |
| 13:08.45 | pb_ | ah |
| 13:09.26 | sinterklaas | schurig: if i'm back home I'll send them the 'normal' way |
| 13:10.23 | sinterklaas | look at our tax-cow : http://www.fam-riemersma.nl/gallery/chess_fotos?page=1 |
| 13:11.20 | schurig | sinterklaas: now, I found uudecode in the uudeview RPM, I'm about to commit them to bk |
| 13:11.51 | schurig | sinterklaas: eeek, your're working for the enemy (greedy municipal govermnents :-) |
| 13:12.45 | sinterklaas | schurig: hehe, free parking for me ;-) |
| 13:14.16 | schurig | sinterklaas: I pushed ppp. iptables.oe is already there. Did you make changes to get it working? |
| 13:14.19 | sinterklaas | schurig: I'll quess your barcode scanners are for the poor ? :) |
| 13:14.56 | schurig | sinterklaas: why for the poor? |
| 13:15.08 | sinterklaas | yeah, the original version wouldn't work, after playing around I decided to link it statically |
| 13:15.23 | sinterklaas | schurig: cheap grosseries ? |
| 13:16.10 | schurig | sinterklaas: they are mostly not used in stores, but in warehouses |
| 13:16.52 | schurig | sinterklaas: e.g. you won't see http://www.mn-solutions.de/holotrak.html at grocery stores :-) |
| 13:17.17 | sinterklaas | schurig: we have a customer working with barcode scanners, I'll tip him... |
| 13:19.01 | schurig | sinterklaas: but they won't run OE, only http://www.mn-solutions.de/mnci.html does |
| 13:19.57 | schurig | the web site is brand new (based on Typo3), I didn't have time to test it against all 9312 different browsers. So if you see an error, please tip me :-) |
| 13:23.09 | schurig | sinterklaas: committed your iptables as well. I did not do any tests, I hope it works. If not, we can revert the change :-) |
| 13:24.45 | schurig | hmm, that beast doesn't commit and tells that there is nothing to push |
| 13:26.49 | sinterklaas | schurig: I have to go, maybe I want a bk thinggy, I'll let you know. btw. our customer is working for http://www.transtrack.nl/sites/default/default.asp |
| 13:27.24 | schurig | ahh, now it committed |
| 13:46.55 | *** join/#oe joy_ (~joy@pD9E73F26.dip.t-dialin.net) |
| 13:48.27 | *** join/#oe mrz80 (1000@caledonia.nerdc.ufl.edu) |
| 13:52.53 | pb_ | hi mickey |
| 13:54.18 | schurig | hmm, which IPK has /etc/ipkg.conf and /etc/arch.conf? |
| 13:58.52 | *** join/#oe prpplague (~billybob@12.148.134.14) |
| 13:59.21 | pb_ | hi plague |
| 14:01.20 | prpplague | pb_: morning |
| 14:01.33 | prpplague | pb_: hows biz today? |
| 14:09.33 | pb_ | not too bad. trying to debug some audio dma problem on this Samsung board. |
| 14:09.57 | prpplague | lovely |
| 14:10.49 | prpplague | pb_: ever heard of these guys - http://www.dspdesign.com/products/product_detail?product_id=110 |
| 14:14.19 | pb_ | nope. sounds cute though. |
| 14:16.20 | prpplague | pb_: yea, sounds like a nice dev board, just can't anyone to respond to emails |
| 14:16.27 | pb_ | doh |
| 14:16.49 | *** join/#oe joy_ (~joy@pD9E73F26.dip.t-dialin.net) |
| 14:19.37 | *** join/#oe bipolar (~bflong@146.145.26.90) |
| 14:21.24 | CosmicPenguin | morning |
| 14:22.33 | prpplague | CosmicPenguin: morning |
| 14:24.40 | CosmicPenguin | prpplague: hows tricks? |
| 14:24.53 | prpplague | CosmicPenguin: ssdd |
| 14:25.13 | prpplague | CosmicPenguin: suppose to be hacking this smartcard pc setup, but i'm totally bored with it |
| 14:26.59 | CosmicPenguin | It might be time to go to walmart and get myself a floor lamp |
| 14:32.07 | *** join/#oe Shayde (~dbs@seawall.homeport.org) |
| 14:51.13 | *** join/#oe jmau (~jmau@pD9E5DC8D.dip.t-dialin.net) |
| 14:51.18 | *** join/#oe _jmau_ (~jmau@pD9E5DC8D.dip.t-dialin.net) |
| 15:11.06 | *** join/#oe mrz80 (1000@caledonia.nerdc.ufl.edu) |
| 15:21.21 | mickeyl | ~lart sourceforge for telling us mailing list archives will be down for "a few days" since 2004-04-19 |
| 15:21.51 | mickeyl | guess four weeks is still a few days for them |
| 15:22.45 | mickeyl | oh btw. |
| 15:22.56 | mickeyl | anyone notice we hit the 800 mark in .oe files with my latest commit? |
| 15:25.00 | *** join/#oe tomimo (~kurre@ncircle.nullnet.fi) |
| 15:39.53 | pb_ | mickeyl: oh, cool |
| 15:41.56 | *** join/#oe TimRiker (~timr@c-24-1-104-25.client.comcast.net) |
| 15:43.33 | mickeyl | not sure if i like the .oefiles/maintainer ratio though :D |
| 15:44.45 | pb_ | heh, right |
| 15:44.56 | mickeyl | :D |
| 15:45.17 | mjc | blarg |
| 15:46.13 | mjc | stupid glibc-initial problem.. nobody bothered to commit the fix (hackish or not) and I've forgotten what it was |
| 15:47.01 | mickeyl | glibc-initial is out of my scope. i won't touch anything there. need to pester gb2 or kergoth |
| 15:47.34 | mickeyl | besides i can't reproduce the problem since it just works (tm) here |
| 15:48.29 | pb_ | this is the split_packages() thing? that's just too weird. I have no clue what's going wrong there. |
| 15:49.47 | pb_ | someone who actually experiences the problem needs to debug it, I guess. |
| 15:55.11 | lumi | mickeyl: the opie-today-stocktickerplugin should be moved to non-working.... |
| 15:55.43 | mickeyl | lumi: no, i just fixed it upstream |
| 15:56.04 | mickeyl | update cvs and rebuild to check please |
| 15:56.41 | mickeyl | problem was that stockticker contained an application named stockticker |
| 15:56.54 | lumi | mickeyl: oh okay. I'll check. |
| 15:56.55 | mickeyl | due to our DESTDIR= it tried to create stockticker in the ${S} |
| 15:57.02 | mickeyl | which collided with the directory name |
| 15:58.00 | mickeyl | actually i'm surprised that we don't hit something like that earlier... well, probably we have only few binaries in opie left. most are plugins |
| 15:58.08 | mickeyl | s/don't/didn't/ |
| 16:01.01 | lumi | mickeyl: A question: is there a rule about having one .oe per directory? what about putting opie-handwriting-sets_cvs.oe into opie-handwriting/? |
| 16:01.37 | mickeyl | lumi: not sure about that. it would need a rewrite of opie-meta which at the moment only can cope with one oefile pre directory. |
