irclog2html for openzaurus on 2002.09.22

01:11:57Error403lo all
01:11:59Error403kergoth: good news!
01:12:05Error403I'm at kde.org on my PDA right now! :)
01:15:11Error403works like a charm
01:15:17Error403I broke down and bought a CF card and reader
01:18:42TheMasterMind1cool
01:18:51TheMasterMind1flashed oz 2.9.5.5 ?
01:24:57Error403well, yes
01:25:01Error403but that's not the cool part
01:25:07Error403you don't know the whole story :)
01:25:16Error403I'm using a StarTAC modem to do it
01:25:23Error403so it's a new definition of "wireless internet"
01:25:34Error403Where I get a cell signal, I get 14.4kb/s :)
01:29:18TheMasterMind1lol
01:33:31sambatronfrom where should I get the config.in ?
01:33:37TheMasterMind1also, the path for the sources has changed
01:33:57TheMasterMind1its now $(TOPDIR)/packagename/ not $(SOURCE_DIR)
01:34:08TheMasterMind1copy config.in from libtool or something
01:34:11TheMasterMind1or pcmcia-cs
01:34:18TheMasterMind1and change what's necessary
01:34:21TheMasterMind1name, depends, desc etc
01:34:28TheMasterMind1default will be n
01:34:30sambatronthere is no config.in file in there
01:34:44sambatronoh, yes
01:34:44TheMasterMind1arsenic:/zaurus/buildroot/packages# ls pcmcia-cs/config.in
01:34:45sambatronsorry
01:35:45TheMasterMind1yea, just look at the ones that have been done already and you should be able to figure it out
01:35:51sambatronsorry, what's the vi command to make a word uppercase =
01:35:52TheMasterMind1like libtool and libjpeg don't create ipks
01:35:52sambatron?
01:35:55TheMasterMind1pcmcia-cs does
01:36:05TheMasterMind1stuff like that.. each package has its own quirks
01:36:12TheMasterMind1many have patches that we apply etc
01:36:31sambatronI see
01:36:48TheMasterMind1each packagedir/ should have any patches and sources that are in the buildroot
01:36:56TheMasterMind1most of the sources are downloaded from bk/http/ftp/cvs
01:37:15TheMasterMind1like busybox is cvs, then patched by the buildroot
01:38:19TheMasterMind1i'm going to go chill for a while.. bbiaf
01:38:24sambatronyep
01:38:25sambatron:)
01:59:49sambatronwhat's the vim command to stop autoindenting ...
01:59:50sambatron?
02:01:44TheMasterMind1:se autoindent=0
02:01:45sambatronis there somebody here ? I need some help converting the makefiles
02:01:50sambatroncool :)
02:01:52TheMasterMind1im back now
02:01:52TheMasterMind1what's up
02:02:03sambatronSuposedly I have finished with it
02:02:19sambatroncould you take a look (how do I commit it ?)
02:02:25TheMasterMind1no fucking clue :D
02:02:27TheMasterMind1i'm new to bk
02:02:35TheMasterMind1and i can't figure out how to commit my changes either
02:02:40TheMasterMind1i'm waiting for kergoth to show up
02:02:42kagemushaTheMasterMind1: did you get java working?
02:02:44sambatron:)
02:02:46TheMasterMind1did you try compiling it?
02:02:51TheMasterMind1change packages/config.in
02:03:01myuuargh
02:03:02TheMasterMind1so it reads fbvncserver's config
02:03:03TheMasterMind1.in
02:03:08TheMasterMind1then run make
02:03:10myuui hate this
02:03:15TheMasterMind1it will prompt [y/N] for it
02:03:16TheMasterMind1say y
02:03:17myuurm lib*
02:03:19TheMasterMind1and see if it compiles
02:03:21TheMasterMind1etc
02:03:24myuuno space left on device
02:03:28TheMasterMind1myuu: heh
02:03:34TheMasterMind1that's sad
02:03:39sambatronMakefile:13: /Vars.make: No such file or directory
02:03:39TheMasterMind1you can delete some of the terminfo files
02:03:47myuuwhere?
02:03:50TheMasterMind1/usr/share/terminfo/
02:04:04TheMasterMind1delete all dirs except v and V
02:04:09TheMasterMind1and i think you'll be fine
02:04:16TheMasterMind1i dunno
02:04:27TheMasterMind1vt10x are the only terms that are actually used i think
02:04:46TheMasterMind1the ncurses Makefile was fixed so it doesn't produce 5megs of worthless terminfos
02:05:10TheMasterMind1sambatron: you had include $(TOPDIR)/Vars.make
02:05:10TheMasterMind1 ?
02:05:22myuuno space left on device
02:05:23TheMasterMind1did you bk get ?
02:05:26TheMasterMind1myuu: ouch
02:05:29TheMasterMind1that sucks
02:06:04myuuis there an oz version that runs off the SD
02:06:14TheMasterMind1hmm, no
02:06:18sambatronit doesn't find that file...
02:06:21TheMasterMind1you could install stuff to SD
02:06:24TheMasterMind1it'd get icky
02:06:37TheMasterMind1sambatron: does it exist?
02:06:42TheMasterMind1Vars.make in the root of your buildroot
02:06:53TheMasterMind1if it doesn't, do bk get in the root of the buildroot
02:06:53myuuargh
02:06:54TheMasterMind1and it will get it
02:07:08myuugot some links for roms anyway?
02:07:18TheMasterMind1myuu: the problem with putting opie on the sd is that if something happens, nothing will start
02:07:30TheMasterMind1i guess we could have it so you get dropped in a shell
02:07:35TheMasterMind1actually, you already do
02:07:37sambatronyes, it does exist... maybe the variable TOPDIR is not defined ?
02:07:38TheMasterMind1it might work
02:07:45TheMasterMind1sambatron: should be
02:07:52myuusomething with sd as /
02:07:54sambatronwhere is it defined ?
02:08:24TheMasterMind1what did you put for depends?
02:08:26TheMasterMind1in config.in
02:08:32sambatronlibc6
02:08:35TheMasterMind1ok
02:08:41TheMasterMind1myuu: sd as / ?
02:08:42TheMasterMind1hmm
02:08:50TheMasterMind1that won't work i don't think
02:08:54TheMasterMind1you need stuff to mount sd
02:09:00TheMasterMind1and if stuff is on sd itself, you have a problem
02:09:13TheMasterMind1you can try crow's rom
02:09:19TheMasterMind1: excuse me?
02:09:19TheMasterMind1ibot crow
02:09:22TheMasterMind1: sorry...
02:09:22TheMasterMind1ibot crows rom
02:09:25myuu=\
02:09:27TheMasterMind1google it
02:09:34TheMasterMind1: sorry...
02:09:34TheMasterMind1ibot crow rom
02:09:44well, rom is Read Only Memory
02:09:44myuuibot rom
02:10:17i guess wiki is http://www.handhelds.org/z/wiki/HandheldsWiki or http://www.zauruszone.farplanet.net/wiki/ or a user-editable web site
02:10:17TheMasterMind1ibot wiki
02:10:19TheMasterMind1look there
02:10:25TheMasterMind1sambatron: any luck?
02:10:30sambatronnopes
02:10:37sambatronwhere is that topdir defined ?
02:10:48sambatronmy shell ?
02:11:07TheMasterMind1no
02:11:11TheMasterMind1buildroot/Makefile
02:11:16TheMasterMind1the main make file
02:11:18TheMasterMind1which calls the other ones
02:11:29TheMasterMind1export TOPDIR:=${shell pwd}
02:11:29TheMasterMind1include $(TOPDIR)/Vars.make
02:11:41sambatronoh, I see I was calling directly the makefile of the fbvnc...
02:12:02TheMasterMind1oh
02:12:03TheMasterMind1don't
02:12:09TheMasterMind1add it to the main config.in
02:12:12TheMasterMind1and then run make
02:12:21TheMasterMind1it will ask if you want to compile fbvncserver
02:12:22sambatronbut the main makefile fails
02:12:22TheMasterMind1say y
02:12:26TheMasterMind1and it will compile stuff
02:12:45TheMasterMind1what does it fail on?
02:12:47TheMasterMind1what does it say
02:12:50TheMasterMind1copy/paste
02:12:53sambatronmake[2]: *** No rule to make target `/home/jorge/prog/zaurus/oz/buildroot-oz/stamps/.ipkg-x86.sourcetree', needed by `/home/jorge/prog/zaurus/oz/buildroot-oz/stamps/.ipkg-x86.binary'.  Stop.
02:13:13TheMasterMind1try
02:13:15TheMasterMind1make distclean
02:13:18TheMasterMind1and redo it
02:13:26TheMasterMind1as in, type make again
02:14:12sambatroncool. Now I have to add fbvnc to the main config.in ?
02:14:22sambatrondo I ?
02:14:31TheMasterMind1yes
02:14:36TheMasterMind1for it to recognize that it exists
02:14:39TheMasterMind1packages/config.in
02:14:50TheMasterMind1not buildroot/config.in
02:14:57TheMasterMind1 buildroot/packages/config.in
02:15:44sambatron vi packages/fbvncserver/config.in
02:15:45sambatron???
02:15:57TheMasterMind1vi packages/config.in
02:15:59myuu=/
02:16:09TheMasterMind1bk get in packages/ if that doesn't exist
02:17:28sambatronfails, I need a user and pass for busybox cvs repo
02:17:54sambatronsorry, for ipk
02:17:55TheMasterMind1anoncvs
02:18:09TheMasterMind1hmm, it should have gotten .cvspass file
02:18:12TheMasterMind1and not ask your for pass
02:18:20sambatronI have no .cvspass file
02:18:24TheMasterMind1sounds like you didn't download everything from the tree properly
02:18:36TheMasterMind1arsenic:/zaurus/buildroot# bk get
02:18:36TheMasterMind1.cvspass 1.3: 3 lines
02:18:43TheMasterMind1it gets it for me
02:19:20sambatronyes, it got it... maybe the distclean removed it
02:19:31TheMasterMind1maybe
02:19:46sambatronmake: *** /home/jorge/prog/zaurus/oz/buildroot-oz/build/ipkg-x86/C/busybox-0.60.3: No such file or directory.  Stop.
02:19:46sambatronmake: Entering an unknown directorymake: Leaving an unknown directorymake[2]: *** [/home/jorge/prog/zaurus/oz/buildroot-oz/stamps/.ipkg-x86.binary] Error 2
02:19:51sambatronnew error
02:20:05TheMasterMind1hmm
02:20:06TheMasterMind1wtf
02:20:11TheMasterMind1sounds like it didn't download from cvs
02:20:20TheMasterMind1i sugesst up and start over :D
02:20:27TheMasterMind1rm -rf stamps/.*
02:20:45TheMasterMind1that should rebuild stuff
02:21:01TheMasterMind1and maybe rm -rf build/ipkg*
02:21:15myuuTMM-play with X?
02:21:22TheMasterMind1nope
02:21:23TheMasterMind1haven't
02:21:28TheMasterMind1i'm waiting for picogui
02:21:30TheMasterMind1that's going to own
02:21:37myuureally!
02:21:42myuuany screenshots
02:21:54TheMasterMind1picogui.org
02:22:06TheMasterMind1not many zaurus screenshots
02:22:12TheMasterMind1its more of a run it anywhere gui thing
02:22:16TheMasterMind1but its really awesome
02:22:22myuuwhens the release?
02:22:31TheMasterMind1soon
02:22:35TheMasterMind1heh
02:22:37TheMasterMind1a week maybe
02:22:42myuuwowo
02:22:44TheMasterMind1its been 2.5years developing already
02:22:46sambatronsame problem
02:23:02myuuive seen it a couple times
02:23:26TheMasterMind1hmm
02:23:26myuui thought it was a frontend for pico =X
02:23:33TheMasterMind1maybe the cvs is down?
02:24:30sambatron /buildroot-oz/build/ipkg-x86/ this directory doesn'T exist
02:25:17TheMasterMind1and you rm -rf stamps/.ipkg*
02:25:18TheMasterMind1?
02:25:22sambatronyep
02:26:24sambatrongot it O:)
02:26:34sambatronI had to copy the .cvspass to my home O:)
02:27:01sambatronthis will take a while, right ?
02:28:33TheMasterMind1maybe
02:31:07sambatronguess I'm going to bed... it's 5:37 here
02:31:27TheMasterMind1k
02:31:36sambatronthanks for the insight
02:31:40TheMasterMind1np
02:31:57sambatronI'll come often here to help, I'd love to develop something for this project
02:32:02myuuhttp://picogui.org/download/puchuu_lain.jpeg
02:32:32sambatronbye
02:32:40myuubastards, appealing by using lain
02:34:02TheMasterMind1later
02:35:20TheMasterMind1smacks kergoth
02:35:33TheMasterMind1wake up you lazy whore
02:35:55MSpinawwww...OZ brings out the best in all of us:)
02:37:16myuuwell TMM you've got me excited =/
02:37:30TheMasterMind1heh
02:37:41TheMasterMind1btw, scanline is the picogui guy
02:37:48TheMasterMind1you can talk to him when he's around
02:38:09TheMasterMind1someone gimme kergoth's phone #
02:57:15Error403kergoth: There? :)
02:58:12Error403I just want to know when a new kernel is going to be released for OZ so that I can actually run really intense applications
02:58:19Error403or have an uptime longer than a day :)
02:58:39Error403(due to the memory problems with the current kernel)
05:35:39Onyx4hey
05:35:44kergothhey
05:36:53Onyx4anything new since friday evening?
05:37:18kergothnah, i spent all day out with friends. helping people move, seeing a movie, that sort of thing
05:37:28kergothright now i'm beating on the buildsystem, seeing if i can make it fall on its face
05:37:32Onyx4I went to see a race today at laguna seca
05:37:52kergothnice
05:38:40Onyx4I need to upgrade my opie, it's so old compared to what they are running now
05:41:03Onyx4we would need tmake for the build system first
05:41:13Onyx4then maybe a few other things, I think pam is required by opie
05:41:53kergothOnyx4: nope
05:42:10kergothOnyx4: quick -D on the compile line and opie-login uses normal auth methods
05:42:13kergothOnyx4: :)
05:42:18kergothOnyx4: the patch is in packages/opie
05:42:32Onyx4hehe ok
05:42:56Onyx4oh looks like they are building opieirc application
05:42:58Onyx4that's kind of cool
05:43:29kergothyeah
05:43:35kergothsince zic doenst seem to be being developers
05:43:39kergother
05:43:41kergothdeveloped
05:43:42Onyx4right
05:43:43kergothdamn fingers, obey me!
06:11:21scanlineis back (gone 06:19:17)
06:11:27scanlineoh wow.. I didn't think I was gone that long
06:11:44kergothhehe
06:11:48scanlineIt would be easy to port xchat to Qt
07:56:49Onyx4: sorry...
07:56:49Onyx4ibot: mastermind
07:56:55i think themastermind1 is the master at the mind. or a 15 year old oz developer or a l33t haxxor
07:56:55Onyx4ibot: themastermind1
08:30:32scanlineis away: sleep or something...
09:06:32Wemblyhello?
09:06:40Wemblywhy is it when i upgrade busybox... it hates me
09:06:49Wemblyand refuses to perform an ls or anything?
09:07:10Wemblykeeps saying illegal instruction
10:42:07sambatronhello
10:51:08sambatronnobody in here ?
11:28:05TheMasterMind1hmm
11:28:09TheMasterMind1who's alive now?
11:28:28TheMasterMind1Onyx4, KeyserSoze ?
11:28:35TheMasterMind1err
11:28:36TheMasterMind1kergoth`zzz:
11:28:57sambatronhello !
11:29:04TheMasterMind1sup
11:29:18sambatronmy sources don't compile :)
11:29:37sambatronwould you like to see the error ?
11:29:45sambatronis not the fbvncserver package
11:30:19TheMasterMind1ok
11:30:21TheMasterMind1show
11:31:31sambatron(making...)
11:31:49TheMasterMind1bk pull btw
11:31:51TheMasterMind1kergoth did some work
11:32:24sambatronentry-armv.S: Assembler messages:
11:32:24sambatronentry-armv.S:899: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
11:32:24sambatronentry-armv.S:901: Error: Internal_relocation (type 208) not fixed up (IMMEDIATE)
11:32:24sambatronentry-armv.S:948: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
11:32:24sambatronentry-armv.S:949: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
11:32:24sambatronentry-armv.S:950: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
11:32:26sambatronentry-armv.S:1238: Error: Cannot represent SWI relocation in this object file format
11:32:28sambatronmake[4]: *** [entry-armv.o] Error 1
11:33:22sambatronbtw why does it recreates the dependencies everytime I run make ?
11:33:48TheMasterMind1its being stupid..
11:33:59TheMasterMind1it shouldn't, it should create stamps
11:34:03TheMasterMind1and know that it already made em
11:34:09TheMasterMind1what package is this error in?
11:34:41sambatronbuildroot-oz/build/linux/arch/arm/kernel
11:36:13TheMasterMind1hmm
11:36:15TheMasterMind1that's not good
11:36:32TheMasterMind1compiles fine for me
11:36:39TheMasterMind1has been compiling for a while
11:36:43sambatronmaybe I'm doing something wrong...
11:36:47TheMasterMind1we haven't changed anything in the kernel
11:39:15sambatronhow could I be sure that my sources are synch'ed ?
11:40:15TheMasterMind1i take taht back
11:40:19TheMasterMind1i just got the same error
11:40:24TheMasterMind1someone fucked something up
11:40:42sambatroncan everybody write to the repo ?
11:40:46TheMasterMind1btw, you do have your toolchain and stuff setup
11:40:54TheMasterMind1nope, you need access to write
11:40:55TheMasterMind1afaik
11:41:06sambatronwhat's toolchain ?
11:41:19Wemblyhey mastermind..
11:41:28Wemblywtf is up with busybox revision2?
