T2 IRC Log: 2010-07-22
This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.
--- Log opened Thu Jul 22 00:00:14 2010
00:08 -!- Asmundr [~jamie@h113.143.102.166.static.ip.windstream.net] has joined #t2
01:22 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has quit [Read error: Connection reset by peer]
08:44 -!- mjungwirth [~mjungwirt@213.174.234.68] has quit [Quit: Nettalk6 - www.ntalk.de]
08:46 -!- mjungwirth [~mjungwirt@213.174.234.68] has joined #t2
09:33 < CIA-29> aldas * r36877 /trunk/package/php/php/php.desc: * updated php (5.3.2 -> 5.3.3)
09:33 < CIA-29> aldas * r36878 /trunk/package/develop/git/git.desc: * updated git (1.7.1.1 -> 1.7.2)
10:28 < CIA-29> rene * r36879 /trunk/package/develop/gambas/gambas.desc: * updated gambas (1.0.19 -> 2.18.0)
10:32 < CIA-29> rene * r36881 /trunk/package/contrib/beep/beep.cache: * added beep.cache
10:41 < CIA-29> rene * r36880 /trunk/package/contrib/beep/beep.desc: * updated beep (0.0.1 -> 1.3)
11:06 < CIA-29> rene * r36882 /trunk/package/network/tcpdump/ (tcpdump.conf tcpdump.desc):
11:06 < CIA-29> Mario Jungwirth:
11:06 < CIA-29> * improved tcpdump to cross compile
11:10 < CIA-29> rene * r36883 /trunk/package/base/htop/ (configure.patch.cross htop.conf htop.desc):
11:10 < CIA-29> Mario Jungwirth:
11:10 < CIA-29> * improved htop to cross compile
11:13 < CIA-29> rene * r36884 /trunk/package/x86/syslinux/syslinux.desc: * updated syslinux (4.01 -> 4.02)
11:13 < CIA-29> rene * r36885 /trunk/package/database/sqlite/sqlite.desc: * updated sqlite (3.6.23.1 -> 3.7.0)
11:33 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has joined #t2
11:33 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has quit [Changing host]
11:33 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has joined #t2
11:38 -!- arete74 [~arete@143.225.107.252] has joined #t2
12:04 < CIA-29> rene * r36886 /trunk/package/editors/texmacs/texmacs.desc: * updated texmacs (1.0.7.3 -> 1.0.7.5)
12:12 -!- uid9909 [~uid9099@2001:470:9a41:107:4cda:5044:257c:e722] has joined #t2
12:24 [Users #t2]
12:24 [@ChanServ] [ CIA-29 ] [ koan ] [ rxr ] [ TobiX ]
12:24 [ arete74 ] [ digitaloktay] [ mjungwirth] [ Stealth] [ uid9909]
12:24 [ Asmundr ] [ dileks ] [ mtr ] [ T-One ] [ zod ]
12:24 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
12:44 * rxr to lunch - cu
12:54 -!- arete74 [~arete@143.225.107.252] has quit [Quit: leaving]
13:37 -!- rogermason [~user@CPE001b639d4751-CM0011aea11450.cpe.net.cable.rogers.com] has joined #t2
13:39 < rogermason> I'm working through Emerging the dependencies to get emacs installed. glitz fails for lack of GL/gl.h. Is that provided by mesa? Thanks.
14:07 -!- uid9909 [~uid9099@2001:470:9a41:107:4cda:5044:257c:e722] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.7/20100713130626]]
15:06 < koan> curious, the Wikipedia notability nazis have deleted the page about T2 SDE: https://secure.wikimedia.org/wikipedia/en/wiki/T2_SDE
15:10 < rxr> oh my
15:11 < rxr> sucks
15:14 < koan> the german wikipedia moderators are more sane
15:39 < koan> rxr: the build intro page still calls 6.0 the "latest" release: http://t2-project.org/documentation/buildintro.html
15:40 < koan> (there may be other outdated pages too)
15:42 < rxr> updated
15:45 < koan> maybe grep for "6.0" and "7.0" through the website's sources :-)
15:45 < rxr> tried, pretty much noise
15:51 < rxr> if you have questinos jsut ask :-)
15:53 < koan> yup, I'm just skimming through the website now to build the skeleton of the article
15:55 < koan> hm I hadn't seen the whitepaper before
16:00 < rogermason> Got past the GL issue. Now atk won't build: CDPATH="${ZSH_VERSION+.}:" && cd . && /bin/sh /root/t2-8.0/src.atk.SDE-CONFIG.20100722.112349.1679.gem/atk-1.30.0/missing --run aclocal-1.11 $
16:00 < rogermason> aclocal-1.11: couldn't open directory `//share/aclocal': No such file or directory
16:00 < rogermason> make: *** [aclocal.m4] Error 1
16:00 < rogermason> Due to previous errors, no 9-atk.log file!
16:00 < rogermason>
16:01 < rxr> which GL issue ?
16:04 < rogermason> I'm working through Emerging the dependencies to get emacs
16:04 < rogermason> installed. glitz fails for lack of GL/gl.h. Is that provided by
16:04 < rogermason> mesa? Thanks.
16:05 < rogermason> But I built mesa and that fixed the "issue".
16:37 -!- mjungwirth2 [~mjungwirt@213.174.252.112] has joined #t2
16:54 < koan> rxr: is the long-term goal of support for other open source kernels still being pursued?