| 16:01.50 | mickeyl | lumi: generally, there's no problem with grouping files. |
| 16:01.52 | mickeyl | i did that for libsdl |
| 16:01.56 | mickeyl | and python, of course |
| 16:01.56 | lumi | mickeyl: got it. |
| 16:02.47 | mickeyl | good work on the latest .oe's btw., nearly didn't touch anything |
| 16:03.05 | lumi | mickeyl: thanks.... |
| 16:03.47 | treke | lumi: opie-handwriting-sets should probably be a part of opie-handwriting.oe |
| 16:05.19 | lumi | treke: I'm missing the sophistication to combine the multiple .oe functions. I'm happy to learn from samples..... |
| 16:06.20 | treke | libopie2 is probably one |
| 16:06.54 | mickeyl | basically just have to adapt PACKAGES and FILES_<packagename> |
| 16:07.02 | treke | http://openembedded.bkbits.net:8080/packages/anno/libopie/libopie2_cvs.oe%401.25?nav=index.html|src/.|src/libopie |
| 16:07.32 | treke | yeah, it's pretty easy when there is just a handful of packages that are fixed in advance :p |
| 16:08.26 | lumi | treke: I'll check. |
| 16:13.49 | mickeyl | gb2: ping |
| 16:13.55 | *** join/#oe reenoo_ (~r@p508676CD.dip.t-dialin.net) |
| 16:14.27 | reenoo_ | afternoon |
| 16:14.34 | mickeyl | afternoon reenoo_ |
| 16:15.16 | reenoo_ | yo mickeyl |
| 16:15.18 | reenoo_ | brb |
| 16:32.54 | pb_ | doh, something is still a bit funky with shlibdeps. |
| 16:33.04 | pb_ | <PROTECTED> |
| 16:33.04 | pb_ | <PROTECTED> |
| 16:33.04 | pb_ | [...] Depends: glibc-locale*, libc6 (>= 2.3.2+cvs20040519) |
| 16:33.15 | reenoo_ | re |
| 16:52.02 | CosmicPenguin | pb_: I can't believe that xorg compiled our geode stuff with very minimal changes |
| 16:52.09 | CosmicPenguin | Thats scary |
| 16:53.55 | pb_ | CosmicPenguin: heh |
| 16:55.43 | CosmicPenguin | pb_: did you have any luck? I had to do wierd things with xkbfile to make it compile and then I had to edit the .h post install to make xorg compile |
| 16:57.40 | pb_ | CosmicPenguin: mm, good question. I think my build got bogged down recompiling glibc or something, and then I forgot about it. |
| 16:57.42 | pb_ | I'll check the logs in a bit. |
| 16:58.51 | CosmicPenguin | nod |
| 17:02.26 | mickeyl | patcher rules |
| 17:02.35 | mickeyl | at least it did in the times it work... |
| 17:03.42 | treke | just grab the one from buildroot |
| 17:03.43 | treke | it works |
| 17:04.02 | treke | http://ent.brooks.edu/~ggilbert/patcher |
| 17:04.45 | mickeyl | yep |
| 17:07.46 | mickeyl | damn |
| 17:08.01 | mickeyl | that recent embedix kernel is hosed for everything else than tosa |
| 17:08.56 | treke | I guess that's good in its own way |
| 17:09.24 | mickeyl | i can't see how that could be good, 'cause it contains interesting fixes also for the others |
| 17:09.41 | treke | encourages work on the jl tree :p |
| 17:10.03 | mickeyl | hah |
| 17:10.12 | mickeyl | i don't buy that :) |
| 17:38.26 | *** join/#oe jochen (~jochen@pD9E27F56.dip.t-dialin.net) |
| 18:02.26 | *** join/#oe pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) |
| 18:03.21 | pb_ | re |
| 18:03.41 | treke | er |
| 18:14.57 | *** join/#oe [wobble]_ (~martin@80.131.32.175) |
| 18:33.51 | *** join/#oe CosmicPenguin (~nobody@aus-ext-proxy01.amd.com) |
| 18:34.05 | *** join/#oe fido_hl (~hl@193.171.34.174) |
| 18:48.34 | mickeyl | pb_: ping |
| 18:48.51 | pb_ | mickeyl: yo |
| 18:48.58 | mickeyl | pb_: any idea what happens if a PREFERRED_PROVIDERS line contains two entries for the same virtual providers ? |
| 18:49.29 | treke | The end of the world |
| 18:49.44 | pb_ | mickeyl: it will pick one at random |
| 18:49.52 | mickeyl | ouch |
| 18:49.53 | mickeyl | too bad |
| 18:50.04 | mickeyl | it would've been nice if the second would override the first |
| 18:50.32 | pb_ | mickeyl: I think the right thing is to redo PREFERRED_PROVIDERS to have one variable per virtual. |
| 18:50.33 | mickeyl | concrete working problem is this: |
| 18:50.39 | pb_ | PREFERRED_PROVIDER_foo = bar |
| 18:50.53 | mickeyl | hmm overrides |
| 18:51.10 | pb_ | well, not actually overrides, but it would look the same. |
| 18:51.14 | mickeyl | yeah |
| 18:51.20 | mickeyl | concrete example is tosa.conf |
| 18:51.36 | mickeyl | i needed to rename openzaurus-pxa to openzaurus-tosa |
| 18:51.44 | mickeyl | tosa.conf could include conf/zaurus-pxa.conf |
| 18:51.51 | pb_ | right |
| 18:51.53 | mickeyl | but would need to specify another provider |
| 18:52.13 | mickeyl | which it can't do, cause we use PREFERRED_PROVIDERS_append |
| 18:52.20 | pb_ | you want to have a go at implementing the broken-down PREFERRED_PROVIDER thing in oemake? |
| 18:52.22 | mickeyl | (cause otherwise we would override those in local.conf) |
| 18:52.46 | mickeyl | pb_: ah not really... i didn't look at oemake for quite a while. |
| 18:53.20 | mickeyl | for now i use a verbose tosa.conf... i'd rather see the kernel fixed to work for all machines though |
| 18:53.48 | mickeyl | add one annoying issue to the neverending list of embedix-kernel oddities |
| 18:54.30 | jochen | Hmm, opensimpad kernel crashes immediately after starting init on CL4 |
| 18:54.48 | mickeyl | florian reported the same |
| 18:54.57 | jochen | And bootloader misdetects the memory size as 32MB |
| 18:55.06 | jochen | (CL4 only has 16) |
| 18:55.10 | mickeyl | which bootloader, btw? |
| 18:55.21 | jochen | the "new" one :-) |
| 18:55.29 | mickeyl | hh.org ? |
| 18:55.34 | jochen | yes |
| 18:55.40 | mickeyl | ah welcome on board |
| 18:55.42 | mickeyl | unfinished |
| 18:55.46 | jochen | i know |
| 18:55.52 | mickeyl | no one able to fix |
| 18:56.02 | mickeyl | Chicken did his best but he lacks information sources |
| 18:56.22 | jochen | iirc, CL4 has only one bank, so i gave it a try :-) |
| 18:56.30 | mickeyl | jochen: try building 2.4.25-vrs2-pxa1-jpm1 with a known good cross-gcc 3.3.2 from hh.org. i think we have a toolchain issue here |
| 18:56.51 | mickeyl | jochen: help me pestering chicken to finish lab support for SIMpads :) |
| 18:57.13 | mickeyl | but at the moment, he has "too much fun" working on jornada-kernels :) |
