T2 IRC Log: 2009-10-18

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.

« prev | next »

--- Log opened Sun Oct 18 00:00:43 2009
00:49 -!- mqueiros_ [n=mqueiros@217.70.75.179] has joined #t2
03:48 -!- koan [n=koan@unaffiliated/koan] has quit [Read error: 60 (Operation timed out)]
03:49 -!- koan [n=koan@unaffiliated/koan] has joined #t2
11:44 -!- mpp_ [n=mpp@2a01:48:251:107:203:dff:fe13:6c9e] has joined #t2
12:57 < rxr> re
13:01 -!- dpc [n=dpc@chello089075192131.chello.pl] has joined #t2
13:11 < CIA-38> rene * r34576 /trunk/package/base/sysfiles/sysfiles.conf: * do not extra log btee compilation, skip O2, standard options and transforms apply, anyway
13:11 -!- dpc [n=dpc@chello089075192131.chello.pl] has left #t2 []
13:13 < rxr> koan: in case you want to publish about t2 8.0, please hold of some days until the -rc2
13:14 < rxr> some first-all-cross-build falloff dust has to settle :_)
13:20 < CIA-38> rene * r34577 /trunk/package/base/sysfiles/sysfiles.conf: * also run the postsysfile scripts on cross build
13:27 -!- LMJ [n=serwou@laf31-4-82-236-42-164.fbx.proxad.net] has left #t2 ["Leaving"]
13:28 < CIA-38> rene * r34578 /trunk/package/base/sysfiles/sysfiles.conf:
13:28 < CIA-38> * reindent sysfiles.conf after the previous conditional removal
13:28 < CIA-38> * polish some T2-style whitespace and newlines on the way
13:40 -!- dpc [n=dpc@chello089075192131.chello.pl] has joined #t2
13:40 < dpc> Hi. Anybody available?
13:40 < rxr> yes
13:40 [Users #t2]
13:40 [@ChanServ] [ koan ] [ mpp_ ] [ rxr ] [ T-One]
13:40 [ CIA-38 ] [ mjungwirth_] [ mqueiros_] [ Stealth ] [ TobiX]
13:40 [ dpc ] [ mpp ] [ mtr ] [ synchris]
13:40 -!- Irssi: #t2: Total of 14 nicks [1 ops, 0 halfops, 0 voices, 13 normal]
13:41 < dpc> Cool.
13:41 < dpc> What binary iso do I need to install in a VM to get a system in which building T2 will actually work?
13:41 < dpc> I've tried on my Ubuntu and I was close, but I'm tired now.
13:41 < CIA-38> rene * r34579 /trunk/package/emulators/gxemul/gxemul.desc: * updated gxemul (0.4.6.5 -> 0.4.7.2)
13:41 < rxr> well, in theory it should build on any, Debian, SuSE, etc.
13:42 < dpc> In theory. :)
13:42 < rxr> for t2 it would be best to grab the 7.0-rc2 Id day
13:42 < rxr> say
13:42 < dpc> minimal-i486 ?
13:43 < dpc> ftp://ibiblio.org/pub/linux/distributions/t2/stable/7.0-rc2/ - I'm looking here
13:43 < rxr> yes
13:44 < dpc> t2-7.0-rc2-minimal-i486.iso I have installed this one. But when doing ./scripts/Build-Target it failes on perl something.
13:44 < dpc> I remember having this one on Ubuntu due to lack of some perl related package.
13:45 < rxr> hm
13:45 < rxr> just try on ubunutu
13:45 < rxr> I try to fix bugs that you encounted on the way
13:46 < dpc> The official guide says it is good to grab t2-rescue.
13:46 < rxr> the only thing that might be an issue is the dash or what it was default sh shell
13:46 < rxr> IIRC
13:46 < rxr> yes, that is indeed good
13:46 < rxr> if you have the resoruces grab the 7.0-rc2
13:46 < dpc> But I don't know how to install resouce.
13:46 < dpc> It start live-cd.
13:46 < dpc> *starts
13:47 < rxr> yeah, the t2 livecd have no installer
13:47 < rxr> the minimals in your link above come with a normal installer
13:47 < dpc> Yes. I've just started it.