11:41:48TheMasterMind1no clue
11:41:57TheMasterMind1i'm trying to figure this pos bk crap out
11:42:00Wemblyhehe
11:42:12Wemblyyou have busybox revsision2 on yours?
11:42:17TheMasterMind1nope
11:42:20TheMasterMind1i'm running really old stuff
11:42:26Wemblybah
11:42:33TheMasterMind1can't upgrade until buildroot is working again
11:42:40Wemblywhy not?
11:43:09TheMasterMind1cuz its too much work
11:43:18Wemblywhy not use ipkg?
11:43:38TheMasterMind1where are the ipks?
11:44:52TheMasterMind1grrr
11:44:56TheMasterMind1kergoth`zzz: WAKE UP
11:48:28TheMasterMind1anyone around who knows how to work bitkeeper?
12:04:21TheMasterMind1aha
12:04:27TheMasterMind1i figured out how to edit files properly
12:04:29TheMasterMind1bk vi filename
12:07:16Wemblyomg
12:07:21Wemblyyou have to edit files that way?
12:07:27TheMasterMind1yea
12:07:30Wemblyouch
12:07:32TheMasterMind1or do bk edit filename
12:07:34TheMasterMind1and then edit
12:07:38TheMasterMind1and then somehow unlock the file
12:07:42TheMasterMind1can't figure out how to unlock
12:07:45Wemblybk release?
12:16:48TheMasterMind1nope
12:16:49TheMasterMind1its weird
12:16:54TheMasterMind1i fucked up my bk tree
12:17:01TheMasterMind1ran lots of weird commands
12:17:05TheMasterMind1trying to figure it out
12:17:08TheMasterMind1now its all screwed up
13:00:42doQmenthi.
15:56:16jmhodgesheya kergoth
16:00:01TheMasterMind1kergoth:
16:00:02TheMasterMind1ok
16:00:04TheMasterMind13 things
16:00:08TheMasterMind11) how the hell do i use bk
16:00:17TheMasterMind12) how do i do sourcetreedepends
16:00:25TheMasterMind13) how do i patch with -p0
16:00:53TheMasterMind1also, sambatron is doing makefile portinf for us too
16:01:06TheMasterMind1i update some, but i can't figure out how to push changes
16:01:14TheMasterMind1and bk citool and al the other tools don't work
16:01:16kergoth1) bk -r edit -q .. edit files .. bk ci -y"message" filename .. eventually when you've got a bunch of files in and you want to give them to someone, or to me, you create a 'changeset' by doing bk commit
16:01:21TheMasterMind1Cannot find wish to run
16:01:21kergothbk citool works, its X though
16:01:25kergothwish is tcl
16:01:28kergothapt-get install tcl
16:01:37TheMasterMind1ok
16:01:39kergoth2) unknown, let me think about it. it'll be a bitch
16:01:51kergoth3) easy, but not added yet, give me about 15 mins
16:01:53kergoth:)
16:01:56TheMasterMind1ok
16:02:04TheMasterMind1i need sourcetreedepends too
16:02:10TheMasterMind1hmm
16:02:25kergothwell, i can give you sourcetreedepends, but it'll be . interesting
16:02:27kergothgive me a sec
16:02:29kergothbrb
16:03:02TheMasterMind1my tree is all messed up. i ran a bunch of weird commands to try to get stuff working and made a bunch of changes, but now there's a different version that i have and a different one in the repository
16:03:10TheMasterMind1also, kernel isn't compiling
16:03:13kergothokay, check this
16:03:26kergothif you edited a file, but didnt bk ci it, you run 'bk unedit filename' to throw away your changes
16:03:41TheMasterMind1ok
16:04:06kergothsourcetreedepends.. let me explain a bit here
16:04:11kergothright now, all dependencies are handled in the config.in files
16:04:13kergothwith 'depends'
16:04:16TheMasterMind1right
16:04:21kergothso that you cant say, select bash without having selected ncurses
16:04:31kergoththe problem is, I'm going to autogenerate our actual makefile dependencies from that
16:04:31TheMasterMind1ok
16:04:41kergothand i dont see a way to reflect sourcetreedepends in config.in files
16:04:44kergothyou see what i mean?
16:04:48TheMasterMind1yea
16:05:00kergothso we'll need a manual way to do it for just this one, or i need to think up a sane way to do it overall
16:05:04TheMasterMind1we could do it some icky way
16:05:09kergothI have an idea, let me test it
16:05:20TheMasterMind1can you just make sourcetree on something
16:05:30kergothcheck this
16:05:32kergothfrom the root
16:05:36kergothyou can do make sources to d/l *all* the sources
16:05:45kergothand make sourcetrees to extract everything
16:05:45TheMasterMind1hmm
16:05:48kergoth:)
16:05:52TheMasterMind1can't you do it for one package only
16:05:57TheMasterMind1then we can call that from qte's makefile
16:06:04kergothwhich means, each makefile in packages/blah has a 'sources' and 'sourcetrees' target
16:06:06TheMasterMind1make source opie or something
16:06:09TheMasterMind1exactly
16:06:12kergothso make -C $(TOPDIR)/packages/opie sourcetrees;
16:06:14kergothboom
16:06:16kergoth:)
16:06:18TheMasterMind1alright
16:06:19TheMasterMind1perfect
16:06:29TheMasterMind1fix up patching and tell me how to set that option
16:06:33kergothI made sure i had that working, cause i knew we'd run into this
16:06:36kergothokay
16:06:53TheMasterMind1and how do i commit.. need a password or rsa key setup or what?
16:07:05kergothhave yoiu figured out editing and ci'ing in your own tree yet?
16:07:07kergoth:)
16:07:09TheMasterMind1yea
16:07:12TheMasterMind1bk vi file
16:07:16TheMasterMind1i like it that way
16:07:21TheMasterMind1so other stuff is locked
16:07:29TheMasterMind1and bk ci/citool
16:07:30kergothwell, other stuff doesnt need to be locked on your personal tree
16:07:36TheMasterMind1ok
16:07:39kergothunless you give someone else access to it
16:07:42TheMasterMind1but if i make everything edit
16:07:46kergothcause then they couldnt do a damn thing, cause everything is locked
16:07:49kergothhehe
16:07:51TheMasterMind1lol
16:07:51TheMasterMind1ok
16:07:59TheMasterMind1make[4]: Entering directory `/zaurus/buildroot/build/linux/arch/arm/kernel'
16:08:00TheMasterMind1arm-linux-gcc -D__ASSEMBLY__ -D__KERNEL__ -I/zaurus/buildroot/build/linux/include -mapcs-32 -march=armv4 -mno-fpu   -c -o entry-armv.o entry-armv.S
16:08:00TheMasterMind1entry-armv.S: Assembler messages:
16:08:00TheMasterMind1entry-armv.S:899: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
16:08:00TheMasterMind1entry-armv.S:901: Error: Internal_relocation (type 208) not fixed up (IMMEDIATE)
16:08:01TheMasterMind1entry-armv.S:948: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
16:08:03TheMasterMind1entry-armv.S:949: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
16:08:07TheMasterMind1entry-armv.S:950: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
16:08:07kergothanyway, once you have a set of changes checked in with bk ci
16:08:07TheMasterMind1entry-armv.S:1238: Error: Cannot represent SWI relocation in this object file format
16:08:09TheMasterMind1make[4]: *** [entry-armv.o] Error 1
16:08:11TheMasterMind1make[4]: Leaving directory `/zaurus/buildroot/build/linux/arch/arm/kernel'
16:08:15TheMasterMind1kernel isn't compiling
16:08:17TheMasterMind1for me or sambatron
16:08:17kergothyou create a ChangeSet with bk commit/citool
16:08:20TheMasterMind1laptop:/zaurus/buildroot# bk citool
16:08:21TheMasterMind1Cannot find wish to run
16:08:23kergothTheMasterMind1: make distclean the kernel
16:08:23TheMasterMind1:\
16:08:27TheMasterMind1i installed tcl
16:08:28kergothTheMasterMind1: cd build/linux; make distclean;
16:08:32TheMasterMind1wait, no i didn't
16:08:55TheMasterMind1E: Package tcl has no installation candidate
16:09:01TheMasterMind1tcl8.3
16:09:08kergothgo with tcl8.3
16:09:10kergoththat should be fine
16:09:14kergothits what i'm using here
16:09:17TheMasterMind1k
16:09:21kergotheach makefile has a 'install' target too
16:09:28TheMasterMind1ok
16:09:30kergothmake -C packages/blah install; will install its ipk into the rootfs :)
16:09:36kergothkinda nifty, if i do say so myself
16:09:47kergothmake image calls that on every makefile before creating the initrd.bin
16:09:50TheMasterMind1also, what's up with the clone vs. get thing
16:09:56TheMasterMind1why do you have to bk get after the clone
16:10:02kergothokay check this
16:10:04kergothyou've used cvs right?
16:10:09TheMasterMind1citool still don't work
16:10:16TheMasterMind1yes, i've used cvs
16:10:22TheMasterMind1only with buildroot
16:10:24kergothwith cvs, there's one repository, and its upstream
16:10:30TheMasterMind1yea
16:10:33kergothwith bitkeeper, you use clone to copy the whole repository locally
16:10:37kergothnow you have a local repository
16:10:39TheMasterMind1alright
16:10:41kergothand you interact with it like you would with cvs
16:10:44kergothcvs co
16:10:46kergothetc
16:10:52TheMasterMind1ok
16:10:52kergothonly slightly different commands
16:11:04kergoththen you use bk pull/bk push to send/receive ChangeSets between repositories
16:11:12kergoththere's nothing special about the one on bkbits, its just another clone
16:11:23kergothbut it happens to be web accessable :)
16:11:24TheMasterMind1ok
16:11:37TheMasterMind1so clone just clones the rep, but get gets the files?
16:11:42TheMasterMind1something like that..
16:11:56kergothyou have the rep, but you need to checkout the files
16:11:59kergothlike what cvs co did
16:12:04kergothbk -r co -q will do it
16:12:04TheMasterMind1ah
16:12:05TheMasterMind1ok
16:12:07kergothand its closer to cvs
16:12:09kergoth'co'
16:12:16kergothso use that, to avoid getting confused
16:12:30TheMasterMind1aha
16:12:39TheMasterMind1it needed tk8.3
16:12:46kergothahhh forgot about that
16:12:57kergothmost things do need tcl/tk both
16:13:04TheMasterMind1yea
16:13:33TheMasterMind1whats that command that goes through all changed files and lets you comment them?
16:13:44TheMasterMind1oh, and how do i get access to commit?
16:13:45kergothwell, you can do that when you bk ci
16:13:48kergothbk ci -y "message" blah;
16:14:02kergothyou can commit now, it commits the changeset into *your* repository, it just wont be in mine
16:14:04kergothyet
16:14:17TheMasterMind1and how do i push it to the bkbits one?
16:14:35kergothbk push
16:14:38kergothbut
16:14:43kergothyou cloned teh http tree right
16:14:46TheMasterMind1yea
16:14:47kergothyou cant push to http
16:14:51kergothcause its the web you know
16:14:51TheMasterMind1oh
16:14:56kergothbut
16:15:03kergothif you change the parent of the repository to the one that'll let you push,
16:15:06TheMasterMind1ok
16:15:09TheMasterMind1what do i change it to
16:15:10kergoththen it prompts you for your key password on every bk pull
16:15:12kergothwhich is a pain in the ass
16:15:25TheMasterMind1hmm
16:15:30kergothyou can also bk push [parent] .. specify where to push
16:15:41kergothso its bk push bk://openzaurus@openzaurus.bkbits.net/buildroot
16:15:47TheMasterMind1ah
16:15:47TheMasterMind1ok
16:15:51kergoth*after* you create teh changeset with commit/citool
16:15:56TheMasterMind1and i have to setup my key and such?
16:16:00kergothand I need to add your sshv1 key
16:16:03kergothyeah
16:16:09kergothgive me a 'identity.pub' you want to use for this
16:16:29kergothbk is pretty fucking sweet, but it takes a little getting used to :)
16:18:37kergothTheMasterMind1: done, you have the ability to push
16:18:44kergothTheMasterMind1: just make sure you dont screw anything up :-P
16:19:06TheMasterMind1heh
16:19:29kergothhehe
16:19:41kergothjust like cvs, i can always revert if someone does something stupid, so its not a prob
16:20:18TheMasterMind1laptop:/zaurus/buildroot# bk ci packages/libjpeg62/Makefile
16:20:18TheMasterMind1delta: packages/libjpeg62/SCCS/s.Makefile is not locked.
16:20:20TheMasterMind1what's tha tmean?
16:20:58kergothits not locked? that means its not locked :)
16:21:07kergothif you edited it, you did it without doing a bk edit or bk vi
16:21:10kergothi.e. forced the changes
16:21:13TheMasterMind1ok
16:21:14kergothif thats the case
16:21:17kergothdo this
16:21:19kergothmv Makefile Makefile.new
16:21:23kergothbk edit Makefile
16:21:25kergothmv Makefile.new Makefile
16:21:30kergothbk ci Makefile
16:21:30TheMasterMind1alright
16:21:32kergoth:)
16:24:17TheMasterMind1hmm
16:24:19TheMasterMind1when you ci
16:24:21TheMasterMind1it dissapears
16:24:40TheMasterMind1should i do openlogging?
16:24:46kergothyes, do it
16:24:49kergothoh yeah
16:25:00kergothwhen you ci, it goes into the repository, but doesnt co it for you automatically
16:25:07kergothjust bk -r co -q after ci'ing
16:25:11TheMasterMind1laptop:/zaurus/buildroot# bk push bk://openzaurus.bkbits.net/buildroot
16:25:11TheMasterMind1ERROR-BAD CMD: push_part1, Try help
16:25:13kergothor bk -r edit -q, or just bk co that one file
16:25:19kergoththats the wrong command
16:25:27kergoth[12:15] < kergoth> so its bk push bk://openzaurus@openzaurus.bkbits.net/buildroot
16:25:31TheMasterMind1oh
16:25:35kergothfyi
16:25:39kergothyou can do the patch -p0
16:25:41kergothalready
16:25:43kergothi forgot i added it
16:25:46kergothhehe
16:25:58TheMasterMind1how though?
16:26:06TheMasterMind1ok i commited some stuff
16:26:11kergoth$(Sources) = blah.tar.gz; \
16:26:15kergothblah.patch pnum=0;
16:26:18kergoth:)
16:26:32TheMasterMind1ok
16:26:38kergoth<source><space>options<semicolon>
16:26:40kergoththats the format
16:26:44kergothin general
16:26:46kergothi.e. cvs
16:26:47TheMasterMind1gotcha
16:26:55kergoth:pserver:blah@blah:/cvsroot module=blah;
16:27:22kergothwhere 'module' is a supported option in cvs
16:27:34kergoththere's others, like setting a tag or date to pull from cvs at
16:31:40kergothTheMasterMind1: btw, its $(MAKE) not $(Make)
16:31:56TheMasterMind1hmm ok
16:32:02TheMasterMind1i saw $(Make) somewhere
16:33:11kergothbk pull, corrected
16:33:16kergothhave you tested make image yet?
16:34:08kergothTheMasterMind1: btw, apmd depends on libtool. now that libtool is converted we can convert apmd and add depends LIBTOOL in its config.in
16:34:11kergoth:)
16:34:48TheMasterMind1wasn't there a command that went through all changed files..
16:34:54TheMasterMind1and asked for comments
16:34:58TheMasterMind1bk ci you have to specifiy file
16:35:27TheMasterMind1packages/qte-2.3.4/config.in not checked in, use -i flag.
16:35:29TheMasterMind1i'm so confused..
16:36:04TheMasterMind1oh, it doesn't exist
16:36:06TheMasterMind1so i bk add
16:36:24kergothyou specify per file comments when you ci
16:36:34kergothyou can also specify a comment for the whole changeset you generate
16:36:42kergothbk citool lets you change the per file comments after teh fact though
16:36:48kergothand it lets you set the changeset comment
16:36:59kergothi dont use the guis much, but they're quite nifty
16:37:16TheMasterMind1ok so i made some ci's
16:37:19TheMasterMind1now the files are gone
16:38:35kergothyeah, then bk co them again
16:38:37kergothor do what i do
16:38:42kergothi just bk -r co -q after doing a ci
16:38:50TheMasterMind1hmm
16:38:50TheMasterMind1ok
16:38:52kergothbtw, -q is silent, otherwise it'll dump every filename it co's
16:38:54kergothhehe
16:39:16TheMasterMind1yea
16:39:42TheMasterMind1i'll figure this all out eventually
16:46:47sambatronhello
16:46:59sambatronI'm back...
16:47:06kergothgreetings
16:47:37sambatrongreetings kergoth, I've benn talking to TheMasterMind1
16:47:45sambatronI would like to help on the project
16:48:04kergothIt'd be greatly appreciated
16:48:23sambatronI downloaded bitkeeper, and have the sources...
16:48:49sambatronyesterday with the help of TheMasterMind1 I converted fbvncserver Makefile to the new form
16:48:59kergothdid you convert libvncserver yet?
16:49:04kergothfbvncserver requires it :)
16:49:13kergoth'build' dependency, so to speak. uses libvncserver.a
16:49:34sambatronno I haven't
16:49:43kergoththatd be next, if you're up for it
16:49:45kergoth:)
16:50:00sambatronbut how do I test if I did it correctly with fbvncserver ?
16:50:15kergothmake config
16:50:22kergothhit 'y' for fbvncserver
16:50:25kergoththen make
16:50:28kergothand see whether it explodes
16:50:30kergoth:)
16:50:31sambatron./scripts/lkc/conf -s ./config.in
16:50:31sambatroncan't find file packages/libtool/config.in
16:51:29sambatronwhy is that ?