| 18:59.53 | jochen | Hmm, i still have an old 2.95.4 based toolchain installed, but i guess that one is too old and the current 3.3.2 from openembedded is b0rked... |
| 18:59.58 | *** join/#oe CIA-5 (micah@hyperreal.info) |
| 19:00.48 | mickeyl | jochen: the 3.3.2 from hh.org is known to produce working kernels. we would need to be sure if the 3.3.2 in oe is broken, so could you do that? |
| 19:08.18 | mickeyl | any sa-usb experts here? |
| 19:08.22 | mickeyl | usb_send: Not TPC: UDCCS2 = 1 |
| 19:08.22 | mickeyl | udc_ep2_int_hndlr: hmm, what to do here? |
| 19:08.22 | mickeyl | udc_ep2_int_hndlr: Buffer is c01ebdf0 length 154 |
| 19:08.22 | mickeyl | udc_ep2_int_hndlr: Retransmitting |
| 19:08.22 | mickeyl | udc_ep2_send: busy sending 154 - returning -EBUSY |
| 19:08.22 | mickeyl | NETDEV WATCHDOG: usbf: transmit timed out |
| 19:08.24 | mickeyl | usbeth_xmit_timeout: Resetting Tx endpoint |
| 19:08.27 | mickeyl | usbeth_xmit_timeout: Sending packet |
| 19:08.29 | mickeyl | udc_ep2_send: busy sending 154 - returning -EBUSY |
| 19:08.38 | mickeyl | and then keeps iterating the last messag |
| 19:08.41 | jochen | I'll prepare a build environment on an Intel box somewhere |
| 19:08.48 | mickeyl | always dropping usb packets then |
| 19:10.32 | jochen | My normal build machine is an Alpha, but for some reason compiling an ARM kernel took a long time on that box. Even compiling a full native XFree86 Debian package was faster :-/ |
| 19:11.47 | *** join/#oe CosmicPenguin (~nobody@aus-ext-proxy02.amd.com) |
| 19:11.52 | CosmicPenguin | join #geode-amd64 |
| 19:15.41 | CosmicPenguin | sorry bout that |
| 19:15.58 | CosmicPenguin | can you combine <var>_$machine and <var>_append? |
| 19:16.17 | CosmicPenguin | So if I wanted to add something to SRC_URI for machine "geode"? |
| 19:17.55 | pb_ | I don't think so. The obvious syntax would be 'var_append_geode', but this probably won't do the right thing because of the magical way that _append works. |
| 19:18.10 | pb_ | You might have to write a bit of Python to munge the variables the way you want. |
| 19:18.35 | CosmicPenguin | or more correctly, I think, just write a new .oe that inherits the old one |
| 19:18.43 | pb_ | Or that, yes |
| 19:18.45 | *** join/#oe florian (~fuchs@pD9E5F08C.dip.t-dialin.net) |
| 19:18.49 | florian | hi all |
| 19:20.10 | mickeyl | ho florian |
| 19:20.19 | mickeyl | we were just talking about the SIMpad dying in init() |
| 19:20.28 | mickeyl | florian: did you do any tests with a known-good-compiler yet? |
| 19:20.34 | florian | Ahh... |
| 19:21.19 | florian | mickeyl: no, not yet... i didn't have that much time this week. |
| 19:21.47 | mickeyl | k, nm |
| 19:21.50 | jochen | florian: i just had the same "feature" |
| 19:21.51 | florian | Aditionally i still don't know that much about oe.. how do i select the compilers to use? |
| 19:22.08 | mickeyl | no i recommend compiling the kernel manually to be sure |
| 19:22.15 | mickeyl | however, see KERNEL_CCSUFFIX |
| 19:22.19 | mickeyl | and KERNEL_LDSUFFIX |
| 19:22.26 | florian | ah, ok |
| 19:23.18 | florian | compiling manually should be easy, i have one of jameys crosstoolchains installed which defintely works for kernels. |
| 19:23.48 | florian | Maybe i have some time later this evening to try this :-) |
| 19:24.05 | mickeyl | yeah |
| 19:24.11 | CosmicPenguin | pb_: with the work on the module loader (presumably) happening, do you think its a worthwhile patch to add an xorg hack to choose which drivers to build? |
| 19:24.23 | mickeyl | florian: all my manually compiled 2.4.25-vrs2-pxa1-jpm1 boot fine. |
| 19:24.30 | pb_ | CosmicPenguin: maybe you should check with the #freedesktop dudes |
| 19:24.31 | mickeyl | although i didn't test with an OE-built one |
| 19:24.42 | florian | mickeyl: ok, another reason to check this... |
| 19:24.58 | pb_ | CosmicPenguin: daniels is the guy who has been mostly working on xorg |
| 19:25.23 | florian | We don't have samba in oe, right? |
| 19:25.28 | CosmicPenguin | Matthew is in there, but he's the only guy I know |
| 19:25.28 | pb_ | florian: right |
| 19:27.42 | florian | hmm... then it will be a more complex job to get libsmbclient. :-/ |
| 19:32.03 | pb_ | florian: it'd probably be easy to add samba to oe |
| 19:32.06 | pb_ | maybe you'd like to try that. |
| 19:33.02 | CosmicPenguin | florian: oh, that would be nice. Less work for me later... :) |
| 19:33.11 | florian | pb_: Sounds like a good idea... i started adding gkdial, but its configure script sucks. I'll try again later... |
| 19:33.30 | pb_ | florian: heh, right |
| 19:33.48 | florian | First i want to see if it works on my iPaq :-) |
| 19:34.05 | pb_ | :-) |
| 19:39.09 | kergoth`bbl | CosmicPenguin: _append_override and _override_append both *should* work, though I expect they probably dont at this point. I wrote test cases to verify it, and they didnt all pass. |
| 19:40.48 | pb_ | heh |
| 19:41.58 | pb_ | actually, looking at the code, _append_override has more of a chance than I thought it did. |
| 19:42.38 | kergoth`bbl | yeah, there was always logic to handle both cases. if it doesnt, its just a bug. |
| 19:43.13 | CosmicPenguin | kergoth`bbl: once I thought about it, I figured that for machine dependent patches like that, an inherited .oe is much more intellegent |
| 19:44.13 | kergoth`bbl | yep. always multiple ways to get the job done |
| 19:45.24 | CosmicPenguin | I'm strugggling with spreading out machine dependent fixes throughout the set of packages |
| 19:45.40 | CosmicPenguin | Personally, I think I would rather have them all together in one spot, but I realize thats probably sixes |
| 19:45.52 | kergoth`bbl | its ugly. itd be nice if they could all be splipt out into a seperate repo for each target |
| 19:45.57 | CosmicPenguin | you either have machine dependant stuff spread out, or you have package dependent stuff spread out, so... |
| 19:45.57 | kergoth`bbl | split rather |
| 19:55.36 | kergoth`bbl | florian, CosmicPenguin: there's a samba build or two in nonworking.. dont remember why it doesnt build atm, but at least you've got a starting point |