16:55 < rxr> yes, but I'm not yet sure which one to use
16:55 < rxr> i'm not fully happy with either of them
16:55 < rxr> minix is too lagging
16:55 < rxr> draonfly the better one so far
16:55 < koan> I remember you looked at DragonFly BSD yes
16:55 < rxr> but I started with mingw / win32 support :-)
16:56 < koan> and how difficult is it to port T2 to another kernel? which pieces have to be changed?
16:56 < rxr> I'm actually thinking about re-using L4 bits and writing my own multi-server kernel as a hobby
16:57 < rxr> well, beside patching some of the packages (pkgsrc like) to build with the other kernel at all, it is not so difficult
16:57 < rxr> mostly only the kernel sources need to be extracted from their archives (most are not as nicely tarball packaged as linux, but just resist in their SVN, etc. tree)
16:57 < rxr> and with the kernel the c library and maybe some system control tools
16:57 < rxr> it's actually not sooo much work
16:57 < rxr> even beginners could do so
16:58 < rxr> it's mostly that I personally need something more micr-kernel (read stable) than Linux, and I have not yet pinpointed the best solution, yet
16:58 < koan> hm ok, and I guess they can look at how the mingw/win32 port was done for a start?
16:59 < rxr> yes, though mingw does not come with a kernel, ... but the basics are the same
16:59 < rxr> especially wheter the gcc release tarball can produce working executables for said platform it's an easy game, just package the system header and libc and you're already halfway there
17:00 < rxr> e.g. most normal packages should already build, ... and then just collecting all their external kernel sources into a nice kernel packages, ...
17:01 < koan> well, this idea of one userland (T2) with different choices of kernel has always intrigued me, so I really hope I still see it happening once :-)
17:01 < rxr> well, for most part t2 is like those pkgsrc trees, ...
17:01 < rxr> ports collection
17:02 < rxr> btw. the Haiku folks even considered using T2^W ROCK Linux back in their early days when they still named itself openbsd
17:02 < koan> interesting
17:03 < koan> why did they decide to not do it?
17:06 < digitaloktay> yeah
17:06 < digitaloktay> http://omploader.org/vNTBlMA
17:09 < rxr> guess while writing their kernel they found it comfortable to glue it together was they went
17:10 -!- T-One [~T-One@194.183.141.238] has quit [Quit: Verlassend]
17:11 < rogermason> Does anyone have any idea why atk cannot find aclocal-1.11? aclocal-1.11: couldn't open directory `//share/aclocal': aclocal-1.11 is certainly present in /usr/share. This is on a machine installed with 8.0, from a downloaded iso. Thanks.
17:38 < rxr> hm
18:49 < rxr> damn, I need a way more performance storage architecture here
18:51 < koan> uhm the minimal ISO image doesn't contain svn? how can I install packages then?
18:51 < rxr> it should contain svn
18:51 < rxr> which image? x86_64?
18:52 < koan> what's the path?
18:52 < koan> yes
18:52 < rxr> err, on my ppc64 I just type on it's /usr/bin/svn
18:52 < koan> nope, it isn't here
18:52 < rxr> I think I tested a partial t2 rebuild from svn checkout in both x86 images
18:52 < rxr> did you do a full install?
18:52 < rxr> theoretically it is possible to leave out packages in the "isntaller"
18:53 < koan> hm I chose the minimal package set
18:53 < rxr> aha!
18:53 < koan> ok, I thought minimal was enough, my fault
18:53 < rxr> (while it is nice that this still, works, haven't used that in years :-), use he full test :-)
18:53 < rxr> well, certainly enough for very small setups
18:54 < koan> 729M used
18:54 < rxr> you could use stone and the mine frontend therein to install more bianry packages from the CD
18:54 < rxr> however, I even do not use that often, probably not super smooth :-(
18:54 < rxr> this "minimal set" was more used when we used to build very full desktop / server binaries sets some decade ago
18:55 < rxr> nowadays I intentionally publish minimal builds as most people tend not to need 66% of the former big desktop/server set
18:55 < rxr> so it just wasts build time and traffic
18:56 < rxr> we decided some years ago that the "openbsd"-alike minimal target is a nice base system to start to emerge xfce & co
18:56 < koan> ok I misunderstood the minimal package set, as I thought it equaled the full package set of the minimal ISO
18:56 < rxr> yeah, I completely forgot that the installer has that option :-)(
18:56 < rxr> guess I should make it more clear
18:56 < rxr> btw. the installer currently can not show the packages sizes anymore since we removed some historic clutter from the packages
18:57 < rxr> I had no time to fix the installer to read the sizes from the more sorted, straight-forard tarball files, yet :-(
18:57 < rxr> but our plan was to write this binary handling anyway ....
18:57 < koan> ok, installing all packages now from stone
18:57 < koan> ok that was fast, there seems no big difference between minimal and all :-)
18:58 < koan> and now svn is there too
18:58 < rxr> yes as I wrote the "new", "official" isos are already a minimal package selection
18:58 < rxr> the installer just happens to know another, old minimal set for the former huge (DVD, multi-CD) desktop/server builds, ...
18:59 < koan> seems fair to me, T2 feels a lot like BSD systems this way
18:59 < rxr> patches to remove that, or improve it welcome :-)
18:59 < rxr> maybe I get some more inspiration from the openbsd installation
18:59 < rxr> once I saw it I liked some parts of it
19:00 < koan> btw, I'm using FreeBSD now as my main desktop since half a year, I really like it
19:00 < rxr> :_)
19:00 < rxr> I had many friends at university who enjoyed it
19:00 < rxr> however, personally, I'd rather not go the route to another monolythic kernel
19:01 < rxr> I'm anooyed enough when my rotuer crashes just because of some strange, yet untracked, typo in the PPPoE code of Linux, ...
