irclog2html for #oe on 20040123

`17:11.33`	`*** join/#oe ibot (ibot@c-24-1-99-18.client.comcast.net)`
`17:11.33`	`*** topic/#oe is OpenEmbedded/OpenZaurus Developer Lounge`
`17:11.46`	`kergoth`	`ibot: wb`
`17:11.48`		`thx`
`17:29.22`	`mickeyl`	`grrr`
`17:29.41`	`mickeyl`	`~lart XMame team for using includefiles named memory.h`
`17:35.54`	`mickeyl`	`mmh that helps`
`17:38.40`	`mickeyl`	`phew. i added so much packages in the last days, i gonna have to spend a weekend trying if they actually work and not just compile :D`
`17:38.46`	`kergoth`	`hehe`
`17:38.57`	`kergoth`	`i didnt test 99% of my packages at runtime till i built an image for ti`
`17:39.05`	`kergoth`	`and it turned out everything worked. was spooky`
`17:39.09`	`mickeyl`	`and how much did you have to tweak`
`17:39.13`	`mickeyl`	`oh really, cool`
`17:39.40`	`kergoth`	`the binaries anyway. some packages were missing a file here and there due to packaging`
`17:39.48`	`kergoth`	`modutils was putting binaries in /usr/sbin, which wasnt critical, just annoying`
`17:39.49`	`kergoth`	`etc`
`17:40.06`	`mickeyl`	`yeah. oh btw., I'm going to make OE the preferred distribution for gamers... added so much sdl game stuff. Need a central emulator GUI now with plugins and stuff.`
`17:40.08`	`CosmicPenguin`	`do all of hte OZ images have Opie in them?`
`17:40.14`	`CosmicPenguin`	`Are there no minimal images?`
`17:40.24`	`kergoth`	`mickeyl: we still have like 8,100 unpackaged files when you do a full oe metadata build`
`17:40.24`	`treke`	`there used to be minimal images`
`17:40.34`	`mickeyl`	`kergoth: oh you can measure that? cool`
`17:40.42`	`mickeyl`	`CosmicPenguin: just depends on your .config.`
`17:40.47`	`kergoth`	`mickeyl: files are -moved- from tmp/base/blah/image to tmp/base/blah/install/[package]`
`17:40.52`	`treke`	`CosmicPenguin: there were never any that are made reliably :)`
`17:40.57`	`kergoth`	`mickeyl: so any files that still exist in tmp/base/*/image werent packaged`
`17:40.57`	`mickeyl`	`kergoth: aah i see. cool`
`17:41.04`	`CosmicPenguin`	`mickeyl: my client messed up, and his Ipaq is missing stuff`
`17:41.07`	`kergoth`	`mickeyl: i've been spending some time working through them`
`17:41.10`	`kergoth`	`mickeyl: one really annoying thing..`
`17:41.19`	`CosmicPenguin`	`And I just needed him to get back to a minimal spot - but oh, well`
`17:41.23`	`kergoth`	`mickeyl: autoconf doesnt seem to have a --docdir by default, so people jsut put their html docs -wherever-`
`17:41.34`	`kergoth`	`mickeyl: /usr/doc, /usr/share/html, /usr/lib/package, blah`
`17:41.38`	`kergoth`	`no consistency`
`17:41.40`	`mickeyl`	`kergoth: ah right. that sucks`
`17:41.44`	`kergoth`	`yeah`
`17:41.51`	`kergoth`	`so i'm thinking, we add a quick autoconf macro`
`17:42.11`	`kergoth`	`to add the cmdline arg, default it to ${datadir}/doc/packagename-version/`
`17:42.20`	`kergoth`	`then use it in every last autoconf build`
`17:42.26`	`kergoth`	`pain in the ass.. but..`
`17:42.33`	`mickeyl`	`better in the long run`
`17:42.40`	`kergoth`	`yep`
`17:42.46`	`kergoth`	`and upstream maintainers would probably accept that`
`17:43.24`	`mickeyl`	`yeah, sure`
`17:43.28`	`mickeyl`	`bbl`
`17:49.23`	`kergoth`	`heh`