13:47 < dpc> In VirtualBox.
13:47 < dpc> Installed, svn checkout (trunk), required files downloaded.
13:48 < rxr> sounds good
13:48 < dpc> I'm root and I do ./script/Build-Targe
13:48 < dpc> It build linux-headers
13:49 < dpc> The config is to skip Paranoia, do a crossbuild (it's x86 ->x86 but I seen somewhere that crossbuilding should be used all the time and it worked better even on my ubuntu when doing x64 -> x64).
13:49 < dpc> Generic, minimal system.
13:49 < dpc> Let's wait a bit I'll paste the error.
13:49 < rxr> ok
13:51 < dpc> I was trying to build T2 on my Ubuntu 9.10 beta for a long time and I was close, but it's beyond my T2 understanding at the moment.
13:51 < dpc> However I think you should really do try get it built on other distros.
13:52 < rxr> we do and I was not aware there still problems
13:52 < rxr> I even do built t2 on very old T2 or ROCK Linux versions
13:53 < rxr> e.g. one test server still has bash-2.05b.0, gcc-3.2.3 and was a base install from ROCK Linux 2.0.1 (2004/03/13)
13:53 < rxr> we continously adapt to foreign distro problems, ...
13:53 < dpc> From the T2 minimal install:
13:53 < dpc> Can't locate strict.pm in @INC (@INC contains ... )
13:54 < rxr> I think that was an issue on that t2 release, some perl files missing
13:54 < dpc> I'm unable to copy paste.
13:54 < rxr> yo can just rebuild perl
13:54 < dpc> How? :)
13:54 < rxr> ./scripts/Emerge-Pkg -deps=none perl
13:54 < rxr> with a config matching your cpu
13:54 < dpc> default should be OK I guess. As this is x86 VM building x86 target.
13:55 < rxr> yeah, then just emerge with the above commands
13:55 < dpc> I've spotted some problems when building in Ubuntu.
13:55 < rxr> -deps=none just avoids updating half of the system :-)
13:55 < dpc> package/security/openssl/openssl.conf: (had to add:)
13:55 < dpc> +hook_add premake 6 "touch Makefile"
13:55 < rxr> we'll have new 8.0 releases in the next days
13:56 < dpc> As Makefile was older then some other files and error was printed.
13:56 < rxr> h
13:56 < rxr> hm
13:56 < rxr> intresting issues
13:56 < dpc> Same in the package/base/libpcap/libpcap.conf
13:56 < dpc> +hook_add premake 5 "touch Makefile"
13:57 < dpc> And I remember that one config failed because Makefile contained -ltermcap (or something like this) while I had to change it to -lncurses (I've followed the FixFileDiff instruction)
13:57 < dpc> One sec and I'll tell which one was it.
13:58 < rxr> it's best to always immediately post that on the mailing lsit so it's not forgotten
13:58 < dpc> package/base/procinfo/fix-termcap-to-ncurses.patch
13:58 < dpc> Which list should I use?
13:59 < rxr> there is only one for posting, no? -svn and -update should be read-only aumtatics, so the regular t2 lsit
13:59 < rxr> list
13:59 < rxr> t2 at t2-project dot org
13:59 < dpc> OK. I'll register now.
14:01 < CIA-38> rene * r34580 /trunk/package/emulators/pearpc/gcc43.patch: * fix pearpc to build for x864-64 with gcc-4.3, again
14:15 < dpc> Is t2 mailinglist sends copy of an email to a sender as well?
14:15 < rxr> I think so
14:16 < rxr> I got an mail from ucore dot info
14:16 < dpc> Good. It's me.
14:17 < rxr> maybe, if you have a very strict mail server you might not get mails as the t2-project MX has no revertse dns ptr for the ip :-(
14:18 < dpc> Ulalala. Gmail do I use to host my email. Shall the messages pass, you think? :)
14:18 < rxr> yes, gmail accepts them :-.)
14:19 < dpc> Good to hear that is. :)
14:19 < dpc> I've received an email from minimalist so it should be OK.
14:20 < dpc> T2 VM installation is not building fine. Thank you.
14:20 < rxr> s/not/now/ ?