16:51:52kergothpester TheMasterMind1
16:51:59kergothhe forgot to add config.in for libtool :)
16:52:06kergothor forgot to include it in the changeset when he did the push
16:52:34TheMasterMind1bk pull
16:53:11kergothstarting work on qt/e eh?
16:53:23TheMasterMind1yea
16:53:38TheMasterMind1i wanna try out the new opie stuff
16:54:44sambatronmake config ... does not show me fbvncserver
16:54:47TheMasterMind1grr. this is quite annoying
16:54:56TheMasterMind1sambatron: add it to packages/config.in
16:55:03TheMasterMind1whydo i have ot bk co after every bk ci
16:55:20kergothTheMasterMind1: it doesnt automatically check it back out of your repos
16:55:25kergothTheMasterMind1: you could do a quick and dirty shell script
16:55:59kergothlike function bkci () { bk ci -y "$2" $1; bk co $1; }
16:56:03kergothput that in your .bashrc
16:56:03TheMasterMind1hmm
16:56:05sambatrondamn, all the changes I did have been overwritten !
16:56:08kergoththen run bkci filename message
16:56:15TheMasterMind1ick
16:56:20kergothsambatron: bk -r edit -q before making changes
16:56:59TheMasterMind1hmm
16:57:04TheMasterMind1is apmd supposed to be in the new format?
16:57:11kergothi dont think its been converted yet
16:57:17TheMasterMind1ok
16:57:24kergothTheMasterMind1: btw, please compile test before commit
16:57:30kergothTheMasterMind1: you had $(Make) in libtool, failed out
16:58:04TheMasterMind1yea
16:58:07TheMasterMind1i'm getting used to this still
16:58:20kergothquite a few changes all at once
16:58:24sambatronfor every file do I have to do the bk -r edit ?
16:58:27kergothnew config system, changed paths, bk..
16:58:31TheMasterMind1yes
16:58:37kergothsambatron: -r is recursive. it does bk edit on every file in the repository
16:58:37TheMasterMind1sambatron: bk edit file
16:58:41TheMasterMind1or -r recursive
16:58:46sambatronok
16:58:49TheMasterMind1hmm, that bkci needs to edit the file too
16:58:56kergothhehe
16:58:58kergothchange it then
16:59:05TheMasterMind1function bkci () { bk ci -y "$2" $1; bk co $1; bk edit $1 }
16:59:12kergothyou can just bk edit
16:59:15kergothit does the co for you
16:59:17kergoth:)
17:02:18TheMasterMind1after a bunch of ci's
17:02:22TheMasterMind1i commit then push yes?
17:02:26kergothyup
17:02:29kergothwhat i do
17:02:39kergothfunction bkpush () { bk commit -d && bk push bk://openzaurus@openzaurus.bkbits.net/buildroot; }
17:02:41kergoththat :)
17:02:49TheMasterMind1-d
17:02:51TheMasterMind1?
17:03:01kergothcommit -d uses the default changeset comment, which is a list of files you changed and each file's comment from bk ci
17:03:08kergothI just use ci to comment each file and use the default changeset comment, myself
17:03:15kergothbut i dont use bk citool
17:03:18kergothso i try to save time that way
17:03:51TheMasterMind1yea
17:03:53TheMasterMind1i don't like bk citool
17:04:15kergothyeah, it works, but its not pretty
17:04:22kergothhave you tried bk revtool yet?
17:04:24kergothkinda nifty
17:05:01TheMasterMind1not yet
17:05:38sambatronI have a stamp-rootinstall on pcmcia-cs Makefile and but in fbvncserver Makefile I have SubBinaryInstallCmds and BinaryInstallCmds
17:05:49sambatronhow does that match
17:05:50sambatron?
17:05:52TheMasterMind1intersting
17:06:10TheMasterMind1sambatron: you're supposed to update the BinaryInstall stuff to the new format
17:06:44sambatronwith a "stamp-subbinaryinstallcmds" ?
17:07:08TheMasterMind1no, there's no subbinary
17:07:20sambatronor as "stamp-rootinstall" and "stamp-subrootinstall"
17:07:22sambatron?
17:08:11kergothsambatron: read CONVERSIOn
17:08:18sambatronI have
17:08:21kergothsambatron: packages that use SubBinaryInstallCmds are multi-ipk builds.
17:08:50kergothsambatron: split the commands in SubBinaryInstallCmds into seperate $(stamp-packagename-rootinstall) targets
17:08:58TheMasterMind1look at ncurses
17:09:01kergothsambatron: then make $(stamp-packagename-ipk) targets depend on them
17:09:01TheMasterMind1it does multiple ipks
17:09:06kergothyeah, exactly
17:09:14sambatronok
17:09:53TheMasterMind1i need a way to see what files i have edited and not ci'ed yet
17:09:56TheMasterMind1i can never keep track ofthat
17:10:00kergothoh i can tell you that
17:10:04kergothbk status
17:10:23kergothwill show you how many files are edited but not ci'd, and how many are ci'd but not committed
17:10:38kergothyou can do bk -r diffs to see a recursive diff of the changes you've made but not ci'd
17:10:38TheMasterMind1     2 files modified and not checked in.
17:10:43kergothso thats how i tell what i've done so far
17:12:04sambatronI'm sorry but I can't see two different packages being created on fbvncserver...
17:12:12kergoth?
17:12:17kergothfbvncserver is one ipk
17:12:25kergothlast i checked anyway
17:12:33kergothhold up
17:12:41kergothokay
17:12:50kergothfbvncserver is structured like the multiipk builds were, but is only one ipk
17:13:03kergothin that case, you can translate SubBinaryInstallCmds directly to $(stamp-rootinstall)
17:13:14sambatronok
17:13:22kergothand swipe Description and the like from the BinaryInstallCmds and put them up top with the rest of the variables
17:15:15sambatronwhat happens with the variables that are double declared like name ?
17:15:23kergothsambatron: dont do that.
17:15:25kergothsambatron: just set it once
17:15:39sambatronwhich one is the good one ?
17:16:10kergothdo this
17:16:16kergothkeep Name at the top as fbvncserver
17:16:23kergothbut, I want the ipk called fbvncserver-oz
17:16:26kergothso do what ncurses does
17:16:34kergoth$(stamp-ipk) : Name=fbvncserver-oz
17:16:38kergoththat'll change it just for the ipk build
17:16:41kergoth:)
17:16:50sambatroncool
17:16:53kergothjust put that in along with everything else, dotn replace the existing target with it
17:18:56TheMasterMind1rofl
17:18:57TheMasterMind1http://www.google.com/search?num=20&hl=en&lr=&ie=UTF-8&oe=UTF-8&q=%22go+to+hell%22
17:19:54kergothlol
17:21:11TheMasterMind1btw, we need to put/symlink the FAQ to openzaurus.sf.net/ so they can just view it instead of going through sf's download thing
17:21:17TheMasterMind1that's really annoying
17:23:54TheMasterMind1and cloudchaser said she sent you some changes for the FAQ
17:23:56TheMasterMind1so update it
17:27:13TheMasterMind1see the new need_conversion ?
17:27:16TheMasterMind1it should be named todo now
17:31:28sambatronI get this with the make
17:31:29sambatronentry-armv.S: Assembler messages:
17:31:29sambatronentry-armv.S:899: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
17:31:30sambatronentry-armv.S:901: Error: Internal_relocation (type 208) not fixed up (IMMEDIATE)
17:31:30sambatronentry-armv.S:948: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
17:31:30sambatronentry-armv.S:949: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
17:31:31sambatronentry-armv.S:950: Error: Internal_relocation (type 210) not fixed up (OFFSET_IMM)
17:31:33sambatronentry-armv.S:1238: Error: Cannot represent SWI relocation in this object file format
17:31:43TheMasterMind1sambatron: cd build/linux; make distclean
17:31:43kergothsambatron: do make -C build/linux distclean from buildroot
17:31:50kergothhehe
17:32:22sambatronok :)
17:32:59kergothTheMasterMind1: i merged need_converting into the root TODO
17:33:02kergothTheMasterMind1: bk pull
17:33:17TheMasterMind1ok
17:34:56TheMasterMind1i think udhcpc support is done no?
17:34:59TheMasterMind1ifupdown uses it
17:35:02TheMasterMind1and busybox provides it
17:37:28kergothHave you tested it on a fresh rom?
17:37:32kergothits not done until we confirm it works :)
17:37:41TheMasterMind1heh
17:37:42TheMasterMind1ok
17:37:57TheMasterMind1i know udhcpc works
17:38:01TheMasterMind1i dunno if it works from ifupdown
17:38:23TheMasterMind1is someone actually working on improving ipkg or not/
17:38:34kergothjamey maintains it
17:38:37kergothyes, he's working on it
17:38:40TheMasterMind1hmm ok
17:38:43kergothhe works for crl
17:38:48kergothiirc
17:38:51TheMasterMind1i think i'm going to do a couple of quick patches
17:38:54TheMasterMind1add symlinks and such
17:40:28TheMasterMind1waits for compile..
17:40:57TheMasterMind1does this update from cvs now?
17:41:29kergothno
17:41:53kergothi may yank that functionality in favor of cvs export instead of cvs co and bk export instead of bk co
17:41:55kergothhmm
17:42:31ljphmm
17:43:28TheMasterMind1what's the difference?
17:43:45kergothTheMasterMind1: bk clone takes a long fucking time, especially for a linux kernel tree
17:43:55kergothTheMasterMind1: because its getting the whole repository, old versions of everything and all
17:44:01TheMasterMind1ah
17:44:08kergothTheMasterMind1: bk export and cvs export just get a certain version of the tree as of now, without CVS and BitKeeper dirs
17:44:19TheMasterMind1that's good
17:44:24TheMasterMind1but then you can't update?
17:44:26kergothit'll make the bk downloads faster, like for 2.4.7
17:44:28kergothexactly
17:44:31TheMasterMind1hmm
17:44:39kergoththen you'd have to wipe build/blah* sources/blah* stamps/.blah.source*
17:44:40TheMasterMind1you could do cvs co and bk export
17:44:48TheMasterMind1or even let it choose in the Makefile
17:44:53kergothor jsut wipe the source and sourcetree and it'll re-export again
17:44:55TheMasterMind1opie/busybox would be good to have update..
17:44:58kergoth:)
17:45:03TheMasterMind1or maybe export is fine
17:45:13TheMasterMind1exporting the whole thing everytime might be slower
17:45:24TheMasterMind1we can make it so you can choose
17:45:30TheMasterMind1based on which is faster for that package
17:45:30kergoththe initial checkout/clone will be slow as dirt the first time
17:45:33kergothbut faster afterwards
17:45:40kergothwehereas export will be a bit slower each time
17:45:43kergothits a toss up
17:45:45TheMasterMind1yea
17:45:50TheMasterMind1choosing = good
17:45:51kergothyeah, maybe i'll make it optional or something
17:45:55kergothput it in the config.in
17:45:57kergoth:)
17:46:02kergothi'll add it today
17:46:25TheMasterMind1k
17:47:03TheMasterMind1busybox needs a patch update
17:49:31sambatronchecking LD_LIBRARY_PATH variable... contains current directory
17:49:31sambatronconfigure: error:
17:49:31sambatron*** LD_LIBRARY_PATH shouldn't contain the current directory when
17:49:31sambatron*** building glibc. Please change the environment variable
17:49:31sambatron*** and run configure again.
17:49:36TheMasterMind1yea
17:49:41TheMasterMind1you need to change that va
17:49:41TheMasterMind1are
17:49:44TheMasterMind1make sure there's no :: in it
17:49:49TheMasterMind1or : at the end
17:49:54sambatronok
17:49:59kergothor unset LD_LIBRARY_PATH, unless you need it
17:50:01kergoth:)
17:50:30TheMasterMind1we need to add lib and binary paths for /mnt/card /mnt/cf and /mnt/ram to oz-base
17:50:52TheMasterMind1profile, env vars, root/.profile, ld.so.conf
17:51:03kergoth?
17:51:10kergothld.so.conf is useless unless they run ldconfig after
17:51:17kergothand they're already in profile, iirc
17:51:34TheMasterMind1i'll make sure
17:51:39TheMasterMind1ldconfig runs on bootup no?
17:51:42kergothno
17:51:55kergothldconfig runs in some of the ipk'spostinsts though, so it gets run during the Reconfiguring packages..
17:51:59TheMasterMind1does it any time after they flash and before they start usig it
17:52:04TheMasterMind1ok
17:52:07TheMasterMind1that's good enough
17:52:19kergothbut if you add it in oz-base's ld.so.conf, we need ldconfig in oz-base's postinst
17:52:26kergothi.e for people upgrading their oz-base version
17:52:27TheMasterMind1oh
17:52:28kergothto make sure it happens
17:52:28TheMasterMind1hmm
17:52:28TheMasterMind1ok
17:52:31kergothactually
17:52:33kergothyou cant do that
17:52:39kergothif they change cf cards
17:52:43kergoththe ld.so.cache will be wrong
17:52:48kergoththey'd have to run ldconfig each time they change cards
17:52:51TheMasterMind1ok
17:52:53kergoththats why we use LD_LIBRARY_PATH instead
17:52:53TheMasterMind1we can do that..
17:52:55kergothits dynamic
17:52:57kergothno
17:53:01TheMasterMind1fine
17:53:10TheMasterMind1putting it in ide and sdcontrol would be easy..
17:53:18kergothits *grossly stupid*
17:53:23kergothLD_LIBRARY_PATH exists for a reason
17:53:23TheMasterMind1yes master
17:53:25kergothfor shit like that
17:53:31kergoththe ld cache is too static
17:53:36kergothnot meant for shit changing left and right
17:56:44TheMasterMind1==> PASV ... done.    ==> RETR qt-embedded-2.3.4-snapshot-20020909.tar.gz ...
17:56:45TheMasterMind1No such file `qt-embedded-2.3.4-snapshot-20020909.tar.gz'.
17:56:56TheMasterMind1also, problem
17:57:01TheMasterMind1how do i make it do the sourcetreedepends thing
17:57:11TheMasterMind1before it applies the patches in Sources
17:59:17kergothTheMasterMind1: add $(stamp-sourcetree) : whee
17:59:19kergoththen
17:59:21kergothwhee:
17:59:30kergoth$(MAKE) -C $(TOPDIR)/packages/opie sourcetrees
17:59:31kergothhehe
17:59:56TheMasterMind1hmm
18:01:16kergothdont put any commands in the $(stamp-sourcetree)
18:01:18kergothor it'll break
18:01:25TheMasterMind1ok
18:01:32kergothbut if you just put it there to add a prerequisite for it, it'll work
18:01:44kergothit'll use the commands form Rules.make, but add the preqrequisite you put there
18:01:46kergothpretty nifty
18:02:04TheMasterMind1coool
18:02:22TheMasterMind1make pakcages/packname doesn't work
18:02:56kergoth?
18:02:58kergothyeah it does
18:03:00kergothi just did it
18:03:04kergothlemme do it agian
18:03:35kergothyep, just did it
18:03:40TheMasterMind1make[1]: *** No rule to make target `packages/qte-2.3.4'.  Stop.
18:03:49kergoththere's no rule if you didnt enable it in the config
18:03:51kergothin your .config
18:03:59TheMasterMind1but i did
18:04:03kergothif its set to y, it'll have a rule, and it'll work
18:04:13kergothdid you change .config without doing a make oldconfig maybe?
18:04:16TheMasterMind1make[1]: Entering directory `/zaurus/buildroot/packages/kernel'
18:04:16TheMasterMind1make[1]: *** No rule to make target `packages/qte-2.3.4'.  Stop.
18:04:16TheMasterMind1make[1]: Leaving directory `/zaurus/buildroot/packages/kernel'
18:04:21TheMasterMind1why is it going into kernel?
18:04:25kergothyou need to bk pull
18:04:34kergothhrm
18:04:39kergothyeah, you need to bk pull
18:04:40TheMasterMind1nothing to pull
18:04:49TheMasterMind1arsenic:/zaurus/buildroot# bk pull
18:04:49TheMasterMind1Nothing to pull from http://openzaurus.bkbits.net/buildroot/
18:04:53kergothhrm
18:04:56kergothhold on
18:04:59kergothi couldve sworn i fixed that
18:07:12kergothwhew
18:07:14kergothfound the bug
18:07:16kergothit works now, fucking cool
18:07:39TheMasterMind1k
18:07:42kergothhold up
18:07:48kergothi'll push it
18:13:03kergothTheMasterMind1: bk pull
18:13:17TheMasterMind1k
18:19:34sambatronok, I have my first problem
18:19:44sambatronO:)
18:19:53sambatrontouch /home/jorge/prog/zaurus/oz/buildroot-oz/stamps/.fbvncserver.source;
18:19:54sambatronError! /home/jorge/prog/zaurus/oz/buildroot-oz/sources/fbvnc/fbvncserver_0.1.1-2.tar.gz not found.
18:19:54kergothwhich is..?
18:20:32kergothsambatron: they're there in my tree.
18:20:38TheMasterMind1You don't have <scsi/scsi.h>
18:20:38TheMasterMind1You have <linux/blkpg.h>
18:20:39TheMasterMind1You have <linux/kd.h>
18:20:39TheMasterMind1You have <locale.h>
18:20:39TheMasterMind1You have <langinfo.h>
18:20:39kergothsambatron: thats the wrong path
18:20:39TheMasterMind1You have <sys/user.h>
18:20:41TheMasterMind1You have <asm/types.h>
18:20:43TheMasterMind1You don't have inet_aton()
18:20:45TheMasterMind1You don't have fsync()
18:20:48TheMasterMind1You don't have getdomainname()
18:20:50TheMasterMind1You don't have nanosleep()
18:20:50kergothsambatron: the sources are now in packages/[packagename]/
18:20:52TheMasterMind1You don't have personality()
18:20:54TheMasterMind1You don't have updwtmp()
18:20:55TheMasterMind1You need <linux/tqueue.h>
18:20:57TheMasterMind1You don't have <term.h>
18:20:59TheMasterMind1You don't have ncurses - I will not make ul and setterm.