`17:49.24`	`kergoth`	for i in tmp/stamps/do_unpack; do compile=`echo $i\|sed -e's,do_unpack,do_compile,g'`; if ! test -e $compile; then echo $compile does not exist; fi; done\|sed -e 's,^./,,g'
`17:49.35`	`kergoth`	`useful to check what builds failed`
`18:00.26`	`CIA-5`	`03kergoth 07 * r1.597 10packages/aalib/aalib-1.4rc5/configure.patch: BUGFIX: Correct syntax issues with the m4 macros. automake/aclocal 1.8 is more picky.`
`18:00.26`	`CIA-5`	`03kergoth 07 * r1.598 10packages/ (5 files in 4 dirs): Dont install some m4s..`
`18:00.49`	`Twiun`	`kergoth: check it in as a utility script :)`
`18:01.17`	`kergoth`	`nah, going to alter the parsers to set a STATUS variable so you can easily see the status of any given .oe`
`18:01.32`	`kergoth`	`then itll be easier to directly query it`
`18:01.35`	`kergoth`	`:)`
`18:22.45`	`Pendalar`	`hi hi`
`18:27.28`	`kergoth`	`hey`
`18:27.48`	`Pendalar`	`busy as usual?`
`18:27.59`	`kergoth`	`yep`
`18:28.04`	`kergoth`	`week from hell`
`18:28.08`	`kergoth`	`thank god its nearly over`
`18:28.51`	`Pendalar`	`whenever I say that, I get pulled into working weekends`
`18:29.00`	`Pendalar`	`so.. I just keep my mouth shut til I get home`
`18:29.03`	`Pendalar`	`and unplug the phone`
`18:29.06`	`kergoth`	`hehe`
`18:29.31`	`Pendalar`	`I think I'm about to the end of my current build run, it just hit grub and now starting back`
`18:29.57`	`Pendalar`	`I wish I was intelligent enough to write a script to catch failures, add them to a grep var, and start over with that grepped out`
`18:30.09`	`Pendalar`	`then at the end dump a list of failed packages and in what stage`
`18:30.21`	`Pendalar`	`well I guess stage isn't necessary all that much, but still`
`18:30.41`	`Pendalar`	`the interactiveness makes a build take 3 days since I can't babysit it`
`18:30.55`	`Pendalar`	`I'm pretty sure even my 600mhz could do it while I sleep per night`
`18:32.58`	`kergoth`	`Pendalar:`
`18:33.05`	`kergoth`	`oemake (without -a, wont abort on failure)`
`18:33.11`	`kergoth`	then for i in tmp/stamps/*do_unpack; do compile=`echo $i\|sed -e's,do_unpack,do_compile,g'`; if ! test -e $compile;
`18:33.15`	`kergoth`	`<PROTECTED>`
`18:33.25`	`kergoth`	`which will show packages that were unpacked, but not compiled (failed)`
`18:33.50`	`gb2`	`kergoth: should make -a the default`
`18:33.57`	`kergoth`	`gb2: agreed`
`18:34.03`	`Pendalar`	`ok`
`18:34.06`	`gb2`	`and an option to continue on error`
`18:34.13`	`kergoth`	`should change to -k, same as make's continue on error`
`18:34.16`	`Pendalar`	`well I don't know about default..`
`18:34.17`	`gb2`	`yeah`
`18:34.22`	`Pendalar`	`because then normal users won't see it`
`18:34.24`	`gb2`	`and perhaps a -n to tell what it's going to do`
`18:34.27`	`Pendalar`	`and assume everything went fine`
`18:34.32`	`Pendalar`	`never think to look in the log`
`18:34.38`	`gb2`	`pendalar: die on error should be the default`
`18:34.43`	`gb2`	`it isn't currently`
`18:34.43`	`Pendalar`	`like make does, it doesn't default to continue on`
`18:35.04`	`kergoth`	`Pendalar: uh, the default right now is to not abort if something fails`
`18:35.09`	`kergoth`	`so yes, you're arguing our point`
`18:35.13`	`Pendalar`	`oh`
`18:35.25`	`gb2`	`no wait, i know.`
`18:35.31`	`gb2`	`we should die if something goes wrong`