| 19:56.29 | CosmicPenguin | kergoth`bbl: thanks. For me, its a bit far away, but I just know that Samba is going to be on the list of requirements sooner rather then later |
| 19:57.32 | *** join/#oe darkschneider (~gab@81-208-36-80.fastres.net) |
| 19:59.01 | CosmicPenguin | kergoth`bbl: you get these feelings when you work for a house that was built on Windows and only recently embraced the way of the penguin |
| 19:59.33 | kergoth`bbl | heh |
| 20:03.48 | florian | kergoth`bbl: ok, great |
| 20:12.39 | Twiun | rehi |
| 20:13.33 | CosmicPenguin | Its not trivial... I'm sure theres a better way to orgianize that |
| 20:22.59 | jochen | Mounted devfs on /dev |
| 20:22.59 | jochen | Freeing init memory: 88K |
| 20:22.59 | jochen | Kernel panic: Attempted to kill init! |
| 20:23.25 | mickeyl | kergoth`bbl: yep |
| 20:23.45 | mickeyl | jochen: this is a kernel with the oe toolchain or with the hh.org 3.3.2 ? |
| 20:23.57 | prpplague | mickeyl: there are used to using m$ time scale |
| 20:24.16 | jochen | That's the one with hh.org. But only the kernel has been built with that release. |
| 20:24.24 | *** join/#oe binaryG3N3S1S (~binaryG3N@67.104.83.13.ptr.us.xo.net) |
| 20:24.31 | prpplague | mickeyl: i.e "12 minutes remaining install" where it actually takes 46 |
| 20:24.36 | mickeyl | jochen: guess we can rule out the toolchain as a problem |
| 20:24.41 | mickeyl | prpplague: heh, absolutely yeah :) |
| 20:24.59 | mickeyl | jochen: could you try to chroot into the environment somehow? |
| 20:25.19 | jochen | oe crashed differently: |
| 20:25.20 | mickeyl | or try to execute some of the files via network ? |
| 20:25.20 | jochen | pc : [<40147000>] lr : [<400184f0>] Not tainted |
| 20:25.20 | jochen | sp : bffffafc ip : 4002f000 fp : bffffca3 |
| 20:25.20 | jochen | r10: 000167c8 r9 : 00000000 r8 : 00000000 |
| 20:25.20 | jochen | r7 : 00000000 r6 : 000085f0 r5 : 000167c8 r4 : 000167c8 |
| 20:25.20 | jochen | r3 : 00280006 r2 : 00000023 r1 : 00000224 r0 : 40097476 |
| 20:25.22 | jochen | Flags: nzCv IRQs on FIQs on Mode USER_32 Segment user |
| 20:25.24 | jochen | Control: C001317F Table: C001317F DAC: 00000015 |
| 20:25.26 | jochen | [1] init: obsolete system call 00642e74. |
| 20:25.30 | mickeyl | ouch |
| 20:27.20 | treke | hehehe. |
| 20:27.36 | mickeyl | cool |
| 20:27.42 | treke | The police department forgot to call and tell us the swat team was going to be out for training today :) |
| 20:29.48 | jochen | Hmm, i don't have a working image on the simpad right now. |
| 20:30.29 | jochen | However, "Attempted to kill init" smells like some libc problem or so. |
| 20:31.38 | mickeyl | could you check with another device capable of running arm binaries? |
| 20:34.43 | CIA-5 | 03mickeyl 07 * r1.1460 10packages/ (4 files in 3 dirs): merge handwriting sets into handwriting and opie-dasher languages into opie-dasher |
| 20:37.09 | jochen | I just put the image on http://jochen.scram.de/rootfs.tar.gz |
| 20:37.57 | jochen | It's an 4 MB bootstrap image |
| 20:37.58 | mickeyl | this is built after the broken toolchain week ? |
| 20:38.21 | jochen | It is built 2 days ago. |
| 20:38.31 | mickeyl | from scratch? |
| 20:38.34 | jochen | yes |
| 20:38.36 | mickeyl | good |
| 20:38.46 | kergoth`bbl | uclibc or glibc? |
| 20:38.51 | jochen | glibc |
| 20:39.16 | pb_ | what kernel is that? |
| 20:39.18 | *** join/#oe hufnus (~slonsiki@DSL135-071.LABridge.com) |
| 20:39.35 | jochen | It contains the pp.org built one |
| 20:39.45 | jochen | s/pp.org/hh.org |
| 20:41.04 | jochen | Only the kernel though. All the rest (including System.map) is from oe. |
| 20:41.32 | mickeyl | hmm |
| 20:41.50 | mickeyl | are we really really sure that the uclibc patches we apply to the (glibc-)toolchain don't break anything? |
| 20:42.17 | pb_ | mickeyl: I successfully booted an oe-built rootfs on an arm machine this afternoon. |
| 20:42.23 | kergoth`bbl | i've booted glibc oe root filesystems on arm platforms last week |
| 20:42.32 | mickeyl | ok |
| 20:42.34 | pb_ | so yes, I'm fairly sure |
| 20:42.35 | mickeyl | sure enough for me |
| 20:42.39 | pb_ | jochen: what machine are you on? |
| 20:42.45 | jochen | Alpha |
| 20:43.01 | jochen | 2.6.6-rc1 #10 Sun Apr 25 15:14:04 CEST 2004 alpha GNU/Linux |
| 20:43.02 | pb_ | target machine, I mean |
| 20:43.11 | jochen | Open Simpad CL4 |
| 20:43.28 | *** part/#oe treke (~ggilbert@ip230-225.outside.westmont.edu) |
| 20:43.35 | *** join/#oe treke (~ggilbert@ip230-225.outside.westmont.edu) |
| 20:44.08 | kergoth`bbl | there we are |
| 20:44.09 | kergoth`bbl | NOTE: package php-4.3.6-r0: task do_compile started |
| 20:44.20 | kergoth`bbl | one down |
| 20:44.29 | mickeyl | heh, but couldn't you attempt a more interesting package? |
| 20:44.50 | pb_ | heh |
| 20:44.50 | kergoth`bbl | actually nanoweb is pretty useful. a webserver with php support thats still small is a useful thing |
| 20:44.57 | treke | ~lart mickeyl |
| 20:45.02 | mickeyl | ouch |
| 20:45.17 | treke | we wanted a scripting language that didnt suck like python |
| 20:45.22 | mickeyl | heh yeah |
| 20:45.26 | mickeyl | feel free to rewrite oe in php |
| 20:45.27 | CosmicPenguin | hack in gettext.. :) |
| 20:45.38 | treke | mickeyl: ok. |
| 20:45.42 | jochen | use awk _) |
| 20:46.34 | *** join/#oe T`|Heli (total@GALILEO.ACT.CS.CMU.EDU) |
| 20:47.15 | *** join/#oe T`|Heli (total@GALILEO.ACT.CS.CMU.EDU) |
| 20:51.04 | pb_ | gosh, ipkg sure gets slow when it has the debian archive configured as a feed. |
| 20:51.23 | pb_ | I guess Pigi needs to re-think his data structures. |
| 20:51.53 | mickeyl | hehe |
| 20:52.27 | jochen | pb_: and even worse, when it creates a Packages file, it'll put the oldest packages in there instead of the newest ones. |
| 20:52.42 | pb_ | jochen: eh? ipkg doesn't create Packages. |
| 20:53.13 | jochen | ipkg-utils |
| 20:54.15 | pb_ | I still don't understand exactly what you're saying. Can you be more explicit? |