18:21:02sambatronSources = $(SOURCE_DIR)/fbvnc/$(Name)_$(Version)-2.tar.gz
18:21:02sambatronSourceTree = $(BUILD_DIR)/$(Name)
18:21:03TheMasterMind1You don't have termcap - I will not make more.
18:21:06TheMasterMind1You need -lcrypt
18:21:08TheMasterMind1normal?
18:21:20kergothTheMasterMind1: what are you building?
18:21:24kergothTheMasterMind1: it shouldve found ncurses
18:21:31kergothTheMasterMind1: i'd think anyway
18:21:53TheMasterMind1no clue what's building
18:22:04kergothsambatron: yes, thats wrong
18:22:04TheMasterMind1i just noticed that on the screen
18:22:07kergothsambatron: read what i wrote again
18:22:17kergothsambatron: the sources are now in packages/[packagename]/
18:22:22TheMasterMind1sambatron: $(TOPDIR)/packages/fbvnc/thing.tar.gz
18:22:37sambatronok
18:22:38kergothupdates CONVERSIOn
18:23:36sambatronmake LIBS="-L/home/jorge/prog/zaurus/oz/buildroot-oz/output/staging/usr/lib" \
18:23:41sambatron        INCLUDES="-I/home/jorge/prog/zaurus/oz/buildroot-oz/output/staging/usr/include" \
18:23:41sambatron        CFLAGS_KERNEL="-Wall -D__KERNEL__ -I/home/jorge/prog/zaurus/oz/buildroot-oz/output/staging/usr/include" \
18:23:41sambatron        zaurus_fbvncserver zaurus_tssimd zaurus_kbdsim.o;
18:23:41sambatronmake[3]: Entering directory `/home/jorge/prog/zaurus/oz/buildroot-oz/packages/fbvncserver'
18:23:41sambatronmake[3]: *** No rule to make target `zaurus_fbvncserver'.  Stop.
18:24:37kergothsambatron: READ
18:24:39kergothsambatron: jesus
18:24:45kergothsambatron: add -C $(SourceTree) to that make line
18:25:10kergoth"Also note, that each line within a target starts at the root path
18:25:12kergoth        again, so you'll need to add -C $(SourceTree) to $(MAKE) lines, and
18:25:14kergoth        cd $(SourceTree); to the beginning of other lines like ./configure.
18:25:16kergoth"
18:25:18kergothquoted from CONVERSIOn
18:25:32TheMasterMind1is cvs <-> bk now autosynced?
18:25:42kergothTheMasterMind1: not quite, i've got a script doing it
18:25:48kergothTheMasterMind1: its just not run automatically yet
18:25:51TheMasterMind1Version = $(shell if [ -e $(DL_DIR)/$(Name)-cvsdate ]; then \
18:25:51TheMasterMind1    echo `cat $(DL_DIR)/$(Name)-cvsdate`; else date +%Y%m%d; fi)
18:25:57kergothTheMasterMind1: that isnt used atm
18:26:00TheMasterMind1i'm guessing that won't work
18:26:01TheMasterMind1he
18:26:02TheMasterMind1h
18:26:03kergothTheMasterMind1: remember, that was for our autoupdating
18:26:05kergothyeah
18:26:06TheMasterMind1yea
18:26:11TheMasterMind1so what do i make it
18:26:13TheMasterMind1just date?
18:26:24kergoththat'll work as is
18:26:26kergothit'll just always do the else
18:26:31kergothdate +%Y%m%d
18:26:34kergothhehe
18:26:40kergothfeel free to change it to $(shell date +%Y%m%d) though
18:26:43kergothfor simplicity
18:29:58sambatronmake: *** /home/jorge/prog/zaurus/oz/buildroot-oz/fbvncserver: No such file or directory.  Stop.
18:29:58sambatronmake: Entering an unknown directorymake: Leaving an unknown directorymake[2]: *** [/home/jorge/prog/zaurus/oz/buildroot-oz/stamps/.fbvncserver.binary] Error 2
18:31:37kergothsambatron: ?
18:31:39sambatronyep
18:31:47kergothyour SourceTree is wrong
18:31:50kergothor your make -C is wrong
18:32:09kergothi'd say
18:32:20kergothbut without more of the output or a copy of the makefile i couldnt say for sure
18:32:31TheMasterMind1where is output dir?
18:32:34TheMasterMind1where do i put ipks
18:32:39TheMasterMind1what var
18:32:46TheMasterMind1$(OUTPUT_DIR)/ipks ?
18:33:11kergothyeah
18:33:19kergothtake a look at the other makefiles for examples
18:33:24kergoththats also
18:33:26kergoth$(IPK_DIR)
18:33:27sambatronyou were right, my sourcetree was wrong
18:33:28kergothiirc
18:34:17sambatronmake[3]: Entering directory `/home/jorge/prog/zaurus/oz/buildroot-oz/packages/fbvncserver'
18:34:17sambatronmake[3]: *** No rule to make target `zaurus_fbvncserver'.  Stop.
18:34:25kergoth?
18:34:32kergothpackages/fbvncserver != sourcetree
18:34:47kergothSourceTree = $(BUILD_DIR)/somethingorother
18:35:01kergothdont touch SourceTree when copying it from the old format. its correct regardless
18:35:22sambatronwell it was buildroot-oz/fbvncserver ...
18:35:37kergothno
18:35:38sambatron make: *** /home/jorge/prog/zaurus/oz/buildroot-oz/fbvncserver: No such file or directory.  Stop.
18:35:39kergothit wasnt.
18:35:43sambatronthat was my sourcetree
18:35:44kergothSourceTree = $(BUILD_DIR)/$(Name)
18:35:47Onyx4mmm so you did QTe kergoth?
18:35:50kergoththats from a stock fbvncserver before you touched it
18:35:55kergothSourceTree = $(BUILD_DIR)/$(Name)
18:35:56Onyx4No such file `qt-embedded-2.3.4-snapshot-20020909.tar.gz'.
18:35:56TheMasterMind1kergoth: where should i put the code that moves the opie ipks to IPK_DIR
18:35:59TheMasterMind1under which make target
18:36:00kergoth$(BUILD_DIR) != $(TOPDIR)
18:36:06TheMasterMind1Onyx4: i'm working on qte
18:36:10TheMasterMind1gimme a bit
18:36:17Onyx4ok
18:36:31Onyx4well don't commit your changeset until it's working completely :)
18:36:32kergothTheMasterMind1: hmm
18:36:36kergothTheMasterMind1: here's an idea
18:36:38TheMasterMind1Onyx4: blah blah blah
18:36:39TheMasterMind1i don't hear you
18:36:41TheMasterMind1:D
18:36:43kergothTheMasterMind1: let me think a sec
18:37:17kergothTheMasterMind1: the makefile 'package' target calls the packages/blah/Makefile's ipk target
18:37:20kergothTheMasterMind1: so
18:37:22kergothipk : moveipks
18:37:26kergothmoveipks :
18:37:29TheMasterMind1k
18:37:31kergothmv blah blah;
18:37:36TheMasterMind1and a stamp?
18:37:38Onyx4where did the "need converting" go?
18:37:41TheMasterMind1to tell it its done
18:37:43TheMasterMind1Onyx4: TODO
18:37:44TheMasterMind1in the root
18:37:54kergothTheMasterMind1: yeah, but put the stamp in stamps/ not an ipk filename like the other $(stamp-ipk)'s are
18:37:54Onyx4ah
18:38:17kergothTheMasterMind1: like $(TOPDIR)/stamps/.$(Name).ipk or something
18:38:20TheMasterMind1hmm
18:38:25TheMasterMind1but will it check to see if its there?
18:38:28TheMasterMind1how does that work
18:38:44kergothyep, it checks
18:38:50kergothit'll work like every other stamp
18:38:54kergothdo it this way
18:38:56kergothhere
18:39:01TheMasterMind1moveipks:
18:39:02TheMasterMind1    cd $(SourceTree); sh mkipks
18:39:02TheMasterMind1    rm -rf $(IPK_DIR)/*opie*.ipk
18:39:02TheMasterMind1    rm -rf $(IPK_DIR)/qt-embedded*.ipk
18:39:02TheMasterMind1    cd $(SourceTree); mv *.ipk $(IPK_DIR)/
18:39:02kergothstamp-ipk=$(TOPDIR)/stamps/.$(Name).ipk
18:39:02TheMasterMind1    touch $(TOPDIR)/stamps/.$(Name).ipk
18:39:14kergothchange the name of that rule
18:39:16kergothfrom moveipks
18:39:18kergothto $(stamp-ipk)
18:39:21TheMasterMind1ah ok
18:39:26kergothand set stamp-ipk like i said
18:39:28kergoththen it'll check :)
18:40:06Onyx4I'm gonna do netbase
18:40:19sambatronwhat would be the variable for ./packages/fbvncserver/ ??
18:40:33TheMasterMind1make -C /zaurus/buildroot/packages/opie sourcetrees;
18:40:33TheMasterMind1Use of uninitialized value in concatenation (.) or string at /zaurus/buildroot/scripts/sources/Cvs.pm line 42.
18:40:33TheMasterMind1make[2]: Entering directory `/zaurus/buildroot/packages/opie'
18:40:33TheMasterMind1Makefile:44: *** missing separator.  Stop.
18:40:47TheMasterMind1sambatron: $(TOPDIR)/packages/fbvncserver/
18:41:50sambatronit looks like the sources are not untarred, should I untar them manually ?
18:42:22TheMasterMind1no
18:42:31TheMasterMind1what sources?
18:42:56TheMasterMind1you set the path in Sources:
18:42:58kergothTheMasterMind1: messed up Sources in opie
18:43:03TheMasterMind1that will get the tar and untar etc em
18:43:04kergothTheMasterMind1: cvs format changed
18:43:07TheMasterMind1and put them in build/
18:43:16TheMasterMind1kergoth: ok, what's the new format
18:43:18kergothTheMasterMind1: :pserver:blah@blah:/cvsroot:module changed to :pserver:blah@blah:/cvsroot module=blah;
18:43:35kergothTheMasterMind1: i standardized the means of passing options to the source scripts
18:43:45kergothTheMasterMind1: [source] [options];
18:43:47TheMasterMind1Makefile:44: *** missing separator.  Stop.
18:43:50TheMasterMind1what's that about?
18:44:01kergothTheMasterMind1: missing seperator is you probably forgot a \ within the Sources line or something
18:44:04sambatronlook this is my sources:
18:44:06sambatronSources = $(TOPDIR)/packages/fbvncserver/fbvnc/$(Name)_$(Version)-2.tar.gz
18:44:06sambatronSourceTree = $(BUILD_DIR)/$(Name)
18:44:08kergothTheMasterMind1: Sources = everything has to be on one line
18:44:21sambatronand this is the error:
18:44:24kergothsambatron: so give us some errors
18:44:26kergothyeah
18:44:28kergothheh
18:44:29sambatronmake: *** /home/jorge/prog/zaurus/oz/buildroot-oz/build/fbvncserver: No such file or directory.  Stop.
18:44:29sambatronmake: Entering an unknown directorymake: Leaving an unknown directorymake[2]: *** [/home/jorge/prog/zaurus/oz/buildroot-oz/stamps/.fbvncserver.binary] Error 2
18:44:43kergothsambatron: thats not the error
18:44:48kergothsambatron: give me something farther up in the output
18:44:55kergothsambatron: the real source of the problem
18:44:58kergothunix is deceptive
18:45:07kergothit takes a while to learn how to nail the actual source of a problem in error output
18:45:09sambatronmake[2]: Entering directory `/home/jorge/prog/zaurus/oz/buildroot-oz/packages/fbvncserver'
18:45:09sambatronmake -C /home/jorge/prog/zaurus/oz/buildroot-oz/build/fbvncserver LIBS="-L/home/jorge/prog/zaurus/oz/buildroot-oz/output/staging/usr/lib" \
18:45:09sambatron        INCLUDES="-I/home/jorge/prog/zaurus/oz/buildroot-oz/output/staging/usr/include" \
18:45:09sambatron        CFLAGS_KERNEL="-Wall -D__KERNEL__ -I/home/jorge/prog/zaurus/oz/buildroot-oz/output/staging/usr/include" \
18:45:10sambatron        zaurus_fbvncserver zaurus_tssimd zaurus_kbdsim.o;
18:45:12sambatronmake: *** /home/jorge/prog/zaurus/oz/buildroot-oz/build/fbvncserver: No such file or directory.  Stop.
18:45:14sambatronmake: Entering an unknown directorymake: Leaving an unknown directorymake[2]: *** [/home/jorge/prog/zaurus/oz/buildroot-oz/stamps/.fbvncserver.binary] Error 2
18:45:25TheMasterMind1nm
18:45:35kergothsambatron: ugh. thats not the error either
18:45:38kergothlet me guess
18:45:48kergothsambatron: rm -rf stamps/.fbvncserver.source*
18:45:50kergothsambatron: then do it again
18:46:15sambatronmuch different :)
18:46:27kergothgood good
18:46:31kergoththe stamps determine what is complete
18:46:42kergothif a stampfile exists and has been updated recently, it is done with that step
18:46:45TheMasterMind1$(stamp-rootinstall): $(stamp-staginginstall)
18:46:47TheMasterMind1    touch $@
18:46:47kergothand wont bother trying to do it again
18:46:49TheMasterMind1where is the missing seperator?
18:46:54sambatronmake[3]: Entering directory `/home/jorge/prog/zaurus/oz/buildroot-oz/build/fbvncserver'
18:46:54sambatronarm-linux-gcc -c -I/home/jorge/prog/zaurus/oz/buildroot-oz/output/staging/usr/include -o zaurus_fbvncserver.o fbvncserver.c
18:46:54sambatronfbvncserver.c:21: rfb.h: No such file or directory
18:46:54sambatronfbvncserver.c:22: keysym.h: No such file or directory
18:46:54sambatronmake[3]: *** [zaurus_fbvncserver.o] Error 1
18:47:02kergothTheMasterMind1: the missing seperator isnt on that line,i dont think
18:47:04kergothsambatron: i told you
18:47:09kergothsambatron: why doenst anyone ever listen to me?
18:47:15kergothsambatron: fbvncserver *depends* on libvncserver!
18:47:26kergothsambatron: libvncserver provides rfb.h, keysym.h, and libvncserver.a
18:47:35sambatronI'm sorry kergoth, I try to listen...
18:47:36kergothsambatron: all of which fbvncserver needs and expects to be in place in staging
18:47:42kergothsambatron: so convert libvncserver first
18:47:47sambatronok
18:47:58kergothsorry, I get frustrated
18:48:00TheMasterMind1kergoth: it says Makefile:44
18:48:04kergothparticulary when i repeat myself
18:48:11kergothTheMasterMind1: can you put that somewhere? or dcc it to me?
18:48:16kergothTheMasterMind1: make can be a bitch sometimes
18:49:02TheMasterMind1tjhsst.edu/~agupta/opie-makefile.txt
18:50:02kergothfound it
18:50:12kergothTheMasterMind1: make has to have a tab, not spaces, for indenting commands
18:50:19kergothTheMasterMind1: the touch has spaces in front of it
18:50:25kergothTheMasterMind1: told ya make can be a bitch :)
18:51:00TheMasterMind1hmm
18:51:01TheMasterMind1pos
18:51:16TheMasterMind1cuz i copy pasted
18:51:38kergothif you used y/p in vi to copy, you're fine
18:51:45kergothbut if you copied it out of a terminal in X....
18:51:47kergothhehe
18:51:54TheMasterMind1yea
18:51:54TheMasterMind1heh
18:52:01TheMasterMind1oops
18:52:29TheMasterMind1hmm
18:52:30TheMasterMind1for make image
18:52:38TheMasterMind1will it install the opie ipks from IPK_DIR
18:52:40TheMasterMind1?
18:52:44kergothoh shit, good call
18:52:47kergothhold up, lemme think
18:52:49kergothoh yeah
18:52:51kergoththis is easy
18:52:53kergothokay
18:52:57kergothdid you look at ncurses recently
18:52:59kergoth?
18:53:01TheMasterMind1no
18:53:04kergothat the bottom, there's a ipkinstall : line
18:53:06kergothtake a look
18:53:08kergoththats how you specify what to install
18:53:09TheMasterMind1k
18:53:29TheMasterMind1hmm ok
18:53:33TheMasterMind1what's this do
18:53:47kergothit tells it what ipks to install into the rootfs
18:54:03kergothblah.control tells it to install $(IPK_DIR)/blah_*_arm.ipk
18:54:09kergothso
18:54:11kergothwhat we have to do
18:54:14TheMasterMind1hmm
18:54:14kergothis use $(shell
18:54:21kergothto get a file list of ipks in opie's build dir
18:54:25kergothafter they get generated
18:54:27kergoththen
18:54:32kergothparse the ipk filename to get the packagename
18:54:41TheMasterMind1fun
18:54:43kergothi can do this part for you if you like
18:54:45kergoth:)
18:55:03TheMasterMind1yea
18:55:06TheMasterMind1you do it
18:55:17TheMasterMind1i'll commit in a few
18:55:21TheMasterMind1s/commit/push
18:55:33kergothk
18:55:37kergothjust let me know
18:57:50Onyx4netbase is done
18:58:12Onyx4do we still want to leave hosts.deny with ALL:paranoid in it?
18:58:29Onyx4and also provide a /etc/init.d/portmap script when we don't have portmapper?
18:59:01kergothOnyx4: lets have both hosts.allow and hosts.deny empty of entries
18:59:17Onyx4mmm ok so I would need to make a patch I guess?
18:59:23TheMasterMind1oh also
18:59:27TheMasterMind1oz needs to give out dhcp ips
18:59:30TheMasterMind1so that syncing works
18:59:30kergothyeah probably, or sed it within the makefile
18:59:32kergothTheMasterMind1: no, it doesnt.