`18:35.35`	`gb2`	`so the user doesn't get confused!`
`18:35.45`	`Pendalar`	`sorry I've been using my little grep script thing and wasn't sure what the default was, read gb2's line wrong`
`18:36.28`	`Pendalar`	`I keep thinking about all this water on mars crap, ancient beds and what not..`
`18:36.36`	`Pendalar`	`in the end, I expect some smart ass to say "yeah but the martians got thirsty, drank it all and died, you've wasted 30 yrs of research"`
`18:37.50`	`gb2`	`pendlar: add " now pardon me, i'm going to go get another bottle of water to drink"`
`18:38.16`	`mike_wrk`	`gb2: also should have a dependency list so packages that rely on the failed package don't even try to build`
`18:38.41`	`kergoth`	`mike_wrk: it does that now.`
`18:39.07`	`kergoth`	`anyone know the escape code to erase the current line?`
`18:40.32`	`*** join/#oe robin_ (~robin@debian.demon.nl)`
`18:40.46`	`Pendalar`	`heh`
`18:41.42`	`Pendalar`	`as in ^M sort of junk?`
`18:42.17`	`Pendalar`	`and I assume you mean for bash`
`18:42.57`	`Pendalar`	`esc d does after cursor`
`18:44.00`	`gb2`	`kergoth: \r \r`
`18:44.11`	`Pendalar`	`esc r yeah`
`18:44.16`	`Pendalar`	`or however you mean to use it`
`18:44.31`	`gb2`	`uh. that was a carriage return, buncha spaces, carriage return`
`18:44.32`	`gb2`	`;)`
`18:44.53`	`kergoth`	`looks like \033[2K is erase the current line`
`18:45.07`	`kergoth`	`for vt100 anyway`
`18:46.02`	`treke`	`isnt this what curses is for? :p`
`18:49.19`	`kergoth`	`heh, yeah, but thats overkill`
`18:49.23`	`kergoth`	`just want a damn spinner`
`18:49.27`	`kergoth`	`ah, here we go`
`18:49.37`	`Pendalar`	`what is it?`
`18:50.16`	`kergoth`	`didnt need th eerase line code anyway`
`18:50.57`	`Pendalar`	`then how? I've wanted to know this`
`18:50.59`	`Pendalar`	`;-)`
`18:51.03`	`kergoth`	`i just told you`
`18:51.09`	`kergoth`	`\033[2K erases the current line`
`18:51.16`	`kergoth`	`where \033 is octal for escape`
`18:51.31`	`Pendalar`	`I thought you meant that's what it was for vt100 but you needed it for something else`
`18:51.46`	`Pendalar`	`maybe I need a nap =-x`
`18:57.42`	`Pendalar`	`I assume I need to throw some sleep/waits into a spinner bar script`
`18:57.50`	`Pendalar`	`cause I can't see anything ;-)`
`18:58.45`	`Pendalar`	`although sleep 1 is a bit slow`
`18:59.35`	`kergoth`	`i was using python, so i used time.sleep(.2), which worked pretty well`
`18:59.46`	`kergoth`	`when i was playing with it in testing`
`19:00.14`	`Pendalar`	`hmm it keeps moving over one space`
`19:00.43`	`kergoth`	`then your erase isnt happening`
`19:01.07`	`Pendalar`	`echo "\033[2Ka" .. sleep 1 .. "echo\033[2kb"`
`19:01.22`	`Pendalar`	`it is erasing the a before it does b, but b is one space to the right of the position of a`
`19:01.58`	`Pendalar`	`echo -en ?`
`19:01.58`	`kergoth`	`yopu're using echo, not echo -n`
`19:02.01`	`kergoth`	`so it goes to the next line. that escape code erases the current line, not the previous one`
`19:02.11`	`kergoth`	`yes. echo -n doesnt print a newline`
`19:02.11`	`Pendalar`	`yeah sorry I've got -en in my script`
`19:02.25`	`Pendalar`	`-e to do the parsing`
`19:02.40`	`Pendalar`	`well I guess that's not needed in this case`
`19:02.47`	`kergoth`	`ah, k`
`19:02.49`	`kergoth`	`dunno then`