19:01 < rxr> I'd rahter envision a multi-server micro-kernel OS restart just thtat pppoed, ipv4d, or 3comd, ...
19:02 < koan> of course breadth of hardware support is the main problem with all those microkernels
19:02 < rxr> yeah, which also kept me on Linux instead of an BSD the last decade, ...
19:02 < koan> they are no match for Linux with respect to drivers
19:02 < rogermason> rxr: I got past the aclocal problem in atk by replacing $root/$SDECFG_PKG_GNOME2_CORE_PREFIX/share/aclocal with $root/usr/share/aclocal. Unfortunately I now run into another autotools error: configure.in:166: error: possibly undefined macro: AM_PATH_GLIB_2_0.
19:03 < rxr> though dragonfly certainly has a pretty good support being so *BSD derived
19:03 < rxr> but, actually, developing drivers on a real micro-kernel OS is way more fun and quicker, as not every typo is taking your whole OS down
19:03 < rxr> unless you happen to code on the scheduler, or memory mapper :-)
19:03 < rxr> rogermason: I just happen to re-emerge half of my system on a fresh ppc64 install in the hope to see something simillar as you get
19:05 < rogermason> rxr: I seem to collect errors: it seems that 90% of what I attempt fails :-(
19:05 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has joined #t2
19:05 [Users #t2]
19:05 [@ChanServ] [ digitaloktay] [ koan ] [ mtr ] [ Stealth]
19:05 [ Asmundr ] [ dileks ] [ mjungwirth ] [ rogermason] [ TobiX ]
19:05 [ CIA-29 ] [ ian_wiz ] [ mjungwirth2] [ rxr ] [ zod ]
19:05 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
19:06 < rxr> rogermason: sorry, the emerging from .cache files was a little rough
19:06 < rxr> rogermason: I definetly want to see that improved just work on it
19:06 < rxr> and then the usual .0 release fallout
19:09 < CIA-29> rene * r36887 /trunk/package/ (68 files in 68 dirs):
19:09 < CIA-29> * updated .cache files with invalied gnue* dependencies, also
19:09 < CIA-29> clears off quite some general python noise
19:09 < CIA-29> rene * r36888 /trunk/package/ (23 files in 23 dirs): * clear some more wxpython related noise
19:10 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has quit [Quit: Leaving]
19:17 -!- mjungwirth2 [~mjungwirt@213.174.252.112] has quit [Quit: Nettalk6 - www.ntalk.de]
19:23 < rxr> rogermason: good news! I can reproduce this atk failure on ppc64 :-)
19:35 < CIA-29> rene * r36889 /trunk/package/ (1254 files in 1254 dirs):
19:35 < CIA-29> * updated most .cache files that where from 2005, removes more noise
19:35 < CIA-29> (e.g. gtk+12, and friends) than it adds, ...
19:40 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has joined #t2
19:40 < CIA-29> rene * r36890 /trunk/package/gnome2/ (7 files in 7 dirs): * updated some more gnome2 .cache files
19:41 < CIA-29> rene * r36891 /trunk/package/ (7 files in 7 dirs): * updated .cache files that included the old, split bluez-* references
19:42 < rogermason> rxr: Your good news is good. I think :-)
19:42 < ian_wiz> Hello, n00b here, with a few questions, anyone have a moment to spare?
19:42 < rxr> just type your questions and we'll see
19:43 < ian_wiz> ok, I know the download -cfg sys -req...is no necessary with the 8.0 build
19:44 < CIA-29> rene * r36892 /trunk/package/ (29 files in 29 dirs): * updated yet more .cache files that referenced the aging "fam" package
19:44 < ian_wiz> but in the interest of thorough-ness, i ran it, and dirttree fails bc the address does not resolve. problem, or not so much?
19:44 < rxr> ian_wiz: yes, Build-Target / Emerge now fetch missing downloads on demand (for some years)
19:45 < rogermason> rxr: by the way, there is no need to apologise. After all you are making your hard work available to anyone who wishes to use T2.
19:45 < ian_wiz> got ya, i noticed the PDF manual is slightly out of date, no worries! =)
19:45 < ian_wiz> second, if I have a kernel config from another build system, ie arago/oe/angstrom, can i "cut & paste" into custom kernel config?
19:46 < rxr> yes, either if you enable custom kernel config in the Config menu, or by placing it in your target/.../kernel.conf, I generally advise to create whole new targets in any case to keep all your config, patches, etc. groupped ogether
19:46 < rxr> together even
19:47 < CIA-29> rene * r36893 /trunk/package/ (25 files in 25 dirs): * updated yet more .cache files referencing deprecated gnome1* packages
19:48 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has quit [Read error: Connection reset by peer]
19:48 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has joined #t2
19:48 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has quit [Changing host]
19:48 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has joined #t2
19:49 < ian_wiz> roger that, i e-mailed the list about creating a board config for OMAPl138 Hawkboard and Leopard DM365 (TI-DM365-evm), is there a FAQ on building a new board config?
19:49 < rxr> nope
19:49 < rxr> but for the main part it is "just" getting the 3rd party kernel patches together, and choosing the right (ARM) cpu optimization option, ...
19:50 < rxr> the sad thing about ARM is that you nearly need a custom kernel per device :-(
19:51 < ian_wiz> i hear ya...where does the beagle board config reside, so I can review it's construction?