14:21 < dpc> *now
14:21 < dpc> Right.
14:23 < rxr> all the ppc and sparm emulators (qemu, pearpc) don't like me these days
14:23 < rxr> why the heck are they now more broken then ever
14:23 < rxr> especially qemu
14:24 < rxr> one step forward, 3 steps back
14:24 < rxr> damnit
14:31 < dpc> Crosscomiling is always a hard lifestyle. :)
14:31 < rxr> hm?
14:33 < dpc> Sorry. :)
14:34 < dpc> Whats wrong with a qemu this time?
14:34 < rxr> run time, ppc/sparc openbios broken to no avail
14:34 < rxr> the openhackware hacks worked better, actually loaded anything at all :-)
14:36 < dpc> Heh. I've not played with PPC/SPARC emulation and qemu much.
14:36 < rxr> yeah, of course x86 works way better
14:37 < dpc> I work on SPARC everyday and I must say I like the architecture.
14:37 < rxr> yeah, guess why I do all the non-x86 stuff for T2 :-)
14:37 < dpc> However I'm not sure how does it looks like from the qemu side.
14:37 < rxr> jsut that silo does not load right now and some emulator would be nice(tm) to debug this ...
14:38 < dpc> Oh.
14:38 < dpc> I didn't even know what silo is. Sparc Improved boot LOader...
14:38 < dpc> What's wrong with grub? It's used in Solaris.
14:38 < rxr> yeah, that thing that loads the linux kernel on sparc*
14:39 < rxr> grub until 2.0 beta/alpha did not support sparc
14:39 < rxr> and I bet their proof of concept code in 2.0 is not more stable than silo
14:39 < rxr> grub was until recently x86 only
14:40 < dpc> How come? I can see it .... oh.... right ... . OBP loads Solaris on SPARCS. Right.
14:40 < dpc> But who needs SILO when there is an OBP.
14:41 < rxr> obp can not load linux
14:41 < rxr> for a start you usually have linux on some non OBP support filesystem
14:41 < rxr> and then the loader requirements differ
14:41 < rxr> you got silo to load the linux kernel on sparc*
14:41 < rxr> and yaboot to load linux on powerpc*
14:41 < rxr> aboot for Alpha
14:42 < rxr> elilo for IA64
14:42 < rxr> etc. pp.
14:42 < dpc> Understood. :)
14:43 < rxr> though grub2 starts to support more than x86
14:43 < rxr> how stable that is I leave for futhre experimentation
14:43 < rxr> also I'm not sure if grub2 actually support cd booting on x86
14:43 < rxr> some months ago it did not even support that, ...
14:46 < dpc> What are uses of Linux on Solaris? I work in SUN Microsystem Support and even though I know that you can install Linux on SPARC I wouldn't see much of sense in installing something other than Solaris on these boxes. What are the common purposes of such combo (Linux + SPARC)?
14:47 < rxr> beside fun?
14:47 < dpc> Yes. :)
14:47 < rxr> well, I personally only do fun for t2 on it
14:48 < dpc> I wish I could get a notebook with SPARC inside. That'd be cool.
14:48 < rxr> but I guess bigger institutes do so for the heterogeneous of the infrastructure, same tools as on other linux machines etc.
14:48 < rxr> dpc: there whre sparc notebooks in the past
14:48 < rxr> tough pretty expensive, though
14:49 < rxr> and often loud
14:49 < dpc> I know. I've seen them... on pictures.
14:49 < rxr> http://www.sparcproductdirectory.com/portable10.html
14:50 < rxr> saw some in real :-)
14:50 < rxr> ^ - I
14:50 < dpc> :)
14:50 < rxr> but never owned one, too espensive too loud
14:50 < dpc> Gobi 7 looks cool.
14:51 < CIA-38> rene * r34581 /trunk/package/emulators/pearpc/pearpc.conf: * copy the pearpc config example, without probably no sane person is able to use it
14:51 < CIA-38> rene * r34582 /trunk/package/emulators/pearpc/ (pearpc.cache pearpc.desc): * updated pearpc (0.4 -> 2009-10-18), adds x86-64 JIT, other fixes
14:51 < dpc> However I guess there is not much of the point of using Sparc inside thin client. It could be X64 as well - cheaper and just better.