`19:02.57`	`Pendalar`	`er ya got to have that`
`19:03.10`	`CIA-5`	`03kergoth 07 * r1.375 10oe/bin/oemake: Add a text spinner to oemake during its .oe parsing, so the user gets some feedback.`
`19:04.07`	`Pendalar`	`prettifying..`
`19:06.00`	`Pendalar`	`maybe sleep adds the space`
`19:06.03`	`Pendalar`	`for some gay reason`
`19:06.10`	`Pendalar`	`although shouldn't..`
`19:07.38`	`Pendalar`	`oh well I think it erases the content of the line, but not rewinds it`
`19:08.15`	`mike_wrk`	`Hmmm, the build is spending too much time in python, gotta find a way to speed that up`
`19:08.25`	`kergoth`	`Pendalar: try "\b " before the symbol to spin`
`19:08.33`	`kergoth`	`mike_wrk: it has to parse in every single .oe file into ram`
`19:08.40`	`kergoth`	`mike_wrk: create a dependency digraph from that information`
`19:08.40`	`Pendalar`	`I get a line like this \|\-/\|\-`
`19:08.42`	`kergoth`	`mike_wrk: then go`
`19:08.50`	`kergoth`	`mike_wrk: the -vast- majority of time is spent reading in the .oe files`
`19:08.54`	`Pendalar`	`if I take the clears out, or the same thing but previous chars blanked with it in`
`19:09.01`	`Pendalar`	`ok`
`19:09.06`	`mike_wrk`	`Hmmm, we may need to up the machine from 2G of ram then`
`19:09.29`	`Pendalar`	`heh then what's the point of the clear?`
`19:09.30`	`Pendalar`	`;-)`
`19:09.33`	`kergoth`	`if 2g of ram isnt enough, we've got problems`
`19:09.36`	`kergoth`	`Pendalar: what clear?`
`19:09.50`	`mike_wrk`	`well, both you and I are doing builds simultaneously`
`19:10.05`	`Pendalar`	`<PROTECTED>`
`19:10.05`	`kergoth`	`hehe, i'm not building, just testing oemake functionality`
`19:10.05`	`mike_wrk`	`And the drive on that machine is running as ATA100 so it shouldn't be slow`
`19:10.11`	`kergoth`	`Pendalar: you dont need that if you jsut use \b to erase the previous char`
`19:10.42`	`Pendalar`	`=-}`
`19:11.56`	`kergoth`	`hmm, oemake syntax change is about to break everyone using scripts that call oemake -a`
`19:11.59`	`kergoth`	`hehe`
`19:12.03`	`kergoth`	`maybe i shouldve kept -a as a no-op`
`19:17.21`	`Pendalar`	`would there be an overall way to add an overall process indicator of the build?`
`19:17.22`	`kergoth`	`there, thats better`
`19:17.29`	`kergoth`	`not as things stand now`
`19:17.30`	`Pendalar`	`or at least "Now on [1/1000]"`
`19:17.39`	`kergoth`	`that we could do`
`19:17.41`	`Pendalar`	`doesn't have to be a real time indicator, just x/packages`
`19:17.44`	`kergoth`	`but we'd need to silence oemake even more`
`19:17.50`	`kergoth`	`since it shows 'starting task ..' etc now`
`19:18.04`	`kergoth`	`course we could use ncurses in the oemake output`
`19:18.05`	`Pendalar`	`split screen ;-)`
`19:18.11`	`kergoth`	`show the status in one window, log in another`
`19:18.22`	`Pendalar`	`line 1 at top does [x/total], rest normal`
`19:18.34`	`CIA-5`	`03kergoth 07 * r1.376 10oe/bin/oemake: oemake syntax change. abort on failure is now default, with -k commandline option to continue as much as possible (same cmdline opt as gmake).`
`19:18.34`	`CIA-5`	`03kergoth 07 * r1.377 10oe/bin/oemake: Add --abort/-a back to oemake, as a no-op with a warning.`
`19:18.40`	`kergoth`	`yeah, ncurses would be easiest for that, and probably overkill atm`