19:51 < rxr> we don't have a beagle board config
19:51 < CIA-29> rene * r36894 /trunk/package/ (37 files in 37 dirs): * updated .cache referencing zope and setuptools
19:51 < rxr> if you start a new target you could place the patches in your target/.../pkg_linux26.patch
19:52 < rxr> or if you have many many target/.../package/linux26/*many*file*names*.patch
19:52 < rxr> as this is mostly about getting the right cpu glue bits into the linux kernel anyway
19:52 < rxr> (usually not all upstream)
19:53 < rxr> avr32 and blackfin we used to have some more overlay
19:53 < rxr> but, I could help with a beagleboard example, actually have one lingering on some table here
19:53 < ian_wiz> 10.4 I know we can specify git mirrors into the build process, can we pull down the linux-davinci git repository? (I read the whole manual last night, but am suffering from mental juicy poos right now)
19:54 < rxr> you could indeed pull whole git repositories
19:56 < ian_wiz> ok, i will do homework on that tonight. i appreciate the help.
19:57 < ian_wiz> Endgame is to get Puppy onto ARMv5 and up, the Hawkboard has built in SATA and for DVR purposes this could be nice.
19:57 < rxr> :-)
19:57 < rxr> I know companies are using T2 for ARM products
19:58 < rxr> ARM was one of my first non-x86 stuff to work on, back on the early days, with Psion hardware
19:58 < rxr> http://www.t2-project.org/targets/psion/
19:59 < ian_wiz> It's my first leap into non-x86, and out of Win enviro. Sort of like working on British cars, not easy, but satisfying in the end.
20:00 < rxr> :-)
20:00 < ian_wiz> alright, i'm going to do homework, I appreciate the help rxr and will no doubt be back. cheers!
20:00 < rxr> cu
20:01 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has quit [Quit: Leaving]
20:06 < CIA-29> rene * r36895 /trunk/package/ (6 files in 6 dirs): * updated .cache that depended on *makedep*
20:09 < CIA-29> rene * r36896 /trunk/package/ (4 files in 4 dirs): * updated some more random .cache files
20:11 < CIA-29> rene * r36897 /trunk/package/ (18 files in 18 dirs): * yet more .cache files which do look better, now
20:15 < rxr> rogermason: found the issue, commit in a sec
20:20 < CIA-29> rene * r36898 /trunk/package/ (9 files in 9 dirs): * yet more .cache file updates due overly insane deps
20:26 < rxr> http://www.t2-project.org/packages/gperf.html
20:27 [Users #t2]
20:27 [@ChanServ] [ digitaloktay] [ mjungwirth] [ rxr ] [ zod]
20:27 [ Asmundr ] [ dileks ] [ mtr ] [ Stealth]
20:27 [ CIA-29 ] [ koan ] [ rogermason] [ TobiX ]
20:27 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
20:28 < rxr> note to newcomers, I sometimes paste package matrix links here to increase the relevance for our google custom search, ... just in case someone wonders :-)
20:31 < koan> what is the difference again between generic system and desktop (for the target)
20:35 < rxr> the generic builds nearly all packages without much frills (read patches etc.)
20:36 < rxr> (like vanilla GNU/Linux) nowadays the generic target has package (pre-) selection templates, like minimal, ...
20:36 < rxr> we default to minimal for some time so that people do not build excessi stuff they do not need
20:37 < rxr> the desktop target is a desktop pre-selection and prevously has some prettification pathces applies
20:38 < rxr> nowadays it's superflous, guess I can rather delete it now ...
20:42 < CIA-29> rene * r36900 /trunk/package/gnome2/ (6 files in 6 dirs):
20:42 < CIA-29> * fixed some stray build errors due direct use of a potentially
20:42 < CIA-29> undefined $SDECFG_PKG_GNOME2_CORE_PREFIX, use pkgprefix instead!
20:43 < rxr> rogermason: the last commit should fix your atk error
20:43 < rxr> all the .cache updates should improve Emerge behaviour
20:43 < rxr> I'll further keep an eye on improving just Emerge experience over the next weeks
20:43 < rxr> (while I update my G5, and G4 Cube)
20:44 < rxr> cu then
20:48 < CIA-29> rene * r36899 /trunk/package/gnome2/gobject-introspection/gobject-introspection.conf: * guided gobject-introspection to build without cairo (just for the tests, ...)
21:07 < rogermason> rxr: so I suppose I should check out trunk to get the fix?
21:10 < koan> so I'm working now with 8.0, should I upgrade to trunk? which updates go into 8.0 now and which ones into trunk?
21:13 < koan> btw is it also possible to configure build options for specific packages, like make config does in FreeBSD ports?
21:49 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has joined #t2
21:51 < ian_wiz> Hello, after first build, BusyBox aborted due to bad value for -mtune= switch (arm926ej-s), thoughts?