18:59:35TheMasterMind1does too
18:59:39kergothTheMasterMind1: thats only if the user doesnt want to hardcode an ip in windows
18:59:42kergothTheMasterMind1: read the fucking faq
18:59:44kergothheh
18:59:44TheMasterMind1exactly
18:59:55TheMasterMind1half the dumbass users are too stupid to setup ips
18:59:58kergothits a ease of use thing, theres a definitive reason why I havent done that
19:00:00TheMasterMind1so we need dhcpd
19:00:06TheMasterMind1udhcpd would work
19:00:13kergothbecause usb networking's ip scheme will be configured via the new netsetup
19:00:26kergothand the new netsetup will use /etc/network/interfaces
19:00:33kergothnow if interfaces has a means of setting up a dhcp server, we can do it
19:00:45kergothotherwise, it'll be a pain in the ass, otherwise it'll break if they change their usb network config in the gui
19:00:54TheMasterMind1ARGH
19:00:54kergothand this is coming soon when ben finishes the rewrite in a week or so
19:00:54TheMasterMind1dammit
19:01:06TheMasterMind1Connecting to ftp.trolltech.com[213.203.58.54]:21... connected.
19:01:06TheMasterMind1Logging in as anonymous ... Logged in!
19:01:07TheMasterMind1==> SYST ... done.    ==> PWD ... done.
19:01:07TheMasterMind1==> TYPE I ... done.  ==> CWD /pub/qtopia/snapshots ... done.
19:01:07TheMasterMind1==> PASV ... done.    ==> RETR qt-embedded-2.3.4-snapshot-20020909.tar.gz ...
19:01:07TheMasterMind1No such file `qt-embedded-2.3.4-snapshot-20020909.tar.gz'.
19:01:12kergothTheMasterMind1: date
19:01:12TheMasterMind1what happened to it?
19:01:16kergothTheMasterMind1: there's probably a newer snapshot
19:01:22kergothTheMasterMind1: than 0909
19:02:30Onyx4why do we need to install binaries into staging dir?
19:02:38Onyx4when they are only needed for packaging?
19:02:47kergothwe dont
19:03:01kergothall we *really* need are things that other packages will need if they depend on this package
19:03:11kergothI just did it that way because i didnt feel like working out what things needed and what they didnt
19:03:16TheMasterMind1lol
19:03:22Onyx4oh, ok so then in netbase I should just be able to use the rootinstall and that's it
19:03:23kergothreally, shared and static libs, headers, and scripts are necessary
19:03:27kergothyeah
19:03:36TheMasterMind1kergoth: hmm, what snapshot should i use
19:03:39TheMasterMind1there's one for every day
19:03:45TheMasterMind1seems like they keep last 5 days or so
19:03:55kergothTheMasterMind1: well, you could try looking for todays
19:04:03kergothTheMasterMind1: date +%Y%m%d
19:04:04TheMasterMind1yea, but in 6 days, today's will be gone..
19:04:09TheMasterMind1so should we just use the latest?
19:04:12kergothTheMasterMind1: set it iwth $(shell
19:04:14kergothTheMasterMind1: yeah
19:04:14TheMasterMind1ok
19:04:18TheMasterMind1that works
19:05:04TheMasterMind1hm
19:05:08kergoth?
19:05:19TheMasterMind1that will need to be autoupdated too
19:05:25TheMasterMind1somehow
19:06:18TheMasterMind1trolltech ftp is sloooow...
19:07:40kergothyeah, it is
19:08:36Onyx4do we use the /etc/init.d/networking script?
19:08:42Onyx4and /etc/init.d/portmap ?
19:09:16kergothOnyx4: networking handles the spoofprotection, ecn, forwarding, syncookies options
19:09:29kergothOnyx4: i think its linked in and run, but i dont think any of them are enabled by default
19:09:32Onyx4ok but there aren't any symlinks for /etc/rc<x>
19:09:55kergoth?
19:09:57kergothcheck the postinst.
19:10:07kergothmost packages do the linking/unlinking in the postinst/prerm
19:10:29Onyx4i dont see a postinst on netbase
19:11:11kergothOnyx4: then we should add one :)
19:13:10Onyx4and should we symlink to runlevel 2 only or 2-5?
19:13:27kergothmost users only use 2
19:13:30Onyx4ok
19:13:34kergothbut we should probably link in 2-5
19:13:38kergothfor added flexibility
19:13:51ljpI like run level 11
19:14:56Onyx4mmm the update-inetd script is a perl script, guess that should not be copied :)
19:15:10Onyx4there's no perl on Z right?
19:16:36scanlineis back (gone 10:46:03)
19:16:59TheMasterMind1not yet
19:16:59Onyx4mmm the networking script is old, doesnt even support iptables
19:17:07Onyx4guess we're really not using that stuff
19:17:11TheMasterMind1i saw a perl tar.gz for the Z
19:17:16TheMasterMind1we need to add a makefile for perl
19:17:18TheMasterMind1and php-cgi
19:17:20TheMasterMind1:)
19:17:49kergothOnyx4: hmm?
19:17:53TheMasterMind1hey kergoth
19:17:55kergothOnyx4: there's a seperate iptables script
19:17:56TheMasterMind1if i'm exporting something
19:18:01TheMasterMind1or declare -x
19:18:05TheMasterMind1that won't work?
19:18:09TheMasterMind1does it all have to be on one line
19:18:12Onyx4kergoth: probably not part of netbase?
19:18:13TheMasterMind1like the cd $dir
19:18:14kergothTheMasterMind1: nope. unless you do it on the same line
19:18:18kergothTheMasterMind1: or set a makefile variable instead
19:18:23TheMasterMind1ok
19:18:26kergothTheMasterMind1: export blah=blah up in themakefile itself
19:18:28TheMasterMind1makefile var how?
19:18:29TheMasterMind1hmm
19:18:31kergothTheMasterMind1: make sure to export
19:18:33TheMasterMind1i have that, but its not working
19:18:35Onyx4so basically the stuff in netbase then should be just the /etc/services and config files but no scripts then
19:18:38kergothTheMasterMind1: just like Name and the others at the beginning of hte makefile
19:18:52kergothTheMasterMind1: those are makefile variables
19:19:07kergothOnyx4: well, the networking script, as i said
19:19:09TheMasterMind1/bin/sh: line 1: cd: /zaurus/buildroot/build/qt-2.3.4-snapshot-20020909: No such file or directory
19:19:10kergothOnyx4: iirc
19:19:13kergothOnyx4: but iptables is a seperate deal
19:19:13TheMasterMind1its trying to go to 0909
19:19:16TheMasterMind1i don't know why
19:19:31kergothOnyx4: iptables script is probably part of the iptables package
19:20:52Onyx4ok all the script does it enable up the interface and setup rp_filter
19:21:05kergothyeah. and echo's shit out to proc for forwarding and the like
19:21:19Onyx4right but that's set to "no" by default
19:21:24Onyx4        doopt spoofprotect yes
19:21:25Onyx4        doopt syncookies no
19:21:25Onyx4        doopt ip_forward no
19:21:37kergothOnyx4: yeah
19:21:42kergothOnyx4: unless set in the options file in /etc/network/
19:22:15TheMasterMind1got it
19:22:39TheMasterMind1yay
19:22:43TheMasterMind1look at qte compile!
19:22:55kergothnice
19:23:02TheMasterMind1kergoth: there's going to be a problem
19:23:04TheMasterMind1when the date changes
19:23:04kergoth?
19:23:09kergothoh
19:23:09TheMasterMind1there's a stamp for qte soures
19:23:20TheMasterMind1but the build dir and source file and all change
19:23:21kergothhm
19:23:23kergothand date isnt in the stamp
19:23:28kergothso it wont redownload
19:23:30kergothlol
19:23:31TheMasterMind1yea
19:23:37kergothhrm
19:23:45TheMasterMind1we could add date to stamp
19:23:53TheMasterMind1but then we still have old crap laying around
19:23:56kergothyeah
19:23:56TheMasterMind1though that's not that bad
19:23:59kergothhrm
19:24:08TheMasterMind1dammit
19:24:14TheMasterMind1it crapped out :(
19:24:18kergothwhere at?
19:24:38kergothTheMasterMind1: btw, did the patches from opie's qt/ dir apply fine?
19:24:45TheMasterMind1yea i think so
19:24:49TheMasterMind1ah i see the problem
19:24:51TheMasterMind1my patch is missing
19:24:55TheMasterMind1hm, what about this uic
19:24:58TheMasterMind1it has a stamp also
19:25:11kergoththats alright
19:25:16kergothuse the 0909 uic tarball
19:25:19TheMasterMind1si'll hardcode that
19:25:19kergothor whatever
19:25:22kergothyeah
19:26:00Onyx4kergoth: do you want me to add portmap as part of netbase?
19:26:04Onyx4I mean the binary
19:26:09Onyx4since it was not compiled before
19:26:27kergothOnyx4: if you do, make it a seperate ipk
19:26:32kergothOnyx4: most people wont need a rpc portmapper
19:26:43kergothOnyx4: i mean how many people are running nfs exports on their Z?
19:26:45kergothhehe
19:26:46Onyx4ok , so the portmap startup script I'll make it part of that package
19:26:56kergothyeah
19:26:57Onyx4sometimes you need portmap to mount nfs shares
19:26:58kergothexactly
19:27:08Onyx4like for IRIX systems, they actually do a call to portmapper on the client
19:27:22kergothreally? I didnt realize
19:27:27kergothwell regardless, it deserves its own ipk
19:27:31Onyx4ok cool
19:27:35kergoththats one difference between debian and familiar/oz
19:27:37TheMasterMind1make[1]: Entering directory `/zaurus/buildroot/packages/qte-2.3.4'
19:27:37TheMasterMind1Warning: /zaurus/buildroot/sources/qt-embedded-2.3.4-snapshot-20020922.tar.gz already exists.
19:27:37TheMasterMind1-aborting-
19:27:38TheMasterMind1hmm
19:27:40kergothwe need more granularity on an embedded system
19:27:43kergothTheMasterMind1: hehe
19:27:45TheMasterMind1it doesn't untar
19:27:47TheMasterMind1or apply the patches
19:27:48TheMasterMind1:\
19:27:57kergothTheMasterMind1: the stamp got wiped, but it noticed the source already exists
19:28:00kergothTheMasterMind1: touch the source stamp
19:28:00TheMasterMind1yea
19:28:06TheMasterMind1i deleted the stamp
19:28:11kergothTheMasterMind1: or remove the sources
19:28:12TheMasterMind1i want it to re-untar and apply the patches
19:28:13kergothah
19:28:20ljphow about we lump everything int one huge 32 mb binary?
19:28:21TheMasterMind1but the .tar.gz takes forever to download
19:28:28TheMasterMind1tt ftp is slow
19:28:41kergothTheMasterMind1: touch the stamp for *source*
19:28:43ljpit is on weekends
19:28:44kergothTheMasterMind1: not the one for sourcetree
19:28:48kergothTheMasterMind1: sourcetree = extract
19:28:50kergothTheMasterMind1: source = download
19:29:13TheMasterMind1make[1]: Entering directory `/zaurus/buildroot/packages/qte-2.3.4'
19:29:13TheMasterMind1touch /zaurus/buildroot/stamps/.qt.sourcetree;
19:29:44TheMasterMind1removed the dir
19:29:47TheMasterMind1in build
19:30:12kergothtouch the *source* stamp
19:30:17kergothor remove the *sourcetree* stamp
19:30:26kergoththen it'll see source is newer than sourctreee and reextract
19:30:42TheMasterMind1ah ok
19:30:44TheMasterMind1well
19:30:44TheMasterMind1hmm
19:30:49TheMasterMind1now its not applying the opie patches
19:31:14kergothwhen are those patches applied?
19:31:16TheMasterMind1|diff -bur3 src.orig/widgets/qmenubar.cpp src/widgets/qmenubar.cpp
19:31:17TheMasterMind1|--- src.orig/widgets/qmenubar.cpp      Mon Mar 18 02:11:27 2002
19:31:17TheMasterMind1|+++ src/widgets/qmenubar.cpp   Sun Sep  1 19:31:09 2002
19:31:17TheMasterMind1--------------------------
19:31:28TheMasterMind1it can't find src.orig?
19:31:29kergoththey should be applied during the sourcetree step
19:31:34kergothhm
19:31:37kergoththats odd
19:31:44TheMasterMind1they are being applied during sourcetree
19:31:46TheMasterMind1right after untar
19:33:48kergothodd
19:34:29TheMasterMind1grr
19:34:34TheMasterMind1why is that path screwed up
19:34:46TheMasterMind1someone sucks at making patches
19:35:17Onyx4what S and K number in rc.d should networking have on debian?
19:36:04kergoth /usr/src/coding/projects/user/buildroot$ find /etc/ -name \*networking
19:36:14kergothheh
19:36:16kergothits
19:36:19kergothS35 in rc0 and rc6
19:36:21kergothand S40 in rcS
19:36:23kergothno K
19:36:30TheMasterMind1how do i fix this?
19:36:30Onyx4k
19:36:34TheMasterMind1grr
19:36:39kergothOnyx4: using updatercd form /etc/default/functions?
19:36:56kergothTheMasterMind1: that patch looks fine..
19:37:02kergothTheMasterMind1: i dont see why it wouldnt apply
19:37:15TheMasterMind1because the first list is src.orig
19:37:17TheMasterMind1and not src
19:37:24TheMasterMind1if i type in src/blah/blah
19:37:25TheMasterMind1it works
19:37:27Onyx4kergoth: uh not really :)
19:37:30TheMasterMind1but it can't find src.orig/blah/blah
19:37:41kergothOnyx4: use it, its there for a reason :)
19:37:47kergothOnyx4: check one of the other packages for examples
19:37:51Onyx4kergoth: does pcmcia-cs does it?
19:37:53kergothOnyx4: iirc atd's postinst for example
19:38:05kergothOnyx4: no, pcmcia-cs is still braindead'ly symlinking within the ipktemp
19:38:08kergothOnyx4: iirc
19:38:09Onyx4ok
19:39:06kergothTheMasterMind1: wipe your build dir for qte and sourcetree stamp, then log the output of a make and send it to me
19:40:14TheMasterMind1hmm
19:40:19TheMasterMind1-p is 1 right?
19:40:26kergoth?
19:40:36kergothdepends
19:40:41kergothmost patches should be pnum=1, yes
19:40:43TheMasterMind1input is /zaurus/buildroot/build/opie/qt/qte234-for-opie091-menubar.patch
19:40:43TheMasterMind1output is /zaurus/buildroot/build/qt-2.3.4-snapshot-20020922
19:40:43TheMasterMind1initialize: module set to Patch
19:40:43TheMasterMind1SourceTreeMethod->go: module is Patch
19:40:43TheMasterMind1in_to_out: out is /zaurus/buildroot/build/qt-2.3.4-snapshot-20020922
19:40:44TheMasterMind1can't find file to patch at input line 4
19:40:46kergoth(which is default in patch btw)
19:40:46TheMasterMind1Perhaps you used the wrong -p or --strip option?
19:40:46Onyx4so networking is not in rc2,3,4,5?
19:40:48TheMasterMind1The text leading up to this was:
19:40:50TheMasterMind1--------------------------
19:40:52TheMasterMind1|diff -bur3 src.orig/widgets/qmenubar.cpp src/widgets/qmenubar.cpp
19:40:54TheMasterMind1|--- src.orig/widgets/qmenubar.cpp      Mon Mar 18 02:11:27 2002
19:40:56TheMasterMind1|+++ src/widgets/qmenubar.cpp   Sun Sep  1 19:31:09 2002
19:40:56kergothOnyx4: apparently not
19:41:00TheMasterMind1--------------------------
19:41:20kergothTheMasterMind1: push what you've got into bk and i'll take a look, if you want
19:41:21Onyx4kergoth: ok so it's mostly there to just unmount mounted filesystems then
19:41:51kergothOnyx4: ?
19:42:10Onyx4well if it's only into 0 and 6 it gets called with "stop' but never with start
19:42:21kergothOnyx4: networking != mounting
19:42:22Onyx4the networking script has calls in it to scan /proc/mounts and umount stuff I think
19:42:33kergothnope
19:42:45TheMasterMind1i think it does
19:42:46kergothits just interface manipulation and network options handling
19:42:48kergothno, it odesnt
19:42:53kergothi'm looking at it right now
19:43:17Onyx4since everything is working right now without calling it, I wonder what it would do for us
19:43:21TheMasterMind1it checks for mounts
19:43:25TheMasterMind1if nfs/samba is mounte
19:43:25TheMasterMind1d
19:43:27TheMasterMind1it aborts
19:44:00kergothohhh that, i missed it
19:44:02kergoththanks
19:44:14kergothOnyx4: you mean calling it for shutdown/reboot? or calling it at all?
19:44:35kergothOnyx4: the main advantage is that it cleanly unmounts the network mounts, and that it handles the networking options on boot
19:45:09TheMasterMind1kergoth: pushed
19:45:11Onyx4on but we should call it on boot then on rc2-5 right
19:45:12TheMasterMind1check it out
19:45:16TheMasterMind1and do that ipk thing for opie
19:45:31kergothOnyx4: its in rcS
19:45:39kergothOnyx4: rcS gets called regardless of which rc we boot into
19:45:41kergothOnyx4: iirc
19:45:42Onyx4ah
19:45:49kergothcould be wrong, its been a while since i dug into that cruft
19:45:53kergothcheck the inittab to see  for sure
19:48:08Onyx4mmm there is already a S40networking on my Z interesting
19:48:11Onyx4I wonder how it get there
19:48:29kergothaha
19:48:31kergothi bet i know!