`19:18.45`	`kergoth`	`will look into something like that at some point`
`19:19.07`	`Pendalar`	`I was thinking something close to the way the configuring packages to root thing is`
`19:19.16`	`Pendalar`	`after flashing`
`19:19.22`	`kergoth`	`hmm, i should catch KeyboardInterrupt's in the build, and abort cleanly`
`19:19.34`	`kergoth`	`otherwise ctrl+c'ing teh build looks ugly as hell with the traceback`
`19:20.07`	`Pendalar`	`I rm -rf tmp/* so much I hardly spent time in that area`
`19:20.16`	`kergoth`	`mike_wrk: heh, damn, those python processes sure do use alot of cpu`
`19:20.26`	`mike_wrk`	`yep`
`19:20.43`	`kergoth`	`hmm`
`19:20.55`	`kergoth`	`well, itd help to cache the data in a file, at least thatd cut down on the # of times we need to do it`
`19:21.25`	`mike_wrk`	`yeah, but we don't want to really kill machines with a lot less memory`
`19:21.32`	`gb2`	`is it really slow at just parsing?`
`19:21.38`	`kergoth`	`mike_wrk: well, theres one option`
`19:21.46`	`mike_wrk`	`gb2: python is taking up 100% of one cpu a lot of the time`
`19:21.52`	`gb2`	`profile it?`
`19:21.58`	`mike_wrk`	`In comparise gcc takes no time`
`19:22.00`	`kergoth`	`mike_wrk: we could parse in a single package's metadata, then throw out most of it minus a few items we require, then have it call out oebuild rather than starting the build itself`
`19:22.23`	`Pendalar`	`oh no he's wanting to start over ^_^`
`19:22.27`	`mike_wrk`	`he,he, oemake just took less than 45 minutes`
`19:22.28`	`kergoth`	`err no, thatd be worse`
`19:22.29`	`kergoth`	`heh`
`19:22.39`	`Pendalar`	`we'll call it NextEmbedded`
`19:22.47`	`Pendalar`	`everyone go to #ne ... =-x`
`19:22.57`	`kergoth`	`mike_wrk: oemake actual builds are a fair bit faster due to my removal of the 'parsing the .oe file twice' thing`
`19:23.03`	`kergoth`	`mike_wrk: parsing is still slow of course`
`19:23.14`	`gb2`	`why is parsing so slow?`
`19:23.19`	`gb2`	`does python just suck?`
`19:23.28`	`mike_wrk`	`python does suck`
`19:23.41`	`mike_wrk`	`precompiling python code helps but there's a lot of overhead in there`
`19:23.46`	`gb2`	`regexps are probably being really slow`
`19:23.58`	`gb2`	`is it recompiling regexps each time or something?`
`19:24.32`	`kergoth`	`gb2: nope, it compiles them when the parse .py is 'import'd`
`19:25.10`	`gb2`	`i wonder if there is a python profiler`
`19:27.04`	`mike_wrk`	`gb2: http://ic2.epfl.ch/doc/python1.4/lib/node93.html`
`19:28.55`	`gb2`	`might try that on oemake/build`
`19:29.13`	`gb2`	`do an oebuild showdata on something`
`19:29.22`	`kergoth`	`yeah, thatd be interesting`
`19:30.00`	`gb2`	`could be some stupid O(n^4) algorithm someone forgot about or something`
`19:30.11`	`kergoth`	`hehe`
`19:30.53`	`mike_wrk`	`gb2: I also wonder if its trying to do some clever shit thinking its on ix86`
`19:31.17`	`gb2`	`clever x86 stuff should still be decent on the opteron`
`19:31.23`	`gb2`	`oe* is slow on my P4`
`19:39.39`	`mike_wrk`	`gb2: my python is compiled native for x86_64`
`19:43.16`	`mike_wrk`	`I'll take a look over the weekend, I wanna try to play with the make invocation too so we can do -j2 without killing everything`
`19:43.49`	`kergoth`	`imho we'd be better off with a tool that can spawn off multiple package builds at once`