14:52 < rxr> well, if it would be me to decide, I would can the criptic, complex, engery wasting x86 ISA
14:52 < rxr> and just use PPC, MIPS, ARM, SPARC, etc..
14:53 < dpc> ARM especially. :)
14:53 < rxr> they are also virtualizable without further silicon support, unlike x86 which is not virtualizable per se. and either requires expensive a-head scanning or dedicated further silicon to workaround historic design mistakes ...
14:54 < CIA-38> rene * r34583 /trunk/package/emulators/pearpc/gcc43.patch: * removed pearpc/gcc43.patch, obsolete by CVS update
14:54 < rxr> http://en.wikipedia.org/wiki/Popek_and_Goldberg_virtualization_requirements
14:54 < dpc> It's funny most dominant architecture is the one that is worst of all available. However with mobile devices I guess ARM is even now a leader.
14:55 < rxr> the question is how long
14:55 < rxr> Intel might overturn them in the next years to come
14:55 < dpc> With atoms?
14:55 < rxr> super scaled down Atom, ...
14:55 < rxr> in 3y or so, ... ARM will have an even more difficult time
14:56 < dpc> NOoooooo. ;(
14:56 < dpc> ARM is very nice. SPARC is marvelous in terms of it's instruction set and assembler. And x64 is purely a crap...
14:57 < rxr> it's for a reason smartphones happen to appear on intel roadmaps from time to time in conjunction with the all-in-on-chip (SoC) Atoms after the pinetrail atom
14:57 < rxr> wel, actual x86-64 is not as bad as pure x86
14:58 < rxr> AMD did a pretty good cleanup ...
14:58 < rxr> no wonder their virtualization extension is cleaned, more full featured and performs better than Intel's ...
14:58 < dpc> But is the BIOS to stay with the architecture?
14:58 < dpc> ACPI ?
14:59 < rxr> for some SoC Atoms Intel even plans to can ACPI, let the OS take full control of the chip bits again ...
14:59 < rxr> but the BIOS et al. is not really part of the chip ISA anyway ...
15:00 < dpc> So you can have x64 without BIOS?
15:01 < rxr> sure, with EFI for one
15:01 < rxr> but without fully custom code for another
15:01 < rxr> even with openfirmware if one would want do
15:01 < rxr> to
15:02 < rxr> http://www.google.com/search?q=x86+open+firmware+olpc
15:02 < rxr> http://www.coreboot.org/Welcome_to_coreboot
15:05 < dpc> brb; will read...
15:07 < dpc> BTW. Have you seen this: http://labs.asn.pl/fakebox/
15:36 < rxr> nope
15:48 < rxr> http://www.engadget.com/
15:48 < rxr> heh, amd congo stuff to finally appar ?
15:48 < rxr> appear even
15:53 < rxr> http://rene.rebe.name/2009-10-18/more-amd-congo-stuff-to-appear-soon/
15:57 < CIA-38> rene * r34584 /trunk/architecture/sparc/boot/boot.msg: * use the "right" capitalization in the sparc boot loader message
16:04 -!- capey [n=mika@e81-197-72-70.elisa-laajakaista.fi] has joined #t2
16:04 < rxr> hi capey !
16:04 [Users #t2]
16:04 [@ChanServ] [ dpc ] [ mpp ] [ mtr ] [ synchris]
16:04 [ capey ] [ koan ] [ mpp_ ] [ rxr ] [ T-One ]
16:04 [ CIA-38 ] [ mjungwirth_] [ mqueiros_] [ Stealth] [ TobiX ]
16:04 -!- Irssi: #t2: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal]
16:06 < capey> hi rxr
16:08 < rxr> capey: how's life ?
16:09 < capey> rxr: goog good, thanks for asking, btw. congrats on marriage
16:10 < rxr> thank you!
16:10 < capey> :)
16:11 < rxr> job and such all as the last time we talked?