22:27 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has quit [Ping timeout: 260 seconds]
22:43 < rxr> koan: you can emerge from either trunk or 8.0
22:43 < rxr> currently it doesnot matter much, but trunk has some fixes ahead
22:43 < rxr> most will be merged back to 8.0 if they proof ok
22:43 < rxr> Config has package
22:43 < rxr> options
22:44 < rxr> now we are even slashdotted
22:44 < rxr> kind of first time ever
22:46 < rxr> http://linux.slashdot.org/story/10/07/22/1852234/Open-Source-OCR-That-Makes-Searchable-PDFs
23:25 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has quit [Quit: Leaving]
--- Log closed Fri Jul 23 00:00:14 2010
00:08 -!- Asmundr [~jamie@h113.143.102.166.static.ip.windstream.net] has joined #t2
01:22 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has quit [Read error: Connection reset by peer]
08:44 -!- mjungwirth [~mjungwirt@213.174.234.68] has quit [Quit: Nettalk6 - www.ntalk.de]
08:46 -!- mjungwirth [~mjungwirt@213.174.234.68] has joined #t2
09:33 < CIA-29> aldas * r36877 /trunk/package/php/php/php.desc: * updated php (5.3.2 -> 5.3.3)
09:33 < CIA-29> aldas * r36878 /trunk/package/develop/git/git.desc: * updated git (1.7.1.1 -> 1.7.2)
10:28 < CIA-29> rene * r36879 /trunk/package/develop/gambas/gambas.desc: * updated gambas (1.0.19 -> 2.18.0)
10:32 < CIA-29> rene * r36881 /trunk/package/contrib/beep/beep.cache: * added beep.cache
10:41 < CIA-29> rene * r36880 /trunk/package/contrib/beep/beep.desc: * updated beep (0.0.1 -> 1.3)
11:06 < CIA-29> rene * r36882 /trunk/package/network/tcpdump/ (tcpdump.conf tcpdump.desc):
11:06 < CIA-29> Mario Jungwirth
11:06 < CIA-29> * improved tcpdump to cross compile
11:10 < CIA-29> rene * r36883 /trunk/package/base/htop/ (configure.patch.cross htop.conf htop.desc):
11:10 < CIA-29> Mario Jungwirth
11:10 < CIA-29> * improved htop to cross compile
11:13 < CIA-29> rene * r36884 /trunk/package/x86/syslinux/syslinux.desc: * updated syslinux (4.01 -> 4.02)
11:13 < CIA-29> rene * r36885 /trunk/package/database/sqlite/sqlite.desc: * updated sqlite (3.6.23.1 -> 3.7.0)
11:33 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has joined #t2
11:33 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has quit [Changing host]
11:33 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has joined #t2
11:38 -!- arete74 [~arete@143.225.107.252] has joined #t2
12:04 < CIA-29> rene * r36886 /trunk/package/editors/texmacs/texmacs.desc: * updated texmacs (1.0.7.3 -> 1.0.7.5)
12:12 -!- uid9909 [~uid9099@2001:470:9a41:107:4cda:5044:257c:e722] has joined #t2
12:24 [Users #t2]
12:24 [@ChanServ] [ CIA-29 ] [ koan ] [ rxr ] [ TobiX ]
12:24 [ arete74 ] [ digitaloktay] [ mjungwirth] [ Stealth] [ uid9909]
12:24 [ Asmundr ] [ dileks ] [ mtr ] [ T-One ] [ zod ]
12:24 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
12:44 * rxr to lunch - cu
12:54 -!- arete74 [~arete@143.225.107.252] has quit [Quit: leaving]
13:37 -!- rogermason [~user@CPE001b639d4751-CM0011aea11450.cpe.net.cable.rogers.com] has joined #t2
13:39 < rogermason> I'm working through Emerging the dependencies to get emacs installed. glitz fails for lack of GL/gl.h. Is that provided by mesa? Thanks.
14:07 -!- uid9909 [~uid9099@2001:470:9a41:107:4cda:5044:257c:e722] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.7/20100713130626]]
15:06 < koan> curious, the Wikipedia notability nazis have deleted the page about T2 SDE: https://secure.wikimedia.org/wikipedia/en/wiki/T2_SDE
15:10 < rxr> oh my
15:11 < rxr> sucks
15:14 < koan> the german wikipedia moderators are more sane
15:39 < koan> rxr: the build intro page still calls 6.0 the "latest" release: http://t2-project.org/documentation/buildintro.html
15:40 < koan> (there may be other outdated pages too)
15:42 < rxr> updated
15:45 < koan> maybe grep for "6.0" and "7.0" through the website's sources :-)
15:45 < rxr> tried, pretty much noise
15:51 < rxr> if you have questinos jsut ask :-)
15:53 < koan> yup, I'm just skimming through the website now to build the skeleton of the article
15:55 < koan> hm I hadn't seen the whitepaper before
16:00 < rogermason> Got past the GL issue. Now atk won't build: CDPATH="${ZSH_VERSION+.}:" && cd . && /bin/sh /root/t2-8.0/src.atk.SDE-CONFIG.20100722.112349.1679.gem/atk-1.30.0/missing --run aclocal-1.11 $
16:00 < rogermason> aclocal-1.11: couldn't open directory `//share/aclocal': No such file or directory
16:00 < rogermason> make: *** [aclocal.m4] Error 1
16:00 < rogermason> Due to previous errors, no 9-atk.log file!
16:00 < rogermason>
16:01 < rxr> which GL issue ?
16:04 < rogermason>
16:04 < rogermason> installed. glitz fails for lack of GL/gl.h. Is that provided by
16:04 < rogermason> mesa? Thanks.
16:05 < rogermason> But I built mesa and that fixed the "issue".
16:37 -!- mjungwirth2 [~mjungwirt@213.174.252.112] has joined #t2
16:54 < koan> rxr: is the long-term goal of support for other open source kernels still being pursued?
16:55 < rxr> yes, but I'm not yet sure which one to use
16:55 < rxr> i'm not fully happy with either of them
16:55 < rxr> minix is too lagging
16:55 < rxr> draonfly the better one so far
16:55 < koan> I remember you looked at DragonFly BSD yes
16:55 < rxr> but I started with mingw / win32 support :-)
16:56 < koan> and how difficult is it to port T2 to another kernel? which pieces have to be changed?