`19:44.00`	`kergoth`	`since sooo many packages cant handle make -j, at least we could build a couple of them at once`
`19:44.12`	`mike_wrk`	`I just thought that too`
`19:44.31`	`mike_wrk`	`have to be careful of inter-dependent packages though`
`19:44.49`	`kergoth`	`yep`
`19:45.36`	`mike_wrk`	`he, he, could make oemake understand -j2 itself ( or -j N)`
`19:46.03`	`kergoth`	`heh, yeah, nothing wrong with using a syntax people are comfortable with`
`19:47.53`	`Pendalar`	`hrm`
`19:48.09`	`Pendalar`	`why do I have a /sgml-common dir now`
`19:48.18`	`kergoth`	`why wouldnt we/`
`19:48.22`	`Pendalar`	`in / ?`
`19:48.32`	`Pendalar`	`I don't mean packages/sgml-common`
`19:48.33`	`kergoth`	`oh, dunno`
`19:48.36`	`Pendalar`	`I mean literally`
`19:48.48`	`Pendalar`	`did the build do that?`
`19:48.53`	`kergoth`	`dunno`
`19:48.56`	`Pendalar`	`wasn't there yesterday morning..`
`19:49.15`	`Pendalar`	`yep`
`19:49.16`	`Pendalar`	`0.6.3`
`19:49.18`	`Pendalar`	`=-x`
`19:49.26`	`Pendalar`	`the build installed it to the wrong area or something`
`19:49.50`	`Pendalar`	`I guess anyways`
`19:49.52`	`Pendalar`	`[root@NeXuS root]# ls /sgml-common-0.6.3/html`
`19:49.52`	`Pendalar`	`fdl.html index.html install-catalog.html introduction.html`
`19:50.10`	`treke`	`hmmm`
`19:50.29`	`Pendalar`	`timestamp was the 20th, guess from the last time I toyed with it`
`19:50.31`	`treke`	`looks liek the php version of our conflict checking is too slow. might have to learn how to do php extentions`
`19:50.32`	`Pendalar`	`or something`
`19:51.27`	`mike_wrk`	`treke: they're pretty easy, I've build php modules before`
`19:51.40`	`treke`	`is it? that's good to hear`
`19:51.59`	`mike_wrk`	`yep, a few macros to bind the C functions and figure out param passing`
`19:52.38`	`Pendalar`	`should I leave it kergoth or remove and verify if possible`
`19:53.08`	`kergoth`	`Pendalar: remove and verify if possible`
`19:53.40`	`Pendalar`	`ok this build should be done in a day or two =-x`
`19:56.49`	`*** join/#oe Toi (~pleemans@D577A871.kabel.telenet.be)`
`19:58.52`	`Pendalar`	`hmm my box is pausing more and more between packages`
`19:59.04`	`Pendalar`	`wonder if I'm running out of (insert resource name here)`
`19:59.16`	`mike_wrk`	`what does top reckon ?`
`20:00.03`	`Pendalar`	`top reckon's python sucks`
`20:00.52`	`Pendalar`	`;-) got plenty of swap, and 400mb left for ccache`
`20:04.53`	`Pendalar`	`yeah kergoth`
`20:05.02`	`Pendalar`	`sgml-common-native sticks crap in my /`
`20:05.09`	`Pendalar`	`=-x`
`20:07.25`	`mike_wrk`	`yeah, we gotta look at python`
`20:07.29`	`mike_wrk`	`right, mtg time`
`20:07.31`	`mike_wrk`	`bbl`
`20:17.49`	`kergoth`	`mike_wrk: hmm, at a minimum we can probably have it 'del' the metadata for each package once said package's build completes. it doesnt do that now, so doing htat could at least drop the mem usage over time`
`20:46.03`	`Pendalar`	`I take it you guys have other things on your minds =-x`
`21:00.57`	kergoth`food	`hmm, that helped`
`21:05.09`	`Pendalar`	`the food or the verifying?`
`21:05.29`	kergoth`food	`the food and the oemake deleting metadata as it goes`
`21:05.37`	kergoth`food	`oh, and the latte`
`21:05.39`	kergoth`food	`:)`