19:48:34kergothdo me a faovr
19:48:38kergothremove it from oz-base
19:48:40kergothplease
19:48:42kergoth:)
19:48:48ljpheh the bit shovel
19:48:50TheMasterMind1hmm
19:48:52ljpwhoops
19:48:57TheMasterMind1kergoth: looked at this qte thing?
19:50:13Onyx4yep part of oz-base
19:50:24kergothOnyx4: yank it. oz-base needs trimming anyway
19:50:28kergothOnyx4: :)
19:50:51Onyx4ok so just to make sure I update this thing right, I just untar it make the change, then tar it back using ... ?
19:51:01TheMasterMind1yep
19:51:06Onyx4which tar options to preserve symlinks and ?
19:51:12TheMasterMind1tar czvf opie-base.tar.gz opie-base/
19:51:16kergothcd build; tar -czvf ../packages/oz-base/oz-base.tar.gz oz-base/
19:51:17kergothhehe
19:51:21TheMasterMind1heh
19:51:32TheMasterMind1smacks qte
19:51:50kergothTheMasterMind1: i havnet taken a look yet. let me check
19:52:31TheMasterMind1i think your patch script is broke
19:52:41kergothdoubtful
19:52:45kergothlet me check though
19:52:47TheMasterMind1that's what they all say
19:52:47kergothcould be
19:52:51kergothhehe
19:52:54kergothtrue that
19:53:11Onyx4ah found also portmap there in rcS.d
19:53:16Onyx4that's where it was called
19:53:20kergothyank that too
19:53:28kergothmove it into the portmap ipk
19:53:30kergothor what have you
19:53:42Onyx4yep
19:53:43kergothTheMasterMind1: 60k/s from ftp.trolltech.com, it'll be a bit here
19:53:47TheMasterMind1yea
19:53:48kergothwaits patiently
19:53:48TheMasterMind1slow crap
19:54:07TheMasterMind1=================================== ERROR ====================================
19:54:08TheMasterMind1takepatch: packages/qte-2.3.4/SCCS/s.Makefile is locked w/o writeable gfile?
19:54:08TheMasterMind1Applying   2 revisions to new file packages/qte-2.3.4/qte-2.3.4-patch  
19:54:10TheMasterMind1how do i fix that
19:54:12TheMasterMind1i screwed it up
19:54:34Onyx4how about hwclockfirst procps and hwlock ntpdate bootmisc, they don't have a valid link for oz-base to /etc/init.d
19:54:52kergothTheMasterMind1: move your patch out of the way, then bk unlock qte-2.3.4-patch, then bk edit qte-2.3.4-patch, then move it back
19:55:14kergothOnyx4: the hwclocks .. well, they should be in both busybox, *and* util-linux-hwclock, instead of oz-base
19:55:20kergothOnyx4: like i said, oz-base needs trimming
19:55:31Onyx4yep
19:55:38Onyx4so I take them out?
19:55:57TheMasterMind1ah much better
19:56:03kergothOnyx4: if you get them into the right places
19:56:12kergothOnyx4: i've left them there cause i havnet had time to move them into where they belong
19:56:17kergothOnyx4: there's even a bug report on sf on it
19:56:18kergothOnyx4: hehe
19:56:32kergothOnyx4: hwclock and hwclockfirst need to be in both busybox and util-linux-hwclock though, keep that in mind
19:56:53Onyx4mmm ok i'll let you deal with those, not sure I understand this
19:57:02Onyx4cause of my Z fully installed, they sitll don't point to anything those symlinks
19:57:03kergothtwo packages provide hwclock
19:57:05kergoththe stock one, and busybox
19:57:08TheMasterMind1kergoth: i think its not setting pnum properly or something
19:57:20kergothTheMasterMind1: i'm still wiating for the d/l
19:59:12TheMasterMind1cat /zaurus/buildroot/build/opie/qt/qte234-for-opie091-menubar.patch | patch -p1
19:59:15TheMasterMind1that's the command
19:59:18TheMasterMind1i bet its in the wrong dir
19:59:23kergother
19:59:27kergothi thought you set pnum to 0?
19:59:36TheMasterMind1hmm
19:59:39TheMasterMind1i did
19:59:44TheMasterMind1or did i
19:59:46TheMasterMind1it was before
19:59:51TheMasterMind1hmm
19:59:52kergothcheck Sources
19:59:58kergothmaybe its screwing it up
20:00:06kergothbut make sure the pnum=0 is on the source line for *each* patch
20:00:34TheMasterMind1yea it was wrong
20:00:36TheMasterMind1i fixed it
20:00:38TheMasterMind1but still
20:00:43TheMasterMind1i don't think its setting pnum anywhere
20:00:48TheMasterMind1i grepped and its not setting it anywhere
20:01:02TheMasterMind1even if pnum=0, pnum ends up being 1
20:01:09TheMasterMind1because of that if (!pnum) pnum=1; in there
20:01:18kergothhuh?
20:01:20kergothno
20:01:22kergother
20:01:27kergothdoh
20:01:28kergothyeah
20:01:31kergoththats stupid
20:01:35TheMasterMind1    if(!$invocant->{pnum}){$invocant->{pnum}=1;}
20:01:35kergothit shoudl be checking if its defined, not checking its value
20:01:37kergothyep
20:01:38TheMasterMind1lol
20:01:39TheMasterMind1oops
20:01:41kergothoi
20:01:44kergothbig oops
20:01:47kergothit'd work for any pnum but zero
20:01:50kergothlol
20:01:52TheMasterMind1i'll fix it
20:01:57TheMasterMind1i already edited it
20:02:00kergothyou know how to check whether something is defined in perl?
20:02:03TheMasterMind1nope
20:02:03TheMasterMind1heh
20:02:10kergothlet me get my quick reference quick
20:05:00TheMasterMind1defined
20:05:16TheMasterMind1if (defined $var)
20:07:11Onyx4so I should add also netbase to the def-configs/openzaurus ?
20:07:17kergothOnyx4: yeah
20:07:18TheMasterMind1hmm
20:07:23TheMasterMind1is it being defined to begin with?
20:07:32kergothOnyx4: course it isnt used unless you cp it to .config and run make oldconfig :)
20:07:34kergothTheMasterMind1: ?
20:07:39kergothTheMasterMind1: should be
20:07:47TheMasterMind1    if (!defined $invocant->{pnum}) { $invocant->{pnum}=1; }
20:07:54TheMasterMind1still setting to 1
20:08:48TheMasterMind1Use of uninitialized value in concatenation (.) or string at /zaurus/buildroot/scripts/sourcetrees/Patch.pm line 61.
20:08:48TheMasterMind1pnum is
20:08:52TheMasterMind1yep, its not defined to begin with
20:09:58Onyx4we dont provide libwrap on oz right?
20:10:19TheMasterMind1not yet :)
20:10:22TheMasterMind1do we need it?
20:11:33Onyx4you guys use the bk delta command?
20:11:45kergothi dont
20:11:52Onyx4what's the difference?
20:11:57kergoth?
20:12:02Onyx4between ci and delta
20:12:02kergothdifference between what?
20:12:17kergoth       For humans, easier interfaces are bk ci and bk citool.
20:12:19kergothheh
20:12:26kergothi suspect ci uses delta internally
20:12:40Onyx4ok
20:12:53Onyx4is there a way to say like "bk ci -r" do scan the tree and commit everything changed
20:13:03TheMasterMind1kergoth: fix this
20:13:03kergothbk -r ci -y"message";
20:13:12kergoth-r is a option for bk itself, not for the command
20:13:19kergothTheMasterMind1: i am, wait up
20:13:57TheMasterMind1fix it and tell me when its pushed
20:16:05Onyx4ok netbase is done
20:17:24Onyx4how do you create a new directory with bk?
20:17:38Onyx4or we can just create a file in a new directory and add it and it'll figure it out?
20:17:59kergothbk add dirname
20:18:13kergothit probably will figure it out if you add a file within a dir, never tried
20:19:13Onyx4I'm gonna do portmap
20:19:21Onyx4or we don't really need it right now
20:20:14Onyx4what would you guys think of starting sshd with inetd to save memory
20:20:38kergothbad
20:20:42kergothaccording to bzflag that causes timing issues
20:20:50Onyx4I see
20:20:54kergothi dont recall the details unfortunately
20:20:58kergothhrm
20:21:56Onyx4ok I'll do netkit-base and netkit-ftp ok
20:23:41kergothsource is /usr/src/coding/projects/user/buildroot/build/opie/qt/qte234-for-opie091-menubar.patch
20:23:44kergothoptions is pnum=0
20:23:46kergothhmm
20:26:04TheMasterMind1hmm
20:27:38TheMasterMind1so why isn't it setting
20:30:27Onyx4I'll convert the netkit-base to also use the /etc/default/functions
20:30:42kergothcool, thanks
20:31:52Onyx4any reasons for leaving discard and time enabled in inetd.conf?
20:33:33kergothnot particularly
20:33:35kergothfeel free to alter
20:33:42Onyx4ok
20:34:07TheMasterMind1fixed?
20:34:54kergothpnum was being set, btw
20:35:06TheMasterMind1so what's the prob?
20:35:16kergothit wasnt using defined in the check
20:35:21kergothhold on
20:35:28TheMasterMind1i changed that and it wasn't working still
20:35:28kergothi got it extracting both opie and qt/e without a prob
20:35:32TheMasterMind1hmm
20:35:33TheMasterMind1ok
20:35:36TheMasterMind1push changes
20:35:36kergothI changed like 4 other things in your amkefiles though
20:35:38kergothtweaks
20:35:40kergothwill od
20:35:42kergothdo
20:35:44kergothbtw
20:35:49kergothbk pull -nl is nifty
20:35:53kergothas is bk changes -R -v
20:35:53TheMasterMind1hmm wait a second
20:36:03TheMasterMind1is your qte-2.3.4 Makefile
20:36:07TheMasterMind1pnum=0 ?
20:37:13kergothholy crap
20:37:16kergothits compiling
20:37:18kergoth:)
20:37:19TheMasterMind1heh
20:37:28TheMasterMind1oh btw, fix that opie thing
20:37:35kergoththe opie patches dont work anymore
20:37:38kergothcause they changed the startup script
20:37:43kergothlet me fix that too while i'm at it
20:37:45TheMasterMind1ok
20:38:03TheMasterMind1and the opie ipk install into image thing
20:39:34kergothhm, yeah i need to think about that one
20:39:37kergothhm
20:39:39kergothheh
20:39:47TheMasterMind1copy the ncurses format
20:39:52kergothi know *how*
20:39:58kergoththe question is, how to get a list of the ipks
20:40:01kergoththey dont all start with opie-*
20:40:05kergothlike timezones
20:40:05TheMasterMind1yea
20:40:07kergothqt-e*
20:40:09kergothetc
20:40:11kergothlibopie
20:40:12TheMasterMind1you can just get *.ipk from build/opie/
20:40:13kergothblah blah
20:40:18kergoththey arent there anymore by the time we install'
20:40:21kergothwe moved them out of there
20:40:23kergothremember?
20:40:25kergothhehe
20:40:27kergoththats the point
20:40:29TheMasterMind1generate a list before?
20:40:30TheMasterMind1heh
20:40:36kergoththats what i'm trying to figure out how to do
20:40:57kergothcause we cant set the list in a makefile variable, cause it cant get the list until after the ipks are generated in the stamp-ipk target
20:41:03TheMasterMind1or
20:41:04kergothhrm
20:41:05TheMasterMind1copy not move
20:46:38kergothTheMasterMind1: waiting for qt/e to build so i can test it
20:46:41kergothTheMasterMind1: then i'll pushh
20:47:01kergothTheMasterMind1: problem.
20:47:06kergothTheMasterMind1: we dont want to preinstall all of opie.
20:47:09kergothTheMasterMind1: only some of the ipks
20:47:15kergothTheMasterMind1: all of opie wont even fit into 16mb flash
20:47:18kergothTheMasterMind1: hehe
20:48:05kergothTheMasterMind1: so even if we dont hardcode what items to compile in the config.in's, we still have to hardcode the list of items to preinstall
20:51:14kergothi hate compling qt/e
20:51:15TheMasterMind1yea
20:51:17TheMasterMind1that's true
20:51:58kergothTheMasterMind1: btw, i'm converting opie to use this config system too
20:52:09TheMasterMind1what config system?
20:52:16kergothmake config/menuconfig/xconfig
20:52:19kergoththe kernel config system
20:52:21TheMasterMind1cool
20:52:21kergothscripts/lkc
20:53:57TheMasterMind1is it compiling ok so far?
20:54:14kergothit finished
20:54:16kergothhold up
20:54:22TheMasterMind1push then
20:55:41scanlinebk push daisies.c
20:58:05Onyx4ok inetd aka netkit-base is done
20:58:36TheMasterMind1cool
20:58:40TheMasterMind1goodjob
20:59:08Onyx4I made the install a bit cleaner too like for the atd and netbase package
20:59:33Onyx4see if it builds fine for the two I just added
21:00:24TheMasterMind1k
21:01:07Onyx4networking is one of my expertise, so I'll make sure it's working fine on OZ :)
21:01:09TheMasterMind1yep
21:01:11TheMasterMind1works great
21:01:17TheMasterMind1compiles rather
21:01:33Onyx4BTW we have a very useful command now
21:01:43Onyx4#!/bin/bash
21:01:44Onyx4echo
21:01:44Onyx4ar p $1 control.tar.gz | tar -zxO ./control
21:01:44Onyx4echo
21:01:44Onyx4echo
21:01:44Onyx4ar p $1 data.tar.gz | tar tvz
21:01:46Onyx4echo
21:01:54Onyx4called "ipkgfiles" it lists whats in your ipk
21:02:00Onyx4so you can check if everything made it in
21:02:29Onyx4let me add it to the buildroot hehe
21:02:58kergothOnyx4: i was just going to suggest that :)
21:03:01kergothOnyx4: add ipkgcontrol too, will you?
21:03:07kergothOnyx4: throw em in scripts/
21:03:24Onyx4well my script displays both at same time you'll see
21:03:34Onyx4i just like it that way :)
21:04:03Onyx4ok it's in
21:05:11TheMasterMind1PATH="/opt/QtPalmtop/bin:/sbin:/usr/sbin:/usr/local/sbin:/usr/local/bin:/usr/bin:/bin:/mnt/card/opt/QtPalmtop/bin:/mnt/card/sbin:/mnt/card/usr/sbin:/mnt/card/usr/local/sbin:/mnt/card/usr/local/bin:/mnt/card/usr/bin:/mnt/card/bin:/mnt/cf/opt/QtPalmtop/bin:/mnt/cf/sbin:/mnt/cf/usr/sbin:/mnt/cf/usr/local/sbin:/mnt/cf/usr/local/bin:/mnt/cf/usr/bin:/mnt/cf/bin:/mnt/ram/opt/QtPalmtop/bin:/mnt/ram/sbin:/mnt/ram/usr/sbin:/mnt/ram/usr/loca
21:05:23TheMasterMind1hmm. it cut it off
21:05:33TheMasterMind1long
21:07:56TheMasterMind1kergoth: fixed yet?
21:08:23kergothTheMasterMind1: i'm tweaking things. i just made qt's makefile call out mkipks to generate only the qt ipks
21:08:30kergothTheMasterMind1: instead of relying on opie to generate qt's ipks, which is lame
21:09:05TheMasterMind1heh
21:09:06TheMasterMind1ok
21:09:31TheMasterMind1did qte compile ok?
21:09:44kergothyeha
21:11:26TheMasterMind1goddamit
21:12:35ljptweak
21:13:48TheMasterMind1did opie compile?
21:15:02kergothnifty, it installed qt/e into the rootfs
21:15:04kergothi didnt try opie yet
21:15:39TheMasterMind1k
21:19:18ljpmopie opie
21:22:58kergothOnyx4, TheMasterMind1: did you guys increment revision on the latest changes to oz-base, netbase, etc?
21:23:14kergothOnyx4, TheMasterMind1: we dont want to break people upgrading, when we change things like file locations
21:23:39Onyx4mmm  no
21:23:46ljpya! whip 'em into shape, kergoth!
21:23:49Onyx4but file locations aren't really changing per say
21:24:10Onyx4but looks like we need to bump up revision if we want them to see it as a new package right
21:24:14kergothyeah
21:24:19kergothit wont upgrade them until a revision boost
21:24:23kergothwhich could be good or bad
21:24:30kergothlike moving a file from one package to another
21:24:38kergothif they upgrade in the wrong order, ipkg will say there's a conflict
21:24:52kergothcause two packages will try to own the same package at the same time
21:25:01kergothto handle those situations, we need to depend on certain versions of things
21:25:16Onyx4well in the case of netkit and such, the file is just removed from oz-base, but it not put into netkit , cause it's creted by the postinst script
21:25:17kergothlike, to use the version of one package that now incldues this file, they need the version of the other package that no longer does
21:25:19kergothgot it?
21:25:22Onyx4and ipkg is not aware of it
21:25:26kergothOnyx4: that should be fine
21:25:35Onyx4but I'll keep that in mind for other stuff tough
21:25:38kergothOnyx4: just talking in general, trying to establish a policy on these things
21:25:40kergothyeah
21:26:38Onyx4so if a package doesn't have a "reivison" in the control, I should put one?
21:26:40kergothTheMasterMind1: opie config.in?
21:26:42kergothOnyx4: yeah
21:26:48Onyx4ok adding for netkit-ftp
21:28:58Onyx4hehe I don't think netkit-ftp was ever built, the makefile format was real old something not like the others
21:29:04Onyx4so I made it built now, ready to checkin
21:29:12kergothohhh the olddddddd format
21:29:15kergothhehe
21:29:54Onyx4kergoth why do we put default in config.in and also the def-config/openzaurus ?