| 20:54.31 | pb_ | If you think that ipkg-make-index is doing the wrong thing, please file a bug against it. |
| 20:55.14 | jochen | If you have two different revisions of a package in a directory and create a Packages, only the older package will end up in Packages. |
| 20:55.29 | pb_ | Do you have a testcase? |
| 20:56.00 | jochen | I'll try to construct one |
| 20:56.07 | pb_ | Thanks. |
| 20:56.26 | pb_ | I'm a bit surprised that we haven't run across that problem at handhelds.org. |
| 21:00.06 | fontenot | pb_, are there OE images available for ipaqs ? |
| 21:00.12 | pb_ | fontenot: not currently |
| 21:00.24 | pb_ | reenoo_ would probably be the man to ask about that. |
| 21:01.20 | fontenot | pb_, how much work would it take to make some do you think (not you - me) |
| 21:01.43 | pb_ | probably not a huge amount. |
| 21:01.46 | mickeyl | nothing more than oemake opie-image |
| 21:01.50 | mickeyl | or gpe-image |
| 21:01.52 | pb_ | in theory, just set MACHINE=h3900, oemake {flavour}-image |
| 21:02.12 | pb_ | in practice, there are almost certainly some tweaks needed to the device-specific stuff. |
| 21:03.24 | fontenot | yep |
| 21:03.36 | fontenot | i wonder about ts / X |
| 21:03.55 | pb_ | mm, good question |
| 21:04.14 | pb_ | I ought to get my finger out and finish that XCALIBRATE extension. |
| 21:04.25 | CosmicPenguin | OZ worked as recently as November or December on my 5400 |
| 21:04.30 | CosmicPenguin | It couldn't have changed that much, right? :) |
| 21:04.33 | kergoth`bbl | probably just need a ts.conf for h3900, then you'd need to ts_calibrate before running the server |
| 21:04.53 | kergoth`bbl | damnit, php keeps trying to link in its own internal copy of zlib instead of ours |
| 21:05.17 | mickeyl | CosmicPenguin: OE changed pretty much. We are not even at par with OZ 3.3.6-pre1 |
| 21:05.37 | *** join/#oe eutiro_X (eutiro_X@ppp-82-84-146-55.cust-adsl.tiscali.it) |
| 21:05.49 | mickeyl | yet |
| 21:05.56 | CosmicPenguin | mickeyl: right, but kdrive + tslib should at least be stable, I would think |
| 21:06.17 | CosmicPenguin | But I was half joking, really |
| 21:08.55 | mickeyl | good good. it's a sensible topic. i didn't expect it would take us _that_ long to get into release mode. |
| 21:09.42 | fontenot | pb_, I bet lots of ppl would appreciate it :-) |
| 21:09.52 | kergoth`bbl | well, there were a number of reasons for that. too many packages and too little manpower, primarily. plus the fact that OE itself has been evolving with us as we better determined our needs |
| 21:10.51 | CosmicPenguin | Its building up some decent steam now... though I expect some more growing pains as us non Strongarm people throw monkeywrenches into the process... :) |
| 21:11.48 | kergoth`bbl | checking for ZLIB support... yes |
| 21:11.48 | kergoth`bbl | checking if the location of ZLIB install directory is defined... /home/kergoth/code/build-arm/tmp/staging/arm-li |
| 21:11.52 | kergoth`bbl | nux-uclibc/lib/.. |
| 21:11.54 | kergoth`bbl | checking for gzgets in -lz... yes |
| 21:11.59 | kergoth`bbl | looks just fine, yet it tries to link in a .lo from its internal copy |
| 21:12.00 | kergoth`bbl | hmm |
| 21:12.29 | fontenot | CosmicPenguin, what type of non-ARM person are you? x86 ? MIPS ? |
| 21:12.33 | pb_ | CosmicPenguin: true enough, though different architectures don't scare me much any more. Different OSs are probably going to be more of a challenge, but the only person who seemed really interested in that right now was Neil. |
| 21:13.00 | mickeyl | and we didn't even apply his patches *cough* |
| 21:13.09 | CosmicPenguin | fontenot: x86 - Geode |
| 21:13.16 | kergoth`bbl | heh, need to go back through the oe@ archives and apply them |
| 21:13.27 | kergoth`bbl | i do builds under cygwin on occasion still |
| 21:27.30 | CosmicPenguin | yeah, thats gonna be a bitch, isn't it? |
| 21:27.46 | CosmicPenguin | Maybe I should just fix the ipkg segfault instead |
| 21:27.55 | kergoth`bbl | you should be able to get something basic working pretty quickly |
| 21:27.59 | treke | is kergoth an asshole? |
| 21:28.19 | KingRobot | ibot is kergoth an asshole? |
| 21:28.21 | ibot | bugger all, i dunno, KingRobot |
| 21:30.04 | *** join/#oe Ptonyb (~root@DSL135-071.LABridge.com) |
| 21:31.44 | reenoo_ | pb_: actually I hope to get to OE stuff this weekend. I've been flooded with assignments and organizational stuff recently (making a GPU hacking/grid computing project actually happen) |
| 21:32.03 | pb_ | reenoo_: ah, cool |
| 21:34.30 | kergoth`bbl | php bypasses libtool during the compile phase, creating the .lo files manually |
| 21:34.42 | kergoth`bbl | apparently libtool 1.5+ is more picky about what a .lo is |
| 21:34.55 | lumi | Hi all. I'm building opie packages for oe. i've noticed that, in the ipks I build, the libraries don't get the versioning links (e.g. .so.1.0.0 -> .so) created. Not sure I've explained correctly... |
| 21:35.21 | kergoth`bbl | it should. what links do exist? |
| 21:35.22 | mickeyl | lumi: not needed |
| 21:35.22 | reenoo_ | pb_: unfortunately there's a new fairly evil math assignment in the queue again. even the math lecture on friday will take place (although everybody else will probably be on vacation. even the canteen is closed on friday...) |
| 21:35.27 | kergoth`bbl | oh, plugin, right |
| 21:35.33 | mickeyl | yea |
| 21:35.45 | mickeyl | lumi: those libraries are not loaded by the linker |
| 21:35.54 | mickeyl | they're loaded via directly via dlopen |
| 21:36.00 | lumi | mickeyl: oh... this is a new thing? |
| 21:36.11 | treke | lumi: remember the various plugins arent libraries in the usual sense of shared libraries |
| 21:36.14 | mickeyl | lumi: no. has been like that since the beginning |
| 21:36.18 | treke | lumi: same goes with applications |
| 21:36.20 | mickeyl | (beginning of plugin usage) |
| 21:36.57 | lumi | okay so the links were there before and they just weren't necessary.... |
| 21:37.01 | mickeyl | right |
| 21:37.08 | lumi | mickeyl: got it. |