`21:08.30`	`CIA-5`	`03kergoth 07 * r1.378 10oe/bin/oemake: Make oemake 'del' a package's metadata when said package's build is complete.`
`21:09.05`	`gb2`	`mmm latte`
`21:18.31`	`Pendalar`	`=-/ I was refering to sgml-common-native`
`21:22.33`	`kergoth`	`heh`
`21:22.55`	`kergoth`	`hmm`
`21:23.04`	`kergoth`	`wonder how much more we can cut down the resource usage of oemake`
`21:28.52`	`gb2`	`heh`
`21:32.10`	`*** join/#oe TheMasterMind1 (~aman@h-68-166-68-175.MCLNVA23.dynamic.covad.net)`
`21:32.14`	`TheMasterMind1`	`sup`
`21:32.27`	`TheMasterMind1`	`whats new with oe lately`
`21:32.43`	`kergoth`	`TheMasterMind1: http://openembedded.bkbits.net:8080/oe/ChangeSet@-7d?nav=index.html`
`21:32.54`	`kergoth`	`TheMasterMind1: http://openembedded.bkbits.net:8080/packages/ChangeSet@-7d?nav=index.html`
`21:32.57`	`kergoth`	`:P`
`21:33.08`	`TheMasterMind1`	`heh`
`21:33.31`	`kergoth`	`gb2: heh, every single package build ends up recursing like 7 levels deep, at a minimum`
`21:33.35`	`kergoth`	`hmm`
`21:35.51`	`*** join/#oe DavidM (~david@h24-207-7-221.dlt.dccnet.com)`
`21:37.07`	`kergoth`	`TheMasterMind1: started work on making the c7x0/5600 kernel maintainable`
`21:37.08`	`kergoth`	`fyi`
`21:37.12`	`TheMasterMind1`	`ooh`
`21:37.12`	`TheMasterMind1`	`nice`
`21:37.24`	`kergoth`	`so far have the core machine bits and started on a input layer keyboard driver`
`21:37.31`	`kergoth`	`then ill do touchscreen, cause its a breeze`
`21:37.33`	`kergoth`	`and go from there`
`21:45.53`	`DavidM`	`Is this the channel for questions? or #openembedded?`
`21:46.05`	`kergoth`	`this is the replacement for #openembedded`
`21:46.22`	`kergoth`	`you can ask qeustions here pertaining to the OE tools. not to openzaurus or other distributions built by oe and/or buildroot`
`21:46.48`	`DavidM`	`replacement? shorter name for irc?`
`21:46.48`	`kergoth`	`gb2: this is depressing`
`21:47.06`	`kergoth`	`DavidM: people joined #openembedded thinking it was a general linux embedded discussion channel, not knowing that it was a specific project`
`21:47.12`	`kergoth`	`DavidM: got sick of redirecting people to #elinux`
`21:47.18`	`kergoth`	`plus yeah, its shorter, and we're lazy`
`21:47.18`	`DavidM`	`ahh`
`21:47.52`	`DavidM`	`ok, I just grabbed oe from bitkeeper. Is there any known current issues or am I just doing something really wrong?`
`21:48.13`	`kergoth`	`there are issues, but htye're relatively minor. its definately functional`
`21:48.15`	`kergoth`	`what problem are you having?`
`21:48.36`	`DavidM`	`I grabbed it, and then setup my local.conf file from the instruction son the tiki`
`21:48.45`	`kergoth`	`'it'`
`21:48.48`	`DavidM`	`then tried to run oebuild patcher/patcher-native.oe`
`21:48.48`	`kergoth`	`both repositories?`
`21:48.54`	`kergoth`	`you need both packages and oe`
`21:48.57`	`DavidM`	`both oe and the packages`
`21:49.04`	`kergoth`	`okay, what happens then?`
`21:49.08`	`DavidM`	`running from packages dir`
`21:49.18`	`DavidM`	`should I go to to #flood`
`21:49.29`	`kergoth`	`yeah`
`21:51.01`	`DavidM`	`kergoth: ok, what version do I need?`
`21:52.29`	`DavidM`	`kergoth: I'm just using the one from Debian Woody, python 2.1.3-3.2`
`21:52.29`	`kergoth`	`DavidM: 2.3 preferably, but 2.2 would probably work`
`21:52.35`	`kergoth`	`2.1 will definately not work`
`21:52.37`	`DavidM`	`ok thanks`