21:30:22kergothOnyx4: i added def-config so we can add multiple sets of defaults
21:30:29kergothOnyx4: to generate a full feed, i'll want one with everything enabled, for example
21:30:31kergothOnyx4: def-configs/feed
21:30:35kergothOnyx4: so i added a oz one too
21:30:42Onyx4ok
21:31:00kergothOnyx4: in general, when configuring, the user shouldnt trust the defaults for make *config, but should instead copy the one they should use
21:31:10kergothOnyx4: or configure and select things manually
21:31:25kergothOnyx4: the defaults in config.in are just for convenience
21:32:54Onyx4k
21:34:53TheMasterMind1im' back
21:34:54TheMasterMind1what's up
21:35:57TheMasterMind1do qte/opie compile now?
21:36:02Onyx4you changed oz-base?
21:36:30kergothTheMasterMind1: opie doesnt compile yet, you didnt convert tmake
21:36:35ljpopie's always compiled ;)
21:36:36kergothTheMasterMind1: but no worries, i just did that
21:36:44kergothsmacks ljp
21:36:48Onyx4netkit-ftp is checked in
21:37:41TheMasterMind1Error! /zaurus/buildroot/packages/qte-2.3.4/qte-2.3.4-patch not found.
21:37:42TheMasterMind1Error! /zaurus/buildroot/packages/qte-2.3.4/qte-20020909-uic.tar.gz not found.
21:37:50TheMasterMind1hmm
21:38:06TheMasterMind1i think i forgot to co
21:38:37TheMasterMind1still isn't finding uic
21:39:33Onyx4we dont need net-tools anymore
21:39:41Onyx4it's provided by busybox
21:40:17Onyx4I'll mark it in todo
21:40:34TheMasterMind1kergoth: how is this uic thing supposed to work?
21:40:46kergothjust hold on
21:40:53TheMasterMind1mod->go: outfile is /zaurus/buildroot/build/qt-2.3.4-snapshot-20020909-uic
21:40:53TheMasterMind1SourceTreeModule->go: outfile is /zaurus/buildroot/build/qt-2.3.4-snapshot-20020909-uic
21:40:53TheMasterMind1SourceTreeModule->go: output is /zaurus/buildroot/build/qt-2.3.4-snapshot-20020922
21:40:53TheMasterMind1Recursing! /zaurus/buildroot/build/qt-2.3.4-snapshot-20020909-uic -> /zaurus/buildroot/build/qt-2.3.4-snapshot-20020922
21:40:53TheMasterMind1initialize: module set to Dir
21:40:54TheMasterMind1SourceTreeMethod->go: module is Dir
21:40:55kergothdood
21:40:56TheMasterMind1in_to_out: out is /zaurus/buildroot/build/qt-2.3.4-snapshot-20020922
21:40:58TheMasterMind1mod->go: outfile is /zaurus/buildroot/build/qt-2.3.4-snapshot-20020922
21:41:00TheMasterMind1cp: cannot stat `/zaurus/buildroot/build/qt-2.3.4-snapshot-20020909-uic': No such file or directory
21:41:02TheMasterMind1dir returned 256
21:41:04kergothwill you fucking wait?
21:41:04TheMasterMind1SourceTreeMethod->go->recursion: removing /zaurus/buildroot/build/qt-2.3.4-snapshot-20020909-uic
21:41:06TheMasterMind1SourceTreeMethod->go->recursion: removing /zaurus/buildroot/build/qt-2.3.4-snapshot-20020909-uic.tar
21:41:06kergothi'm fixing the opie build
21:41:08TheMasterMind1heh
21:41:12TheMasterMind1ok ok
21:41:14TheMasterMind1i'll wait
21:41:16TheMasterMind1ok
21:41:19TheMasterMind1this is qte though :)
21:41:25kergothand that was all harmless
21:41:45kergothdir returned 256 because it tried to recurse and realized something b0rked
21:41:46TheMasterMind1the name of the uic.tar.gz was wrong, i fixed it and now its doing that
21:41:47kergothits fine
21:41:56kergothit extracted fine
21:41:59TheMasterMind1ok
21:42:00kergothheh
21:42:28TheMasterMind1ah, so that's where uic goes
21:42:34TheMasterMind1so we need to set that path somewhere
21:42:39kergothwe arleady do
21:42:54kergothPATH+=$(BUILD_DIR)/tmake-1.8/bin:$(BUILD_DIR)/qt-2.3.4-snapshot-$(snapshot)/bin:$(BUILD_DIR)/mtd/util
21:43:00kergoththe qtdir/bin is in the path
21:43:06kergothnow let me work on this
21:43:34TheMasterMind1yea
21:44:04Onyx4mmm do we build ipv6 in the kernel?
21:44:56TheMasterMind1check
21:45:00TheMasterMind1i can't remember
21:45:11Onyx4because busybox is compiled with ipv6 support
21:45:25TheMasterMind1CONFIG_IPV6=y
21:45:26TheMasterMind1yes we do
21:45:29ljpuses ipv7
21:45:42TheMasterMind1heh
21:45:45Onyx4mmm ok none of the config scripts are really setup to use that
21:45:53TheMasterMind1heh
21:45:58TheMasterMind1i don't think anyone uses ipv6
21:45:59Onyx4that's probably a waste of resources, but heh
21:46:07TheMasterMind1yea
21:46:40kergothwaste of resources?
21:46:42kergothwhat is?
21:46:50Onyx4ipv6, I mean kernel memory
21:46:56kergothnot really
21:47:03kergothwe've had at least 8 requests for ipv6 support
21:47:07Onyx4oh ok
21:47:14kergoththough I never got around to dealing with it
21:47:15Onyx4then let's keep it in for future implementation
21:47:15TheMasterMind1kergoth: how much does the new terminfo packaging save? on /
21:47:26kergothTheMasterMind1: about 4.5 megs
21:47:29TheMasterMind1sweet
21:47:32kergothyep
21:47:36TheMasterMind1that's awesome
21:47:38kergothcourse thats compresses
21:47:41kergothcompressed
21:47:42ljpdoes that mean I'll need to update opieftp to use ipv6 functions?
21:47:48kergoththe 16mb flash is really like 25megs or something
21:47:54kergothand it saves 5 of uncompressed space
21:48:01TheMasterMind1oh
21:48:02TheMasterMind1hmm
21:48:17kergothso you wont see as much free as you'd think, but its still a major improvement
21:48:25TheMasterMind1yea
21:48:25TheMasterMind1cool
21:48:27TheMasterMind1hows opie coming
21:48:32Onyx4kergoth: how do you know you exceeded the jffs2 size?
21:48:45kergothOnyx4: right now we dont check, which is bad
21:48:59kergothOnyx4: we should be sanity checking the initrd.bin size to make sure it never exceeds the max
21:49:00Onyx4cause we pad up to 14, so it's hard to know how much we really have :)
21:49:06kergothwell, that 14 is max
21:49:13kergothif your initrd.bin gets larger than that
21:49:17kergothyou've gone too far :)
21:49:22Onyx4hehe ok
21:49:38kergothTheMasterMind1: did you push that uic naming change in the qte makefile upstream yet?
21:49:46TheMasterMind1yep
21:50:00TheMasterMind1and the path fixed
21:50:06TheMasterMind1since its in 0909
21:50:14TheMasterMind1not the latest
21:50:16kergothk
21:51:07scanlineis away: dinnertime
21:52:13Onyx4what depends on libapm.so ?
21:52:50kergothOnyx4: apm
21:52:52kergothOnyx4: :)
21:53:00Onyx4ok thought opie or something used that
21:53:05kergothhrm
21:53:15Onyx4I'm gonna convert apmd now
21:53:16kergothi think opie calls apm or uses ioctl()s on the apm_bios device
21:53:18kergothbut we can check
21:53:20kergothokay
21:53:46ljpsandman supposedly changed the ioctl calls to mimic the apm source
21:54:08kergothyeah, he and i were comparing them trying to find the source of the suspend issue
21:54:12ljpI think that part will need more testing
21:54:18kergothbut does opie use apm_bios to check the battery status?
21:54:23kergothi think it does
21:54:33kergothoh sweet
21:54:38kergothgiFTcurs shows bitrate now
21:54:39Onyx4what else needs libtool so I can check an example of dep
21:54:40ljpit disables /dev/apm_bios
21:55:05kergothljp: it uses apm_bios.
21:55:16kergothljp: in multiple places, it uses ioctl calls to talk to it
21:55:24kergothljp: what it disables is the kernel handling of the power off switch
21:55:58ljpahh
21:56:24Onyx4I can put two lines like this
21:56:28Onyx4depends LIBC6
21:56:31ljphmm that part is no longer in core/launcher/main.cpp
21:56:31Onyx4depends LIBTOOl
21:56:32Onyx4?
21:56:34TheMasterMind1yea i think so
21:56:39kergothOnyx4: it uses libtool in its compilation
21:56:45kergothOnyx4: and we cant rely on the one for the build system
21:56:46TheMasterMind1btw, i've noticed
21:56:50TheMasterMind1the suspend issue doesn't happen
21:56:53kergothOnyx4: so its really a 'build' dependency
21:56:55TheMasterMind1when i do it via the start menu thing
21:57:01TheMasterMind1but when it autosuspends, it does happen
21:57:01kergothTheMasterMind1: 'smart menu thing'
21:57:04kergothTheMasterMind1: ?
21:57:14kergothTheMasterMind1: oh suspend via start menu, gotcha
21:57:14TheMasterMind1the Q in the bottom leftcorner
21:57:16TheMasterMind1themenu
21:57:16kergothheh
21:58:07Onyx4kergoth: right but apmd needs libtool
21:58:13kergothof course it does
21:58:15Onyx4kergoth: so I need to put depends libtool I guess
21:58:19kergothOnyx4: thats what i'm talking about, yeah exactly
21:58:22kergothOnyx4: hehe
21:58:41kergothOnyx4: unfortunately the kernel config system doesnt seperate depends and builddepends, so we just depend on it
21:58:42Onyx4and I can put two "depends" line in config.in?
21:58:44kergothOnyx4: doesnt hurt us much
21:58:53kergothhm
21:58:55kergothhold on
21:58:59kergothwe might have to do
21:59:04kergothDEPENDS = LIBTOOL && LIBC6
21:59:08kergothinstead of two lines
21:59:11kergothbut let me verify
21:59:16Onyx4or I guess if I put "libtool" since libtool depends on libc6 it'll do a recursive dependancy? :)
22:00:31TheMasterMind1heh
22:00:34TheMasterMind1yea that will work too
22:00:39Onyx4cool
22:00:45TheMasterMind1we need something that makes an image
22:00:47TheMasterMind1of all dependencies
22:00:48kergothlibtool doesnt and should not depend on libc6
22:00:51TheMasterMind1just cuz it would look cool
22:00:51kergothits a script
22:00:55kergothit doesnt link against libc6
22:01:01TheMasterMind1oops.
22:01:02TheMasterMind1heh
22:01:02kergothand its a build dependency, it runs on teh build machine, x876
22:01:04kergother
22:01:07kergothx86
22:01:16kergothTheMasterMind1: my dependency generator could produce a jpeg or something
22:01:19kergothTheMasterMind1: a dependency map
22:01:22TheMasterMind1that'd be awesome
22:05:33Onyx4what's this for in apmd compile?
22:05:35Onyx4-I$(CURDIR)/debian/linux-2.2/include
22:06:06kergothheh
22:06:08kergothno idea
22:06:20kergothprobably remnants of the debian patch which we apply
22:06:22kergothor something
22:06:31kergothor remnants of swiped build commands from the debian/rules file
22:06:33kergothhehe
22:06:51Onyx4is CURDIR still valid or I need to replace that by $(SourceTree) ?
22:07:02Onyx4I'll try to take it out and see
22:13:13Onyx4what is apmd started by?
22:13:18kergoth?
22:13:20kergothwe dont use apmd
22:13:29kergothapmd package is two pieces, the apm binary, and the apmd daemon
22:13:36kergothwe only use the former, not the latter
22:13:39Onyx4ah
22:13:41Onyx4gotcha
22:13:54Onyx4so I should not install apmd?
22:14:03kergothi'd split it
22:14:06kergothbuild two ipks
22:14:08kergoth:)
22:14:16kergothapmd-client apmd-server perhaps?
22:14:25kergothor if theres anything shared, add apmd-common?
22:14:27kergothhm
22:14:35Onyx4well if there is no real use for it, why confuse people and give them to option to add apm
22:14:43kergothwe may add apmd later.
22:14:47Onyx4k
22:15:00kergothit depends on whether gaining the ability to use suspend/resume scripts is worth whatever overhead it adds
22:15:13Onyx4the current apmd makefile does not install apmd
22:15:21Onyx4I don't think it builds it either
22:15:26kergothi know
22:15:31kergotheh
22:16:34Onyx4guess I need to change that name in config.in to something like "apm tools" ?
22:16:44Onyx4I wrote apm deamon hehe
22:17:02kergothyeah
22:17:04kergothsomething like that
22:17:10kergothsteal debian's description from debian/control
22:17:12kergoth:)
22:17:59Onyx4install -m 0644 debian/copyright $(TARGET_DIR)/CONTROL/copyright
22:18:01Onyx4what's this for ? :)
22:18:10kergothum, nothing?
22:18:12kergothhehe
22:18:18Onyx4ok :)
22:18:20kergothi dont even know if ipkg installs that into /usr/share
22:20:04scanlineis back (gone 00:28:57)
22:20:39scanlineis away: hiding from all the people that want to talk to him in #picogui
22:20:48TheMasterMind1scanline: port the picogui stuff in the buildroot
22:24:08TheMasterMind1kergoth: how comes opie?
22:24:32kergothTheMasterMind1: had to recompile qt/e, waiting patiently
22:24:58TheMasterMind1k
22:25:12scanlineis back (gone 00:04:33)
22:26:47Onyx4mmm how come the IPK did not strip by libapm.so.1.0.0 ?
22:26:54Onyx4s/by/my
22:27:05TheMasterMind1hmm
22:27:10TheMasterMind1does stuff strip by default or not?
22:27:20Onyx4yes
22:27:40Onyx4because of the build process
22:27:46Onyx4it finds executables files and strip the
22:27:47Onyx4it finds executables files and strip them
22:27:54TheMasterMind1ok
22:27:56kergothOnyx4: lame, we should have it running 'file' to check
22:28:02kergothOnyx4: been on my todo for a while now, keep forgetting
22:28:37Onyx4ok so I wonder if the other libs are not stripped
22:29:15kergothare they excecutable?
22:29:17kergothmost should be 755
22:29:34Onyx4oh wait that's the problem, the library is not executable
22:29:40TheMasterMind1heh
22:29:41TheMasterMind1kjsdf
22:29:46Onyx4ok let me fix that
22:29:51kergothTheMasterMind1: qt/e is still going..
22:30:00TheMasterMind1you think sharp will give us the new zaurus models so we can develop :D
22:30:10TheMasterMind1i hear release date is nov1st
22:31:18ljpno
22:32:52Onyx4so this was actually a bug in the old apmd package, the lib was set 644
22:33:01Onyx4so not stripped, I'll bump up the rev
22:33:20TheMasterMind1:d
22:33:22Onyx4can we do "make clean packagename" ?
22:33:28TheMasterMind1Onyx4 is getting good at this
22:33:36kergothOnyx4: not yet
22:33:45Onyx4apm package is done, let me check it in
22:33:56kergothTheMasterMind1: opie is compiling
22:34:01TheMasterMind1k
22:34:14TheMasterMind1does it do ipk installing into rootfs and stuff?
22:34:16TheMasterMind1the makefilethat
22:34:16TheMasterMind1is
22:34:30kergothnot yet
22:34:30TheMasterMind1s/makefilethat/makefile that/
22:34:32TheMasterMind1pl
22:34:32kergothi dont think anyway, i havent gotten that far
22:34:33TheMasterMind1ok
22:34:35kergothheh
22:35:55Onyx4a nice command I like too is "bk pending" before doing the commit see if everything is there
22:36:21kergothyeah, that and bk -r diffs are great
22:36:32kergothto see what you've done and not ci'd, and then pending to see what you've ci'd but not committed
22:36:35TheMasterMind1whoa
22:36:38TheMasterMind1what's this pending?
22:36:38kergothbk is so nifty
22:36:45kergothTheMasterMind1: shows what is pending a commit
22:36:47TheMasterMind1show me output
22:36:51kergothTheMasterMind1: things that you ci'd but didnt commit
22:36:58TheMasterMind1i see
22:36:59TheMasterMind1cool
22:37:00kergothi cant, my buildroot is making opie atm
22:37:02kergothheh
22:38:05kergothOnyx4: what do you think for individual package clean and mrproper targets?
22:38:19kergothOnyx4: I mean, mrproper should obviously wipe sources, build dir, and everything else
22:38:31kergothOnyx4: but clean... should it only wipe the sourcetree dir and leave sources in place?
22:38:43TheMasterMind1yes
22:39:35TheMasterMind1hmm
22:39:38TheMasterMind1i want more qt fonts
22:39:50Onyx4I was thinking that clean should just wipe the ipkg install temp location maybe?
22:40:16kergothhehe
22:40:16TheMasterMind1ipkg_clean
22:40:19kergothwe have like 4 steps
22:40:30kergoth5 actually. source, sourcetree, binary, staging, ipk
22:40:35TheMasterMind1setup a clean for each :)
22:40:38Onyx4cp: cannot stat `/home/clambert/oz/buildroot-oz/build/opie/qt/qconfig*.h': No such file or directory
22:40:46Onyx4I am missing something?
22:40:47kergothOnyx4: wait up
22:40:51kergothOnyx4: i have patches pending
22:40:54Onyx4ah ok
22:40:57TheMasterMind1it should have source extracted that
22:40:57TheMasterMind1hmm
22:41:00kergothOnyx4: committing when opie finishes compiling
22:41:05TheMasterMind1even without kergoth's patches
22:41:30TheMasterMind1btw, when this is all done
22:41:40TheMasterMind1we're releasing oz 3.5 or something
22:41:46kergothhehe
22:41:54jmhodgessweeet new rom
22:42:02jmhodgesno more freezes! :D
22:42:07kergothi need to finish up the patches to run opie as nonroot, the added groups and changed device permissions and shit
22:42:13Onyx4jmhodges: ?