| 21:37.09 | pb_ | reenoo_: doh |
| 21:37.27 | *** join/#oe jacques (~jacques@66.77.10.109) |
| 21:54.04 | *** join/#oe Hrw (~hrw@sd38.internetdsl.tpnet.pl) |
| 21:54.28 | Hrw | hi |
| 21:54.51 | proti | Hi |
| 21:55.45 | proti | Hrw: Had you been able to launch the qpe from the opie-image you built. |
| 21:57.28 | Hrw | anyone tried to boot opie? |
| 21:57.33 | Hrw | proti: nope - last times I only tried bootstrap-images |
| 21:57.46 | Hrw | time for 1st coffee today (it's 00:02 here) |
| 21:58.10 | mickeyl | d'oh |
| 21:58.18 | mickeyl | if i drink a coffe now, then i can't sleep until 5 |
| 21:58.22 | mickeyl | (00:02 here also) |
| 21:58.27 | CosmicPenguin | Thats probably not healthy.. :) |
| 21:59.07 | proti | Hrw: I was almost there. The qpe complained 'Transformed driver cannot connect.'. Was wondering now if tslib is needed ? |
| 21:59.17 | CosmicPenguin | doh, glibc-inital bug |
| 22:01.21 | mickeyl | proti: i got it |
| 22:01.33 | mickeyl | some missing flags in our qte build |
| 22:01.40 | mickeyl | we need to supply some defines for the collie build |
| 22:01.43 | mickeyl | also for the ipaq builds etc. |
| 22:02.29 | proti | mickeyl: Champagne !!! Have you been able to launch it ? |
| 22:02.36 | mickeyl | no |
| 22:02.44 | mickeyl | i looked at the compile log |
| 22:02.55 | mickeyl | and this is missing the symbols |
| 22:03.33 | mickeyl | let's see if i can fix that by adding some EXTRA_QMAKEVARS_POST_<machine> |
| 22:04.50 | proti | Is the tslib needed ? I can see that OZ unset it for collie-opie but set it for collie-opie-gcc3. |
| 22:05.12 | mickeyl | probably more like collie-kernel-oz |
| 22:07.45 | proti | mickeyl: ?? |
| 22:07.55 | mickeyl | gcc2 or 3 is irrelevant |
| 22:08.00 | mickeyl | kernel is relevant |
| 22:08.08 | mickeyl | embedix uses the built-in driver in qt |
| 22:08.11 | mickeyl | kernel-oz used tslib |
| 22:08.24 | mickeyl | s/uses/relied on/ |
| 22:08.37 | proti | OK. Can we use it without now ? |
| 22:08.48 | mickeyl | which kernels are we using? |
| 22:08.59 | CosmicPenguin | Whats the fix for the split packages bug in glibc-inital? |
| 22:09.13 | proti | mickeyl: openzaurus-sa_2.4.18-rmk7-pxa3-embedix20030509.oe |
| 22:09.16 | kergoth`bbl | CosmicPenguin: none of us have been able to repro it to fix it. |
| 22:09.25 | mickeyl | here you are. embedix builds rely on the built-in driver |
| 22:09.33 | mickeyl | so we need to enable that built-in driver |
| 22:09.41 | mickeyl | or switch on tslib for all builds |
| 22:09.50 | kergoth`bbl | hm? |
| 22:10.20 | Hrw | re |
| 22:10.30 | kergoth`bbl | in oz at least, we've been using tslib across the board for a while now |
| 22:10.46 | mickeyl | i wouldn't be so sure about that for collie |
| 22:10.57 | kergoth`bbl | uh? |
| 22:11.06 | kergoth`bbl | who do you think added support for sharp's collie ts structure to tslib? |
| 22:11.11 | kergoth`bbl | heh |
| 22:11.16 | treke | pb |
| 22:11.27 | CosmicPenguin | kergoth`bbl: I've got a trace |
| 22:12.03 | mickeyl | how come i can't find tslib in colloie-2.4.18-opie_image then? |
| 22:12.11 | mickeyl | (OZ defconfig) |
| 22:12.27 | proti | So, ts or not. Should it be |
| 22:12.29 | kergoth`bbl | why are you asking me? the last release i did was 3.2. |
| 22:12.35 | kergoth`bbl | we want tslib. |
| 22:12.38 | kergoth`bbl | 2003-03-04 Douglas Lowder <dlowder@earthlink.net> |
| 22:12.38 | kergoth`bbl | - Added support for COLLIE and CORGI event types (thanks to Chris Larson |
| 22:12.39 | kergoth`bbl | <PROTECTED> |
| 22:12.47 | kergoth`bbl | as of that date, its supported embedix kernels |
| 22:12.52 | mickeyl | heh then don't question my questioning to be so sure about collie in the latest releases |
| 22:13.11 | kergoth`bbl | so why'd you disable it then? |
| 22:13.11 | proti | Should tslib be enabled unconditionnaly for all targets ? |
| 22:13.17 | mickeyl | no idea |
| 22:13.19 | mickeyl | it's been a while |
| 22:13.35 | kergoth`bbl | i'm telling you the decision that was made. if you reversed it for 3.3.x, thats something i wouldnt know. |
| 22:14.07 | mjc | kergoth`bbl: that glibc-initial bug bit me again |
| 22:14.19 | mickeyl | well, it's not that it's important |
| 22:14.33 | proti | mickeyl: It is still enabled in collie-opie-gcc3. |
| 22:14.50 | mickeyl | good good |
| 22:14.53 | CosmicPenguin | I'm tracking it |
| 22:14.56 | CosmicPenguin | [Errno 2] No such file or directory: '/users/jcrouse/geode/tmp/base/glibc-initial-2.3.2+cvs20040513-r1/image/usr/lib/gconv' |
| 22:15.05 | CosmicPenguin | thats true, its in install, not image |
| 22:15.14 | mickeyl | then someone should start shipping tslib in our opie-image |
| 22:15.30 | mickeyl | no idea if that is sufficient |
| 22:15.33 | mickeyl | i'm not familiar with tslib at all |
| 22:15.47 | mickeyl | i think we should set those defines for qte anyway |
| 22:15.51 | kergoth`bbl | need the machine specific ts.conf |
| 22:15.53 | kergoth`bbl | thats it |
| 22:15.55 | pb_ | mickeyl: if the packages are linked with it, it will get shipped automatically by the dependencies. |
| 22:16.05 | CosmicPenguin | scrat that, its not anywhere in image or install |
| 22:16.06 | kergoth`bbl | yeah, qte should depend on it |
| 22:16.08 | mickeyl | pb_: right. i tend to forget that |
| 22:19.17 | mjc | Filesystem Size Used Avail Use% Mounted on |
| 22:19.17 | mjc | /dev/md0 12G 803M 11G 8% / |
| 22:19.17 | mjc | /dev/hdc1 23M 17M 4.9M 78% /boot |
| 22:19.17 | mjc | none 2.0G 0 2.0G 0% /dev/shm |
| 22:19.17 | mjc | /dev/md1 12G 4.5G 6.5G 41% /usr |
| 22:19.20 | mjc | /dev/md2 9.7G 447M 8.7G 5% /var |
| 22:19.22 | mjc | /dev/sda1 466G 33M 466G 1% /home |
| 22:19.25 | mjc | weee. |
| 22:19.44 | mjc | doh, wrong channel. sorry :p |
| 22:20.07 | Hrw | hmm.. last one sounds nice |
| 22:21.29 | mjc | Hrw: sata, hardware raid5 |
| 22:21.31 | CosmicPenguin | Ahhh... cool |
| 22:21.54 | mjc | CosmicPenguin: what was the hack I had done before? can't remember :p |
| 22:22.22 | jochen | I just recompiled init with the known working hh.org compiler and the kernel no longer crashes :-) |