`21:52.39`	`kergoth`	`np`
`21:52.53`	`kergoth`	`mike_wrk: profile.py is handy.`
`21:53.38`	`kergoth`	`in one .oe parse, looks like 1745 calls to match()`
`21:53.40`	`Pendalar`	OEFILES=`ls /.oe\|grep -E -v 'binutils\|uclibc\|gcc\|avr-libc\|iputils\|popt\|e2fsprogs-libs\|libqpe-opie\|aumix\|e2fsprogs\|linux\|libelf\|pyvorbis\|grub\|fetchmail\|glibc-initial'` oemake -a;
`21:53.42`	`kergoth`	`something is wrong with this picture`
`21:53.46`	`Pendalar`	`for full build this round`
`21:53.57`	`kergoth`	`Pendalar: yeah, thats pretty close to what i'm getting too`
`21:54.02`	`kergoth`	`Pendalar: slowly working through them`
`21:56.41`	`Pendalar`	`on a non RH box?`
`21:56.54`	`kergoth`	`Pendalar: recent suse, so similar failures`
`21:57.11`	`Pendalar`	`hm`
`21:58.50`	`mike_wrk`	`kergoth: damn, I'll take a good look at the profile and match stuff`
`21:58.58`	`Pendalar`	`would installing sgml-common to / be part of the problem of the doc stuff going so rough?`
`22:02.54`	`kergoth`	`mike_wrk: python /usr/lib64/python2.3/profile.py $OEDIR/bin/oebuild showdata blah.oe;`
`22:03.51`	`kergoth`	`<PROTECTED>`
`22:03.52`	`kergoth`	`<PROTECTED>`
`22:03.52`	`kergoth`	`<PROTECTED>`
`22:03.52`	`kergoth`	`heh`
`22:04.07`	`kergoth`	`2281 calls to _compile. can we say "python's regex sucks"?`
`22:04.12`	`kergoth`	`anyway`
`22:04.23`	`kergoth`	`doesnt spend much time there, so who cares`
`22:04.31`	`kergoth`	`brb, soda run`
`22:05.23`	`Pendalar`	`=-p`
`22:07.47`	`mike_wrk`	`<PROTECTED>`
`22:08.01`	`mike_wrk`	`<PROTECTED>`
`22:08.38`	`mike_wrk`	`2879/2826 0.020 0.000 0.030 0.000 data.py:62(getVar)`
`22:08.50`	`Pendalar`	`that doesn't work at all on my box`
`22:08.57`	`Pendalar`	`but then again I have 2.2`
`22:09.00`	`mike_wrk`	`well, you don't have lib64 either`
`22:09.08`	`Pendalar`	`I'm smart enough to change that on my own`
`22:09.09`	`Pendalar`	`=-P`
`22:09.18`	`kergoth`	`heh, i ran oemake and had it quit after parsing`
`22:09.21`	`kergoth`	`and profile dit`
`22:09.23`	`kergoth`	`<PROTECTED>`
`22:09.25`	`Pendalar`	`Traceback (most recent call last):`
`22:09.25`	`Pendalar`	`<PROTECTED>`
`22:09.25`	`Pendalar`	`<PROTECTED>`
`22:09.29`	`Pendalar`	`stuff like that happens`
`22:09.47`	`mike_wrk`	`pendalar: you need the $OEDIR/bin/oebuild`
`22:09.56`	`mike_wrk`	`it won't find it from the path`
`22:10.25`	`kergoth`	`417621/321 4.470 0.000 8.900 0.028 copy.py:161(deepcopy`
`22:10.33`	`kergoth`	`heh, copying metadata around is consuming alot`
`22:10.43`	`ljp`	`rewrite it in perl :)`
`22:10.56`	`kergoth`	`(each package uses the same base config metadata from oe.conf, so we deepcopy that to create the initial data for each package)`
`22:11.23`	`Pendalar`	`yeah.. I know..`
`22:11.30`	`kergoth`	`hmmm`
`22:11.31`	`gb2`	`kergoth: depressing?`
`22:11.34`	`kergoth`	`gb2: quite`
`22:11.38`	`gb2`	`heh`
`22:11.42`	`gb2`	`efficiency issues?`
`22:11.44`	`kergoth`	`gb2: talk about motivation for our new backend..`
`22:11.48`	`gb2`	`python sucking mroe than expected?`
`22:11.49`	`kergoth`	`seems so`
`22:11.50`	`Pendalar`	`python /usr/lib/python2.2/profile.py /root/devel/openembedded/oe/bin/oebuild showdata even`
`22:12.25`	`mike_wrk`	`<PROTECTED>`
`22:12.46`	`Pendalar`	`you can't showdata overall?`