16:56 < rxr> I'm actually thinking about re-using L4 bits and writing my own multi-server kernel as a hobby
16:57 < rxr> well, beside patching some of the packages (pkgsrc like) to build with the other kernel at all, it is not so difficult
16:57 < rxr> mostly only the kernel sources need to be extracted from their archives (most are not as nicely tarball packaged as linux, but just resist in their SVN, etc. tree)
16:57 < rxr> and with the kernel the c library and maybe some system control tools
16:57 < rxr> it's actually not sooo much work
16:57 < rxr> even beginners could do so
16:58 < rxr> it's mostly that I personally need something more micr-kernel (read stable) than Linux, and I have not yet pinpointed the best solution, yet
16:58 < koan> hm ok, and I guess they can look at how the mingw/win32 port was done for a start?
16:59 < rxr> yes, though mingw does not come with a kernel, ... but the basics are the same
16:59 < rxr> especially wheter the gcc release tarball can produce working executables for said platform it's an easy game, just package the system header and libc and you're already halfway there
17:00 < rxr> e.g. most normal packages should already build, ... and then just collecting all their external kernel sources into a nice kernel packages, ...
17:01 < koan> well, this idea of one userland (T2) with different choices of kernel has always intrigued me, so I really hope I still see it happening once :-)
17:01 < rxr> well, for most part t2 is like those pkgsrc trees, ...
17:01 < rxr> ports collection
17:02 < rxr> btw. the Haiku folks even considered using T2^W ROCK Linux back in their early days when they still named itself openbsd
17:02 < koan> interesting
17:03 < koan> why did they decide to not do it?
17:06 < digitaloktay> yeah
17:06 < digitaloktay> http://omploader.org/vNTBlMA
17:09 < rxr> guess while writing their kernel they found it comfortable to glue it together was they went
17:10 -!- T-One [~T-One@194.183.141.238] has quit [Quit: Verlassend]
17:11 < rogermason> Does anyone have any idea why atk cannot find aclocal-1.11? aclocal-1.11: couldn't open directory `//share/aclocal': aclocal-1.11 is certainly present in /usr/share. This is on a machine installed with 8.0, from a downloaded iso. Thanks.
17:38 < rxr> hm
18:49 < rxr> damn, I need a way more performance storage architecture here
18:51 < koan> uhm the minimal ISO image doesn't contain svn? how can I install packages then?
18:51 < rxr> it should contain svn
18:51 < rxr> which image? x86_64?
18:52 < koan> what's the path?
18:52 < koan> yes
18:52 < rxr> err, on my ppc64 I just type on it's /usr/bin/svn
18:52 < koan> nope, it isn't here
18:52 < rxr> I think I tested a partial t2 rebuild from svn checkout in both x86 images
18:52 < rxr> did you do a full install?
18:52 < rxr> theoretically it is possible to leave out packages in the "isntaller"
18:53 < koan> hm I chose the minimal package set
18:53 < rxr> aha!
18:53 < koan> ok, I thought minimal was enough, my fault
18:53 < rxr> (while it is nice that this still, works, haven't used that in years :-), use he full test :-)
18:53 < rxr> well, certainly enough for very small setups
18:54 < koan> 729M used
18:54 < rxr> you could use stone and the mine frontend therein to install more bianry packages from the CD
18:54 < rxr> however, I even do not use that often, probably not super smooth :-(
18:54 < rxr> this "minimal set" was more used when we used to build very full desktop / server binaries sets some decade ago
18:55 < rxr> nowadays I intentionally publish minimal builds as most people tend not to need 66% of the former big desktop/server set
18:55 < rxr> so it just wasts build time and traffic
18:56 < rxr> we decided some years ago that the "openbsd"-alike minimal target is a nice base system to start to emerge xfce & co
18:56 < koan> ok I misunderstood the minimal package set, as I thought it equaled the full package set of the minimal ISO
18:56 < rxr> yeah, I completely forgot that the installer has that option :-)(
18:56 < rxr> guess I should make it more clear
18:56 < rxr> btw. the installer currently can not show the packages sizes anymore since we removed some historic clutter from the packages
18:57 < rxr> I had no time to fix the installer to read the sizes from the more sorted, straight-forard tarball files, yet :-(
18:57 < rxr> but our plan was to write this binary handling anyway ....
18:57 < koan> ok, installing all packages now from stone
18:57 < koan> ok that was fast, there seems no big difference between minimal and all :-)
18:58 < koan> and now svn is there too
18:58 < rxr> yes as I wrote the "new", "official" isos are already a minimal package selection
18:58 < rxr> the installer just happens to know another, old minimal set for the former huge (DVD, multi-CD) desktop/server builds, ...
18:59 < koan> seems fair to me, T2 feels a lot like BSD systems this way
18:59 < rxr> patches to remove that, or improve it welcome :-)
18:59 < rxr> maybe I get some more inspiration from the openbsd installation
18:59 < rxr> once I saw it I liked some parts of it
19:00 < koan> btw, I'm using FreeBSD now as my main desktop since half a year, I really like it
19:00 < rxr> :_)
19:00 < rxr> I had many friends at university who enjoyed it
19:00 < rxr> however, personally, I'd rather not go the route to another monolythic kernel
19:01 < rxr> I'm anooyed enough when my rotuer crashes just because of some strange, yet untracked, typo in the PPPoE code of Linux, ...