| 22:22.32 | mickeyl | like i thought |
| 22:22.51 | mickeyl | something's fishy in the combination of that kernel and our toolchain |
| 22:22.52 | CosmicPenguin | From what I can see, its because of the do_package_prepend in glibc_cvs.oe, but I don't know how to cancel out a _prepend... :) |
| 22:22.56 | jochen | It's definitely a b0rked toolchain. |
| 22:23.10 | Hrw | hmm.. someone know ports 7741 & 57202 are for what? |
| 22:23.18 | CosmicPenguin | obviously, redefining the function won't work |
| 22:23.38 | mjc | I put something in _install that made it work |
| 22:23.41 | mjc | cant remember |
| 22:24.52 | CosmicPenguin | I wonder if we can put some phython checking in glibc_cvs.oe to see if its building -inital or not, and bail out gracefully |
| 22:25.12 | jochen | CosmicPenguin: PACKAGE_NO_GCONV = "1" in glibc-initial_cvs.oe fixed that for me. |
| 22:26.09 | ljp | treke: what is the url for your oe tutorial or whatever it is? |
| 22:28.28 | CosmicPenguin | sweet |
| 22:28.40 | CosmicPenguin | thats even more elegant then the crap I was thinking |
| 22:29.12 | treke | http://treke.net/oe/docs/oe-opie.html |
| 22:29.52 | treke | bits of it are out of date. oe has changed a bit since then |
| 22:31.46 | jochen | CosmicPenguin: the problem is "gconv" being prepended to PACKAGES. It is written into the install/*packages file and will be cached into oe-cache.<user>. From that moment on, the build fails with the error you mentioned. |
| 22:32.19 | CosmicPenguin | yeah, thats pretty close to what I saw |
| 22:33.04 | jochen | So it happens if you interrupt a build, change something in oe and restart, so oe-cache.<user> gets updated. |
| 22:33.10 | pb_ | jochen: ah, good spotting. so, the fix is to only append gconv if some gconv modules were actually split out. |
| 22:33.39 | jochen | Yes. PACKAGE_NO_GCONV = "1" can be used to prevent the append. |
| 22:33.49 | CIA-5 | 03mickeyl 07 * r1.643 10oe/conf/collie.conf: repair syntax error in collie.conf |
| 22:34.16 | jochen | It was the easiest fix :-) |
| 22:35.20 | jochen | CosmicPenguin: you don't run into the problem if you start with an empty /tmp and oe-cache.<user>. |
| 22:35.41 | CosmicPenguin | jochen: probably, I tried to get fancy with it |
| 22:36.11 | mjc | jochen: works |
| 22:36.20 | Hrw | jochen: by "oe-cache.<user>" you mean $CACHE? |
| 22:36.22 | jochen | For me, the problem started after 3 faild builds and clean ups. |
| 22:36.29 | jochen | Hrw: yes. |
| 22:36.44 | CosmicPenguin | Now I'm getting python errors in glibc that I know shouldn't be there |
| 22:36.50 | CosmicPenguin | And I'm not even using a cache! |
| 22:37.13 | kergoth`bbl | heh, if you've fucked with it, bk unedit glibc/glibc_cvs.oe |
| 22:37.14 | mjc | CosmicPenguin: it works fine for me |
| 22:37.22 | kergoth`bbl | thatll revert your local changes and check out a fresh copy |
| 22:37.35 | CosmicPenguin | kergoth`bbl: no bk - but I do have a fresh repo here |
| 22:37.43 | kergoth`bbl | ah |
| 22:38.20 | jochen | CosmicPenguin: CACHE or CCACHE_DIR? |
| 22:38.37 | jochen | They are different beasts. |
| 22:38.39 | CosmicPenguin | What? what are people trying to solve for me? |
| 22:38.41 | CosmicPenguin | I'm ok |
| 22:38.43 | CosmicPenguin | chill |
| 22:39.15 | CosmicPenguin | rebuilding fixes all |
| 22:39.29 | Hrw | "This file is not where BitKeeper thinks it should be." ;) |
| 22:39.56 | jochen | Hrw: don't move files between bk trees :-) |
| 22:40.15 | kergoth`bbl | basically dont fuck with the SCCS dirs in general |
| 22:41.13 | Hrw | basically saying it is my very old repo which I didn't used in last weeks so I forgot what was changed here |
| 22:48.21 | Hrw | bk question: I "bk pull" but has edited file so it fails but leave something in PENDING/ - how to update without "bk pull"? |
| 22:48.21 | proti | Night all. will see tslib tomorrow if it works. |
| 22:48.30 | Hrw | bye proti |
| 22:48.37 | treke | Hrw: You can't. |
| 22:48.43 | Hrw | ok |
| 22:48.44 | mickeyl | it won't work until i add the flags to qte |
| 22:48.45 | treke | bk pull is how you update |
| 22:48.47 | mickeyl | working on it |
| 22:48.53 | treke | it's like saying how do I update without updating. |
| 22:49.18 | kergoth`bbl | Hrw: either unedit the file and lose your changes, or check them in and create a changeset. |
| 22:49.32 | Hrw | kergoth`bbl: I know |
| 22:49.50 | Hrw | but it is >10 screens of text with "errors" at the beginning |
| 22:50.07 | kergoth`bbl | uh? yes, dont pull with local uncommitted changes |
| 22:50.11 | kergoth`bbl | not hard to understand |
| 22:50.24 | Hrw | kergoth`bbl: as I said - it is my very old bk tree |
| 22:50.29 | kergoth`bbl | so what? |
| 22:50.36 | kergoth`bbl | if your changes odnt need to be kept, unedit them |
| 22:50.41 | kergoth`bbl | bk -r unedit; bk pull |
| 22:50.42 | kergoth`bbl | ta da |
| 22:51.14 | Hrw | and few "remove local" |
| 22:51.38 | Hrw | takepatch: 841 new revisions, 0 conflicts in 279 files - done |
| 22:51.46 | pb_ | <PROTECTED> |
| 22:51.48 | pb_ | doh |
| 22:52.05 | mickeyl | pb_: i may have accidantly removed an export there |
| 22:52.22 | mickeyl | ah probably not |
| 22:52.23 | pb_ | mickeyl: it's probably upset because I didn't build a kernel |
| 22:52.27 | mickeyl | right |
| 22:52.30 | pb_ | (this is wireless-tools) |
| 22:52.35 | mickeyl | this is the new base_get_file() |
| 22:52.56 | mickeyl | which says <unknown> if the file to be read is not there |
| 22:53.01 | pb_ | do you know why it wants a path to the kernel source? |
| 22:53.10 | mickeyl | wireless tools? |
| 22:53.11 | kergoth`bbl | probably wants to know your WE version |
| 22:53.16 | mickeyl | because it needs the WE Version |
| 22:53.27 | mickeyl | includes should be sufficient though |
| 22:53.27 | pb_ | hm. it can get that from <linux/wireless.h> |
| 22:54.03 | kergoth`bbl | and it does. just also appends KERNEL_PATH/include to the include path |
| 22:54.09 | kergoth`bbl | (iirc) |