16:12 < capey> well yeah, got ultra busy from work side, but thats nothing new :)
16:13 < capey> just testing the new 8.0-rc and thought on dropping by here and catch all the latest news
16:14 < rxr> yeah - you're welcome
16:15 < rxr> would be great to see you around more often
16:16 < capey> hopefully i'll be here, will be quite busy end of the year,need to finish beta cycle on software
16:38 -!- dpc [n=dpc@chello089075192131.chello.pl] has left #t2 []
17:10 < capey> puuh,didnt remember that it's this tedious to get qt compiled in this other stuff, 3 hours and still compiling
17:11 < rxr> heh
17:12 < capey> that reminds me, i should resurrect my work on getting qt ui for package selection
17:13 < capey> i'm getting to like too much on point & click software
17:13 < capey> that is, when it works
17:31 < rxr> yeh
17:31 < rxr> yeah
17:50 < CIA-38> rene * r34585 /trunk/architecture/sparc/boot/boot.in: * improved sparc boot cd compatibility
18:20 < rxr> http://t2-project.org/packages/elftoaout.html
18:29 < CIA-38> rene * r34586 /trunk/architecture/powerpc/boot/ (boot.in bootinfo.txt ofboot.b): * improved powerpc boot cd compatibility, likewise
19:34 < rxr> dah, damn crap
19:34 < rxr> silo 1.4.14 has again no tarball release
19:34 < rxr> all those looney working for the "big distributions" should get a decent kick into the ass for not properly releasing vanilla tarballs ... !!!
19:42 < CIA-38> rene * r34587 /trunk/package/sparc/silo/silo.desc: * updated silo (1.4.13 -> 1.4.14), some git extracted snapshot in absense of a propper tarball, yuck.
20:03 < capey> hmm... strange, attr fails of missing ar and ranlib
20:03 < capey> arch prefixed that is
20:04 < CIA-38> rene * r34588 /trunk/package/sparc/silo/ (5 files): * updated silo (1.4.14 -> 1.4.14_git)
20:04 < capey> maybe some small binutils problems
20:04 < capey> but i'm off for today, have to test more tomorrow
20:05 < rxr> cu
20:05 < capey> cu
20:05 -!- capey [n=mika@e81-197-72-70.elisa-laajakaista.fi] has quit ["sleep_mode"]
20:07 < CIA-38> rene * r34589 /trunk/package/sparc/silo/register-pseudoop.patch: * removed obsolete silo patch
20:10 < CIA-38> rene * r34590 /trunk/package/sparc/silo/ (be-crosscompilable.patch cross.patch): * renamed silo cross patch for readability and consistency
20:12 -!- mpp_ [n=mpp@2a01:48:251:107:203:dff:fe13:6c9e] has quit ["good night - good fight"]
20:20 < CIA-38> rene * r34591 /trunk/package/gnome2/gtk+/gtk+.desc: * updated gtk+ (2.18.2 -> 2.18.3)
20:20 < CIA-38> rene * r34592 /trunk/package/gnome2/gnome-keyring/gnome-keyring.desc: * updated gnome-keyring (2.28.0 -> 2.28.1)
20:20 < CIA-38> rene * r34594 /trunk/package/xorg/libx11/libx11.desc: * updated libx11 (1.3 -> 1.3.1)
20:20 < CIA-38> rene * r34593 /trunk/package/xorg/xtrans/xtrans.desc: * updated xtrans (1.2.4 -> 1.2.5)
20:20 < CIA-38> rene * r34595 /trunk/package/xorg/libxaw/libxaw.desc: * updated libxaw (1.0.6 -> 1.0.7)
20:23 < CIA-38> rene * r34596 /branches/7.0/misc/luabash/bash/luabash.c: * mreged r33441 from trunk: * added BASH_VERSION 4.x to the known luabash versions
20:42 -!- LMJ [n=serwou@laf31-4-82-236-42-164.fbx.proxad.net] has joined #t2
21:32 -!- dpc [n=dpc@chello089075192131.chello.pl] has joined #t2
22:20 -!- dpc [n=dpc@chello089075192131.chello.pl] has left #t2 []
23:03 -!- LMJ [n=serwou@laf31-4-82-236-42-164.fbx.proxad.net] has quit [Read error: 110 (Connection timed out)]
--- Log closed Mon Oct 19 00:00:43 2009