19:01 < rxr> I'd rahter envision a multi-server micro-kernel OS restart just thtat pppoed, ipv4d, or 3comd, ...
19:02 < koan> of course breadth of hardware support is the main problem with all those microkernels
19:02 < rxr> yeah, which also kept me on Linux instead of an BSD the last decade, ...
19:02 < koan> they are no match for Linux with respect to drivers
19:02 < rogermason> rxr: I got past the aclocal problem in atk by replacing $root/$SDECFG_PKG_GNOME2_CORE_PREFIX/share/aclocal with $root/usr/share/aclocal. Unfortunately I now run into another autotools error: configure.in:166: error: possibly undefined macro: AM_PATH_GLIB_2_0.
19:03 < rxr> though dragonfly certainly has a pretty good support being so *BSD derived
19:03 < rxr> but, actually, developing drivers on a real micro-kernel OS is way more fun and quicker, as not every typo is taking your whole OS down
19:03 < rxr> unless you happen to code on the scheduler, or memory mapper :-)
19:03 < rxr> rogermason: I just happen to re-emerge half of my system on a fresh ppc64 install in the hope to see something simillar as you get
19:05 < rogermason> rxr: I seem to collect errors: it seems that 90% of what I attempt fails :-(
19:05 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has joined #t2
19:05 [Users #t2]
19:05 [@ChanServ] [ digitaloktay] [ koan ] [ mtr ] [ Stealth]
19:05 [ Asmundr ] [ dileks ] [ mjungwirth ] [ rogermason] [ TobiX ]
19:05 [ CIA-29 ] [ ian_wiz ] [ mjungwirth2] [ rxr ] [ zod ]
19:05 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
19:06 < rxr> rogermason: sorry, the emerging from .cache files was a little rough
19:06 < rxr> rogermason: I definetly want to see that improved just work on it
19:06 < rxr> and then the usual .0 release fallout
19:09 < CIA-29> rene * r36887 /trunk/package/ (68 files in 68 dirs):
19:09 < CIA-29> * updated .cache files with invalied gnue* dependencies, also
19:09 < CIA-29> clears off quite some general python noise
19:09 < CIA-29> rene * r36888 /trunk/package/ (23 files in 23 dirs): * clear some more wxpython related noise
19:10 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has quit [Quit: Leaving]
19:17 -!- mjungwirth2 [~mjungwirt@213.174.252.112] has quit [Quit: Nettalk6 - www.ntalk.de]
19:23 < rxr> rogermason: good news! I can reproduce this atk failure on ppc64 :-)
19:35 < CIA-29> rene * r36889 /trunk/package/ (1254 files in 1254 dirs):
19:35 < CIA-29> * updated most .cache files that where from 2005, removes more noise
19:35 < CIA-29> (e.g. gtk+12, and friends) than it adds, ...
19:40 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has joined #t2
19:40 < CIA-29> rene * r36890 /trunk/package/gnome2/ (7 files in 7 dirs): * updated some more gnome2 .cache files
19:41 < CIA-29> rene * r36891 /trunk/package/ (7 files in 7 dirs): * updated .cache files that included the old, split bluez-* references
19:42 < rogermason> rxr: Your good news is good. I think :-)
19:42 < ian_wiz> Hello, n00b here, with a few questions, anyone have a moment to spare?
19:42 < rxr> just type your questions and we'll see
19:43 < ian_wiz> ok, I know the download -cfg sys -req...is no necessary with the 8.0 build
19:44 < CIA-29> rene * r36892 /trunk/package/ (29 files in 29 dirs): * updated yet more .cache files that referenced the aging "fam" package
19:44 < ian_wiz> but in the interest of thorough-ness, i ran it, and dirttree fails bc the address does not resolve. problem, or not so much?
19:44 < rxr> ian_wiz: yes, Build-Target / Emerge now fetch missing downloads on demand (for some years)
19:45 < rogermason> rxr: by the way, there is no need to apologise. After all you are making your hard work available to anyone who wishes to use T2.
19:45 < ian_wiz> got ya, i noticed the PDF manual is slightly out of date, no worries! =)
19:45 < ian_wiz> second, if I have a kernel config from another build system, ie arago/oe/angstrom, can i "cut & paste" into custom kernel config?
19:46 < rxr> yes, either if you enable custom kernel config in the Config menu, or by placing it in your target/.../kernel.conf, I generally advise to create whole new targets in any case to keep all your config, patches, etc. groupped ogether
19:46 < rxr> together even
19:47 < CIA-29> rene * r36893 /trunk/package/ (25 files in 25 dirs): * updated yet more .cache files referencing deprecated gnome1* packages
19:48 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has quit [Read error: Connection reset by peer]
19:48 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has joined #t2
19:48 -!- digitaloktay [~win7@dslb-084-056-245-180.pools.arcor-ip.net] has quit [Changing host]
19:48 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has joined #t2
19:49 < ian_wiz> roger that, i e-mailed the list about creating a board config for OMAPl138 Hawkboard and Leopard DM365 (TI-DM365-evm), is there a FAQ on building a new board config?
19:49 < rxr> nope
19:49 < rxr> but for the main part it is "just" getting the 3rd party kernel patches together, and choosing the right (ARM) cpu optimization option, ...
19:50 < rxr> the sad thing about ARM is that you nearly need a custom kernel per device :-(
19:51 < ian_wiz> i hear ya...where does the beagle board config reside, so I can review it's construction?