| 22:54.11 | pb_ | ah |
| 22:54.30 | pb_ | sounds like the kind of thing that should be stopped. |
| 22:54.59 | kergoth`bbl | heh, would be nice if it could manage we version differences at runtime, really |
| 22:55.01 | kergoth`bbl | :\ |
| 22:55.07 | pb_ | yeah |
| 22:56.06 | pb_ | later |
| 22:56.53 | KingRobot | How can I find out why oe isn't decompressing a tarfile before compiling? |
| 22:57.09 | kergoth`bbl | ? |
| 22:57.41 | reenoo_ | kergoth`bbl: latest wireless tools support that already |
| 22:58.07 | kergoth`bbl | reenoo_: ah really? |
| 22:58.26 | KingRobot | oemake [package] dies at do_configure, because a file is missing... When I look in tmp/base/[package], the sourcefile wasn't extracted. |
| 22:58.37 | kergoth`bbl | KingRobot: ls -l the sourcefiles in your DL_DIR. |
| 22:58.48 | kergoth`bbl | KingRobot: if it failed to unpack, something is wrong with them |
| 22:58.49 | reenoo_ | kergoth`bbl: "The main feature of the latest beta is the ability to handle multiple version of Wireless Extensions without the need to recompile or having multiple version of the tools." |
| 22:59.01 | kergoth`bbl | reenoo_: cool, what version is that? |
| 22:59.14 | KingRobot | kergoth: THe tar.gz is in DL_DIR, and a manual tar -xf works fine. |
| 22:59.28 | reenoo_ | kergoth`bbl: http://www.hpl.hp.com/personal/Jean_Tourrilhes/Linux/wireless_tools.27.pre22.tar.gz |
| 22:59.48 | kergoth`bbl | KingRobot: well dunno what to tell you. something is clearly wrong with the .oe or your environment. |
| 22:59.58 | kergoth`bbl | if the tar.gz is in SRC_URI and is intact, it will be unpacked. |
| 23:00.00 | KingRobot | Hmm, the rest of the packages are working fine. |
| 23:00.01 | kergoth`bbl | thats all there is to it |
| 23:03.25 | Hrw | reenoo_: what is it? |
| 23:03.41 | reenoo_ | D compiler |
| 23:05.38 | jochen | reenoo_: file an RFP (or better ITP) for it :-) |
| 23:06.45 | kergoth`bbl | damnit, php wants to build some things shared, some not, so it tries to bypass libtool for anything it wants non-shared, and thereby ends up creating invalid .lo files for them |
| 23:07.17 | CosmicPenguin | not all libtools are sane |
| 23:07.22 | reenoo_ | jochen: the problem is probably that it seems to be rather x86 cetric... |
| 23:07.33 | reenoo_ | centric even |
| 23:07.54 | CosmicPenguin | stupid opteron, go faster |
| 23:08.30 | kergoth`bbl | i dont see why they couldnt use libtool for compilation |
| 23:08.31 | kergoth`bbl | bleh |
| 23:08.33 | jochen | might be |
| 23:09.02 | CosmicPenguin | whew - finally |
| 23:09.19 | CosmicPenguin | Ok, time to rebuild. Tonight, Shrek 2. Tommorrow, debugging ipkg-cl on x86_64 |
| 23:09.34 | CosmicPenguin | I suspect Shrek will be somewhat more enjoyable |
| 23:09.42 | reenoo_ | hehe |
| 23:10.52 | CosmicPenguin | laters |
| 23:11.41 | Hrw | btw - modutils-collateral was removed from task-bootstrap few days ago - so how to know in which order I have to load modules on collie to not oops? |
| 23:12.54 | mickeyl | need to ask pb_ tomorrow, he removed it |
| 23:13.07 | Hrw | ok |
| 23:13.08 | reenoo_ | 'night all |
| 23:13.12 | Hrw | bye reenoo_ |
| 23:13.56 | KingRobot | kergoth: Got it... If the tmp/stamps/[package].do_unpack exists, it won't try to do it again. |
| 23:14.15 | kergoth`bbl | yep |
| 23:14.33 | kergoth`bbl | doing things behind oe's back is bad. use oebuild clean or oemake -c clean on the package rather than rm -rf'ing its stuff |
| 23:14.54 | KingRobot | Gotcha |
| 23:21.05 | CIA-5 | 03mickeyl 07 * r1.1461 10packages/ (3 files in 3 dirs): |
| 23:21.05 | CIA-5 | apply device specific patches to get keyboard and mouse support in Qt/Embedded. |
| 23:21.05 | CIA-5 | Watch the FIXME's! |
| 23:22.04 | mickeyl | mmh |
| 23:22.10 | mickeyl | s/patches/DEFINES |
| 23:22.15 | mickeyl | it's late... |
| 23:22.16 | mickeyl | g'night |
| 23:23.43 | Hrw | time to go to ircnet |
| 23:28.39 | jacques | jeez |
| 23:28.53 | jacques | conf file is practically unrecognizable since 5-6 |
| 23:32.44 | KingRobot | Hmmm, base-files needs `stat` > 2.5 |
| 23:33.16 | jacques | and things are moved - so howto no longer really valid :-\ |
| 23:34.12 | jacques | where is meta-opie.oe now ? |
| 23:34.27 | KingRobot | kergoth, are we making -native packages for old buildsystems that don't have compatible utils? |
| 23:34.49 | KingRobot | I guess, my question is, would a stat-native be applicable for this case? |
| 23:36.07 | kergoth`bbl | we have in some cases. generally its case-by-case |
| 23:36.16 | kergoth`bbl | jacques: um, in meta/. with every other meta package |
| 23:36.18 | kergoth`bbl | jacques: :) |
| 23:36.30 | kergoth`bbl | it was in the wrong place before |
| 23:37.28 | jacques | kergoth`bbl, what was the minimal meta package you mentioned a couple weeks ago ? |
| 23:37.33 | kergoth`bbl | task-bootstrap |
| 23:37.34 | jacques | task-bootstrap ? |
| 23:37.36 | kergoth`bbl | yup |
| 23:37.36 | jacques | ok thanks |
| 23:37.56 | KingRobot | kergoth: Well, base-files fails with an older stat ( option -c|--format is not available ), but works fine with a new stat (v3.3) |
| 23:38.02 | jacques | you still recommend running oemake from oe-packages-exported ? |
| 23:38.03 | kergoth`bbl | Installing PEAR environment: /home/kergoth/code/build-arm/tmp/base/php-4.3.6-r0/image/usr/lib/php/ |
| 23:38.07 | kergoth`bbl | /bin/sh: line 1: /home/kergoth/code/build-arm/tmp/base/php-4.3.6-r0/php-4.3.6/sapi/cli/php: cannot execute binary file |
| 23:38.09 | kergoth`bbl | make[1]: *** [install-pear-installer] Error 126 |
| 23:38.12 | kergoth`bbl | make: *** [install-pear] Error 2 |
| 23:38.14 | kergoth`bbl | ugh. |
| 23:38.28 | treke | hehehe |
| 23:38.35 | treke | php-native :p |
| 23:38.50 | kergoth`bbl | jacques: nah, we recommend creating a build dir, putting conf/local.conf in there, and adapting OEFILES to point at your packages dir. |
| 23:39.01 | kergoth`bbl | treke: heh, yeah, probably.. grr |
| 23:39.48 | jacques | kergoth`bbl, ok |
| 23:53.30 | *** join/#oe Ptonyb (~root@DSL135-071.LABridge.com) |