22:42:14TheMasterMind1fine
22:42:16kergothi want that done prior to release, but it wont take more than a few minutes
22:42:17TheMasterMind13.2.5 then
22:42:21TheMasterMind1heh
22:42:25TheMasterMind1ok, when that's done
22:42:26TheMasterMind13.5
22:42:28ljpya make it so
22:42:31TheMasterMind1that way everyone upgrades
22:42:37TheMasterMind1and no one's running shitty stuff any more
22:42:42TheMasterMind1so much has improved
22:42:57ljpdoesnt upgrade, only flashes
22:43:24jmhodgesOnyx4: my Z freezes randomly
22:43:51jmhodgesalso it freezes if i interact with apps before the Today app comes up when i come back from suspend
22:43:59TheMasterMind1jmhodges: what rom
22:44:06TheMasterMind1and is it like freezing every couple of seconds?
22:44:09jmhodgesthe latest one
22:44:11jmhodgesno its not
22:44:13TheMasterMind1heh
22:44:15TheMasterMind1"latest"
22:44:21TheMasterMind1from sf/testing/tjhsst.edu ?
22:44:26jmhodgessf
22:44:31kergoththats old
22:44:31TheMasterMind1theres a lot of latests
22:44:33jmhodgessorry latest stable
22:44:33kergothhehe
22:44:34jmhodges:)
22:44:37kergothoz isnt stable
22:44:39kergoththere is no stable
22:44:41kergoth:)
22:44:46jmhodges"stable-r"
22:44:48jmhodgesjackass :)
22:44:53jmhodgesi know im using a beta
22:44:56jmhodges;)
22:45:08ljphow'd oz get from 2.9.5.5 to 2.9.9?
22:45:19kergothljp: TheMasterMind1 is psycho
22:45:24kergothljp: we're about to release 2.9.7
22:45:27kergothljp: :)
22:45:29ljpheh figgered so
22:45:30jmhodgesheheh
22:45:38TheMasterMind1damn you!
22:45:46TheMasterMind1how can you release 2.9.7 ?!
22:45:48TheMasterMind1after all the fixes!
22:46:03kergothheh
22:46:11kergothdepends on how stable opie is looking atm
22:46:23TheMasterMind1i'll prerelease a warez version of 3.5
22:46:24TheMasterMind1heh
22:46:38kergoth3.5 would be insane
22:46:39jmhodgeslol
22:46:51kergothmuch too far ahead
22:46:58TheMasterMind1ok, how about 3.1
22:46:59TheMasterMind1that's fair
22:47:03kergothnow i could see this being 3.0, if opie is stable enough
22:47:15kergothi'll call the release with 2.4.19 4.0 or something
22:47:20TheMasterMind1yes
22:47:24ljpnot 3.0 until apmd works
22:47:25TheMasterMind1so how about 3.0.5
22:47:33Onyx4I did apmd
22:47:41kergothljp: we're not using apmd persay
22:47:53kergothTheMasterMind1: he's got a point. we shouldnt release 3.0 until the suspend issue is fixed.
22:48:00ljpwell, until suspending isnt causing cruft
22:48:11Onyx4there is a suspend issue?
22:48:15kergothOnyx4: yeah
22:48:20ljpsuspend/resume
22:48:21TheMasterMind1suspend issue will be fixed by the time we have all the mks ported
22:48:25kergothOnyx4: opie uses an ioctl to the apm_bios device during suspend, instead of calling apm -s
22:48:27kergothTheMasterMind1: doubt that.
22:48:33kergothTheMasterMind1: most of the opie devs use ipaqs, not Z's
22:48:36kergothTheMasterMind1: and this is Z specific
22:48:43TheMasterMind1so patch it to use apm -s ?
22:48:50TheMasterMind1its probably the kernel anyway
22:48:52kergothTheMasterMind1: that causes a whole different issue
22:48:54TheMasterMind1will be fixed with .19
22:48:57kergothTheMasterMind1: (using apm -s i mean)
22:48:58TheMasterMind1oh?
22:49:01kergothTheMasterMind1: yeah
22:49:06TheMasterMind1what issue
22:49:07kergothTheMasterMind1: i had verxion try it
22:49:15kergothTheMasterMind1: i dont recall what the behavior was off the top of my head
22:49:30TheMasterMind1we can have it do apm -s if there's a cf card, and ioctl otherwise
22:49:39kergothTheMasterMind1: thatd be pointless
22:49:42ljpI hope your guys idea of 'fixing' opie stuff isn't to just add a patch to your opie tree
22:49:54kergothTheMasterMind1: i'm telling you, its just as broken with apm -s as it is with teh ioctl, only differently
22:49:57kergothljp: well obviously.
22:50:08ljpk.. just making sure ;)
22:50:10TheMasterMind1hides
22:50:13kergothi was going to add a patch to mkipks, till ir emembered i have cvs access
22:50:16kergothhehe
22:50:26jmhodgesoooo mkipks?
22:50:33kergothjmhodges: ?
22:50:37jmhodgesim so out of it sorry
22:50:41kergothjmhodges: just a patch to make it handle being run from outside of $OPIEDIR
22:50:46kergothnp
22:50:48jmhodgesok :)
22:51:12jmhodgesim now a freshman in college and ive lost my footing outside of #lunar and campus life :)
22:51:47Onyx4what's all this openobex for?
22:51:51kergothOnyx4: beaming.
22:51:57Onyx4ah
22:52:30kergothOnyx4: seems like you're getting the hang of the new system. what do you think of it?
22:52:43Onyx4hehe :)  I think it's cool
22:52:48Onyx4it takes me just minutes now to do a package
22:53:02Onyx4if there isn't anything qwerky abou tit
22:53:06TheMasterMind1heh
22:53:18TheMasterMind1remember to update TODO
22:53:24Onyx4I always do check it
22:53:36kergoththe qte-opie deal is a bit quirky, thankfully we're almost done with that
22:53:37Onyx4it's all marked up to openssl which is next
22:53:50TheMasterMind1heh
22:53:51Onyx4so I'll try openssl now
22:54:02TheMasterMind1getting qte/opie working originally (in the old buildroot) was a major pain in the ass
22:54:03kergothhm
22:54:13kergothi should wirte a makefile sanity checker
22:54:24kergothto warn you if the makefile doesnt have the targets it needs, or the variables it needs defined
22:54:30kergoththatd be nifty
22:54:55kergothfuck opie takes a long time to compile
22:55:38ljpheh yup
22:55:42ljpdistcc?
22:55:54kergothi only own one box atm
22:55:56kergothheh
22:55:59ljpoh
22:56:00kergothi shouldve done a -j2 though
22:56:28ljpif I get that cool job, I'll send you one of mine
22:57:16TheMasterMind1what's this distcc?
22:57:25ljp: sorry...
22:57:25ljpibot: distcc
22:57:31ljpfudge
22:57:37ljpgoogle it
22:57:47TheMasterMind1heh
22:58:03TheMasterMind1whoa!
22:58:10TheMasterMind1you mean i can use ALL 40 machines to compile?!
22:58:11ljpthat or teambuilder
22:58:32ljpas long as they have the same version of gcc
22:58:32TheMasterMind1sweeet
22:58:34TheMasterMind1how do i do this
22:58:43TheMasterMind1so i'd need arm-linux-gcc on all machines?
22:59:26ljpyes
22:59:45TheMasterMind1and what else
22:59:49TheMasterMind1apt-get install distcc on all of em?
23:00:11Onyx4mmm why does openssl has this "version' and "subversion" control things? should I keep that?
23:00:17Onyx4or just use the version ?
23:00:59ljpread the distcc docs
23:01:05kergothOnyx4: dunno, just use version
23:02:06Onyx4k
23:12:44TheMasterMind1humps distcc
23:12:47TheMasterMind1i'm installing it on all 41 computers
23:13:09kergothTheMasterMind1: school lab?
23:13:09ljpwonders where that ipv4 to ipv6 programming faq was...
23:13:17TheMasterMind1kergoth: yea
23:13:26TheMasterMind1i do all software install/upgrade stuff
23:13:37ljpheh kernel compile in about 2 minutes
23:13:52Onyx4does ssh usesopenssl?
23:14:01TheMasterMind1its so much easier with debian
23:14:04TheMasterMind1used to all run slack
23:14:04kergothOnyx4: yes
23:14:10TheMasterMind1we converted to debian over the summer
23:14:14Onyx4static or dynamix with openssl ?
23:14:43kergothdynamic
23:14:46kergothkonq also uses it
23:14:50Onyx4k
23:14:51kergothfor https
23:16:27Onyx4I think openssh links with openssl but static
23:16:36Onyx4cause ssh works without the openssl ipk installed
23:16:48kergothwe dont want that
23:16:55kergothcan you figure out why?
23:16:55Onyx4yep
23:17:04kergothcause if there's a bug in openssl i want to be able to update it without recompiling half the distro
23:17:06Onyx4normally even on my mandrake system it's like that check ldd /usr/bin/ssh
23:17:07kergothhehe
23:17:11kergothyeah
23:17:19Onyx4is it like that on debian too?
23:17:51kergothno, its not
23:17:53Onyx4oh wait libcrypto.so.0 is part of openssl duh
23:17:54kergoththats fucking stupid
23:17:57kergother
23:18:00kergothyeah
23:18:08kergothit links dynamically against libcrypto, doesnt link against libssl
23:18:08Onyx4thought that was part of the main system
23:18:13kergothi assume it only uses libcrypto
23:18:16Onyx4yep
23:18:18kergothnope, libcrypt is the base system
23:18:20kergothlibcrypto is not
23:18:22kergoth:)
23:18:26Onyx4ok
23:21:50TheMasterMind1phosphorus:/home/agupta# apt-get upgrade
23:21:50TheMasterMind1Reading Package Lists... Done
23:21:50TheMasterMind1Building Dependency Tree... Done
23:21:50TheMasterMind10 packages upgraded, 0 newly installed, 0 to remove and 0  not upgraded.
23:21:55TheMasterMind1heheh
23:22:03TheMasterMind1i just upgraded like 120 packages
23:22:23TheMasterMind1need to reimage all the machiens tommorow
23:24:37Onyx4kergoth: ok since if you look at http://openzaurus.sourceforge.net/feeds/3.0/
23:24:45Onyx4oops http://www.tjhsst.edu/~agupta/oz2.9.9/
23:24:50Onyx4the openssl is named openssl_main
23:25:00Onyx4I should put a "replaces openssl_main" in the control?
23:25:09Onyx4that package is empty :)
23:25:13Onyx4must have not worked
23:25:19kergothopenssl-main was nothing
23:25:21kergothuseless
23:25:28kergothnever was on people's Z's, so its not worth it
23:25:37Onyx4ok
23:25:37kergothOnyx4: whatd you call the apmd package(s)?
23:25:51Onyx4just like they were
23:25:54Onyx4didn't change it
23:26:15kergothso just a 'apmd' package, with only 'apm'?
23:26:22Onyx4and libapm yes
23:26:29kergothokay, cool
23:26:55Onyx4im trying to make less changes as possible compared to the known good packages to we can at least produce the same result we did before
23:27:01Onyx4then we could start enhancing
23:27:08kergothsounds like a plan
23:27:14TheMasterMind1eh
23:27:16kergothi'm still curious if current busybox explodes
23:27:18TheMasterMind1as long as your eiditing
23:27:22TheMasterMind1might as well fix stuff
23:27:30TheMasterMind1that way we can start expanding
23:27:32TheMasterMind1adding enw packages
23:27:34TheMasterMind1instead of fixing old
23:27:36kergothTheMasterMind1: the more variables we change, the more possible shit we break
23:28:28Onyx4kergoth: oh I know where the problem is now with openssl
23:28:38Onyx4the name is libssl
23:28:44kergoth?
23:28:53Onyx4so we have now openssl and libssl
23:29:05kergothahh
23:29:07Onyx4should I just say that openssl replaces libssl or keep the libssl name
23:29:09kergothit shoudl be libssl0.9.6 iirc
23:29:25kergothwe want to follow what it was before, as i was following debian's package naming
23:30:36Onyx4mmm ok now it's weird because if I change the name on top, it won't match the $(Name) being used to download the files and create the build directory
23:30:43kergothokay here's how you do it
23:30:50kergoth$(stamp-ipk) : Name=libssl0.9.6
23:30:54kergothleave the Name up top as openssl
23:31:02Onyx4ah ok
23:31:02kergothyou can pass a seperate variable to the ipk build process
23:31:10kergothi do that for the multi-ipk builds
23:31:13kergothcan do that here too
23:31:21Onyx4that's what I was gonna look for, like the ncurses
23:31:23kergothhmm
23:31:28kergothmy ipkg wont compile anymore
23:31:32kergothdoes yours?
23:31:46TheMasterMind1hows opie?
23:31:53kergothTheMasterMind1: builds, ipks. doesnt install into rootfs yet
23:31:59TheMasterMind1cool
23:32:02kergothTheMasterMind1: i pushed the changes up
23:32:54kergothmake[5]: Entering directory `/usr/src/coding/projects/user/buildroot/build/ipkg/C/replace'
23:32:56kergothmake[5]: *** No rule to make target `malloc.o', needed by `libreplace.a'.  Stop.
23:33:03TheMasterMind1hmm
23:41:18TheMasterMind1 /zaurus/buildroot/build/qt-2.3.4-snapshot-20020922/bin/uic otimepickerbase.ui -o ./otimepickerbase.h
23:41:19TheMasterMind1grr
23:41:23TheMasterMind1its in 0909
23:41:30TheMasterMind1fix that please
23:42:02kergoth?
23:42:06kergothno, its not in 0909
23:42:17kergothi altered the uic tarball
23:42:19kergothheh
23:42:19TheMasterMind1oh
23:42:20TheMasterMind1you suck
23:45:11Onyx4how do you overwrite a variable that's in the makefile with make?
23:45:36kergothOnyx4: for what?
23:45:36Onyx4like MANDIR is defined hardcoded in openssl
23:45:54kergoth$(MAKE) MANDIR=blah
23:45:54Onyx4instead of using $PREFIX
23:46:01kergothit'll prefer the cmdline passed values
23:46:03kergothto its own
23:46:13Onyx4mmm strange it doesn't seem to work
23:46:43Onyx4oh wait
23:46:45Onyx4syntax
23:47:05Onyx4ok misspelled staging
23:51:08TheMasterMind1heh
23:51:34kergothfuck
23:51:37kergothipkg wont compiel for me for arm
23:51:40kergothipkg-x86 builds fine
23:51:42kergothwtf?
23:51:43jmhodges...
23:51:51jmhodgesouch
23:52:00jmhodgeswhats the error kergoth?
23:52:05kergothits a build thing
23:52:14kergothno rule to build replace.o required by libreplace.a
23:52:23jmhodgesah
23:54:15Onyx4mm libssl-0.9.6 is the name of the package, so the version is hardcoded into it which is libssl-0.9.6-0.9.6 the full name
23:55:08TheMasterMind1heh
23:55:27kergothOnyx4: override Version as well
23:55:30Onyx4so I'm tempted to make things right and call it just libssl and say "replaces libssl-0.9.6" ?
23:55:36kergoth$(stamp-ipk) : Version=$(Revision)
23:55:42kergoth$(stamp-ipk) : Revision=
23:55:45kergothhehe
23:55:48kergothOnyx4: thats an idea
23:55:52kergothOnyx4: Conflict and Provides too
23:56:05kergothOnyx4: you can set Replaces makefile variable and it'll add it to the control
23:56:12kergothOnyx4: same with Conflicts and Replaces
23:56:19Onyx4right I was thinking of doing that
23:56:29Onyx4so it would be what for each
23:56:37Onyx4replaces=libssl-0.9.6
23:56:41Onyx4provides=libssl-0.9.6
23:56:41ljpanyone tried oz with uClibc?
23:56:54Onyx4and no conflicts?
23:57:11kergoth?
23:57:14kergothi just told you
23:57:23kergothReplaces=, Conflicts=, Provides=
23:57:28kergothcapital first letter
23:57:33kergothheh
23:57:36Onyx4oh I put them empty?
23:57:53Onyx4I mean what do I put into each for the ipkg to work right
23:58:07Wemblyhey ker.. new package format.. how debian like it is?
23:58:10Wemblyerr is it sorry
23:58:15kergoth?
23:58:19kergothit *is* the .deb format
23:58:22Wemblyfunfun k
23:58:28kergothyou can use dpkg-deb to extract it
23:58:29kergothheh
23:58:30Wemblywhat about the package files?
23:58:34Wemblylike control and such?
23:58:39kergothwhat about them?
23:58:43Wemblyare they the same format?
23:58:47kergothdood
23:58:53kergothwe changed from tar.gz outer layer to ar
23:58:55kergoththats it
23:59:05Wemblyk
23:59:17Wemblywas control ever diff from debian systems?
23:59:22kergothnope
23:59:25Wemblyk
23:59:26kergoththough some of them werent always supported
23:59:35kergothi.e. Conflicts, Replaces, Provides, etc
23:59:37kergothmake[5]: Entering directory `/usr/src/coding/projects/user/buildroot/build/ipkg/C/replace'
23:59:39kergothmake[5]: *** No rule to make target `malloc.o', needed by `libreplace.a'.  Stop.
23:59:41kergothmake[5]: Leaving directory `/usr/src/coding/projects/user/buildroot/build/ipkg/C/replace'
23:59:43kergothdamnit
23:59:54Wemblyso when are we going to move to a system with source packages? ;)

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with infobot logs, split per channel and by date, etc.