19:51 < rxr> we don't have a beagle board config
19:51 < CIA-29> rene * r36894 /trunk/package/ (37 files in 37 dirs): * updated .cache referencing zope and setuptools
19:51 < rxr> if you start a new target you could place the patches in your target/.../pkg_linux26.patch
19:52 < rxr> or if you have many many target/.../package/linux26/*many*file*names*.patch
19:52 < rxr> as this is mostly about getting the right cpu glue bits into the linux kernel anyway
19:52 < rxr> (usually not all upstream)
19:53 < rxr> avr32 and blackfin we used to have some more overlay
19:53 < rxr> but, I could help with a beagleboard example, actually have one lingering on some table here
19:53 < ian_wiz> 10.4 I know we can specify git mirrors into the build process, can we pull down the linux-davinci git repository? (I read the whole manual last night, but am suffering from mental juicy poos right now)
19:54 < rxr> you could indeed pull whole git repositories
19:56 < ian_wiz> ok, i will do homework on that tonight. i appreciate the help.
19:57 < ian_wiz> Endgame is to get Puppy onto ARMv5 and up, the Hawkboard has built in SATA and for DVR purposes this could be nice.
19:57 < rxr> :-)
19:57 < rxr> I know companies are using T2 for ARM products
19:58 < rxr> ARM was one of my first non-x86 stuff to work on, back on the early days, with Psion hardware
19:58 < rxr> http://www.t2-project.org/targets/psion/
19:59 < ian_wiz> It's my first leap into non-x86, and out of Win enviro. Sort of like working on British cars, not easy, but satisfying in the end.
20:00 < rxr> :-)
20:00 < ian_wiz> alright, i'm going to do homework, I appreciate the help rxr and will no doubt be back. cheers!
20:00 < rxr> cu
20:01 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has quit [Quit: Leaving]
20:06 < CIA-29> rene * r36895 /trunk/package/ (6 files in 6 dirs): * updated .cache that depended on *makedep*
20:09 < CIA-29> rene * r36896 /trunk/package/ (4 files in 4 dirs): * updated some more random .cache files
20:11 < CIA-29> rene * r36897 /trunk/package/ (18 files in 18 dirs): * yet more .cache files which do look better, now
20:15 < rxr> rogermason: found the issue, commit in a sec
20:20 < CIA-29> rene * r36898 /trunk/package/ (9 files in 9 dirs): * yet more .cache file updates due overly insane deps
20:26 < rxr> http://www.t2-project.org/packages/gperf.html
20:27 [Users #t2]
20:27 [@ChanServ] [ digitaloktay] [ mjungwirth] [ rxr ] [ zod]
20:27 [ Asmundr ] [ dileks ] [ mtr ] [ Stealth]
20:27 [ CIA-29 ] [ koan ] [ rogermason] [ TobiX ]
20:27 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
20:28 < rxr> note to newcomers, I sometimes paste package matrix links here to increase the relevance for our google custom search, ... just in case someone wonders :-)
20:31 < koan> what is the difference again between generic system and desktop (for the target)
20:35 < rxr> the generic builds nearly all packages without much frills (read patches etc.)
20:36 < rxr> (like vanilla GNU/Linux) nowadays the generic target has package (pre-) selection templates, like minimal, ...
20:36 < rxr> we default to minimal for some time so that people do not build excessi stuff they do not need
20:37 < rxr> the desktop target is a desktop pre-selection and prevously has some prettification pathces applies
20:38 < rxr> nowadays it's superflous, guess I can rather delete it now ...
20:42 < CIA-29> rene * r36900 /trunk/package/gnome2/ (6 files in 6 dirs):
20:42 < CIA-29> * fixed some stray build errors due direct use of a potentially
20:42 < CIA-29> undefined $SDECFG_PKG_GNOME2_CORE_PREFIX, use pkgprefix instead!
20:43 < rxr> rogermason: the last commit should fix your atk error
20:43 < rxr> all the .cache updates should improve Emerge behaviour
20:43 < rxr> I'll further keep an eye on improving just Emerge experience over the next weeks
20:43 < rxr> (while I update my G5, and G4 Cube)
20:44 < rxr> cu then
20:48 < CIA-29> rene * r36899 /trunk/package/gnome2/gobject-introspection/gobject-introspection.conf: * guided gobject-introspection to build without cairo (just for the tests, ...)
21:07 < rogermason> rxr: so I suppose I should check out trunk to get the fix?
21:10 < koan> so I'm working now with 8.0, should I upgrade to trunk? which updates go into 8.0 now and which ones into trunk?
21:13 < koan> btw is it also possible to configure build options for specific packages, like make config does in FreeBSD ports?
21:49 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has joined #t2
21:51 < ian_wiz> Hello, after first build, BusyBox aborted due to bad value for -mtune= switch (arm926ej-s), thoughts?
22:27 -!- digitaloktay [~win7@unaffiliated/digitaloktay] has quit [Ping timeout: 260 seconds]
22:43 < rxr> koan: you can emerge from either trunk or 8.0
22:43 < rxr> currently it doesnot matter much, but trunk has some fixes ahead
22:43 < rxr> most will be merged back to 8.0 if they proof ok
22:43 < rxr> Config has package
22:43 < rxr> options
22:44 < rxr> now we are even slashdotted
22:44 < rxr> kind of first time ever
22:46 < rxr> http://linux.slashdot.org/story/10/07/22/1852234/Open-Source-OCR-That-Makes-Searchable-PDFs
23:25 -!- ian_wiz [~ian@c-98-236-69-69.hsd1.pa.comcast.net] has quit [Quit: Leaving]
--- Log closed Fri Jul 23 00:00:14 2010