WebHosting Paid by #1Payday.Loans
[00:15] <ringo> hi madtux ! [00:16] <ringo> sorry for the late response ... [00:44] blindcod1r (n=blindcod@tor/session/x-5d9e2ad9ad1bb0eb) joined #rocklinux. [00:44] blindcod1r (n=blindcod@tor/session/x-5d9e2ad9ad1bb0eb) left irc: K-lined [00:44] blindcoder (n=blindcod@tor/session/x-fcb6acb75dc11fba) left irc: Remote closed the connection [00:50] _BoS_ (n=BoS@dslb-088-072-036-018.pools.arcor-ip.net) left irc: Remote closed the connection [00:50] _BoS_ (n=BoS@dslb-088-072-033-066.pools.arcor-ip.net) joined #rocklinux. [01:13] kasc_ (n=kasc@dslb-084-060-107-099.pools.arcor-ip.net) joined #rocklinux. [01:20] kasc (n=kasc@dslb-084-060-105-114.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out) [01:20] Nick change: kasc_ -> kasc [01:21] ringo (n=info@ip51cf58c9.direct-adsl.nl) left irc: [01:26] duck_chris (i=manolo@pf0.hostarica.com) joined #rocklinux. [01:26] <madtux> Welcome duck_chris [01:26] <duck_chris> :-D [05:04] duck_chris (i=manolo@pf0.hostarica.com) left #rocklinux. [05:14] madtux (i=miguel@pf0.hostarica.com) left irc: "Leaving" [06:39] blindcoder (n=blindcod@tor/session/x-ecb9fb12e8ef46b6) joined #rocklinux. [06:40] <blindcoder> moin [07:28] <_BoS_> moin [08:22] <netrunner> moin [08:25] <[raphael]> moin [08:28] <blindcoder> >_< [08:29] Action: blindcoder just read linux/Documentation/filesystems/ramfs*initramfs.txt [08:29] <blindcoder> is it really just me or does Linux do quantum leaps _backwards_ in usability? [09:36] <mnemoc> *ACK* [09:36] <blindcoder> thanks [09:37] <blindcoder> I was already thinking that I'm going nuts [09:37] <mnemoc> :) [09:37] <mnemoc> they are [09:57] ija_ (n=ija@62.242.80.212.versanetonline.de) joined #rocklinux. [10:02] ija (n=ija@212.80.242.152) left irc: Nick collision from services. [10:02] Nick change: ija_ -> ija [10:12] <th> good that i've decided to do a rebuild-stage... [10:13] <th> kdepim fails if the etc/env directory already exists. [10:13] <th> hotfixing this with giving "-p" to mkdir [10:13] <th> 633 builds total, 618 completed fine, 0 with errors. [10:23] <netrunner> crystal? [10:24] <th> yes [10:24] <th> with current journal [10:24] <th> and two hotfixes [10:24] <th> the one i just mentioned [10:24] <th> and one caused by devfs removal (alsa tried to put something in etc/devfs.d) [10:27] <owl> moin [10:32] raphael_wels (n=raphael@lehrenetz.fh-wels.at) joined #rocklinux. [10:50] [raphael] (n=raphael@raphael.netpark.at) left irc: "using sirc version 2.211+KSIRC/1.3.12" [10:50] Nick change: raphael_wels -> [raphael] [12:18] <netrunner> why does this not work? https://www.heise.de/tp/r4/artikel/22/22296/1.html [12:18] <netrunner> argl. [12:19] <netrunner> I mean var_remove confopt " " "--enable-libpam" [12:19] <netrunner> aah, it is in extraconfopt. [12:41] [raphael] (n=raphael@lehrenetz.fh-wels.at) left irc: "using sirc version 2.211+KSIRC/1.3.12" [13:02] <clifford> daja77_: https://www.rocklinux.net/smng_test/index.websplb?filter_status=O&filter_search=SECURITY [13:03] <daja77_> nice [13:03] <clifford> .. you are the sucurity team. maybe its time to use your priveligdes to apply this stuff and write security advisories? [13:03] <daja77_> do i have a login there? [13:04] <daja77_> just trying [13:04] <th> daja77_: it's your sm account [13:04] <clifford> yes - but it is still in testing and I'm re-importing the stuff from old sm regulary, flushing all changes done in smng_test [13:05] <daja77_> ok i logged in but i don't find an apply button [13:05] <clifford> but you can also go to https://www.rocklinux.net/submaster/smadm.cgi?a=list and search for 'SECURITY' with your browsers search function. [13:05] <clifford> you can't apply it there. [13:05] <clifford> You can apply it in subversion. [13:06] <clifford> You have write permissions to the repository [13:06] <daja77_> ah ok so i need a svn co too, checking out [13:12] <daja77_> th: sry i am a bit out of sync, i need to co with the username via https, right? [13:21] <th> daja77_: it's a webinterface [13:21] <th> daja77_: like the submaster one [13:21] <th> ahhh [13:21] <th> i'm distracted. [13:21] <th> ignore what i said - i did not read all [13:24] <daja77_> my svn just told me that it has no ssl support, that's why I am asking [13:35] <th> rebuild openssl+neon+subversion [13:36] <th> daja77_: if you want me to JUST APPLY a single smpatch [13:36] <th> daja77_: then you can just tell me the smid and i'll commit it blindly [13:36] <th> like in your name [13:36] <th> if that helps [13:37] <daja77_> well i need ssl in svn sooner or later anyway [14:04] <daja77_> hm shared files between neon and subversion [14:05] <th> where? [14:05] <daja77_> man pages [14:07] <th> which build? [14:07] <daja77_> using Build-Pkg on my host [14:07] <th> ahh ok [14:07] <th> just mine -r neomn [14:07] <th> just mine -r neon [14:07] <th> and subversion ;) [14:07] <th> Error logs from crystal-TRUNK-x86-pentium2-32-crystal-expert: [14:07] <th> 633 builds total, 633 completed fine, 0 with errors. [14:08] <daja77_> :) [14:09] <th> my journal + two minor hotfixes [14:09] <th> that is with kde in /usr [14:10] <th> so i expect some desktop-file issues as posted by stf on the mailing list [14:11] <th> iso-upload in progres [14:11] <daja77_> nice [14:11] <th> 2MBit only - will take some time. [14:19] <daja77_> co in progress [14:30] <th> New Revision: 7259 [14:30] <th> cool [14:30] <daja77_> ok i committed these patches, how can i flag them as applied? [14:30] <th> did you use svn directly? [14:30] <daja77_> yes [14:30] <daja77_> smap in svn [14:30] <th> then you can mark them in the sm webinterface [14:31] <daja77_> i have no button for it there [14:31] <th> smap in svn with the correct settings marks them as applied in sm automatically [14:31] <th> then you need to ask cliff to make you sm admin [14:31] <th> or i can mark them as applied [14:31] <th> lemme do that for the two patches [14:31] <daja77_> yes [14:32] <th> done [14:32] <daja77_> thx [14:34] <th> if you need this again. just pm me. my latency on irc is not that bad. [14:35] <daja77_> ok but i should bug clifford anyway ;) [14:36] <th> yea [14:36] <th> but [14:36] <th> well - perhaps it's ok to wait for smng [14:37] <daja77_> np [14:38] <th> ahh bbsboot is pretty big - i'm afraid it includes tar's and gems' [14:39] <th> there is no sense in having them both on cd [14:39] <th> gems on cd and tar's alone [14:39] <daja77_> indeed [14:39] <th> can i make Create-ISO not to include the tar's? [14:40] <daja77_> dunno, never had both types of packages [14:42] <clifford> daja77_: according to our ''P'' thingy (I'm not writing the word out here ;-) you are supposed to write a security advisory now. [14:43] <blindcoder> aaah, the evil word [14:43] <daja77_> clifford: ok and where to? [14:43] <th> rock-users for a start? [14:44] <netrunner> why is linux24-header splitted as :dev? [14:44] <daja77_> where can I read that P thingie btw [14:44] <clifford> I'd say rock-users is ok. [14:45] <clifford> daja77_: I think we don't have a written version of the current P thingie. [14:45] <daja77_> oh ic [14:46] <clifford> Documentation/Developers/POLICY is from 2003 .. [14:46] <clifford> .. and pretty outdated. [14:47] <daja77_> so only to rock-user? [14:47] <clifford> I'd say for now it is most important that we do realease security advisories. [14:47] <clifford> Its not that important how we do it atm. [14:49] <blindcoder> bury them somewhere in the wiki *eg* [14:49] <daja77_> no [14:49] <daja77_> the question is if i should cross-post them to rock-dev [14:49] <daja77_> hm but i could create a wiki page linking to the advisory mails [14:50] <th> anyone knows how to create an ISO excluding tar.bz2 packages? [14:51] <clifford> th: move the *.tar.bz2 files to a different directory, e.g. [14:51] <th> hmm that would be too easy [14:53] <clifford> ok. then create an extra dir for the gems, create hardlinks for the gems in this extra dir and patch isofs.txt [14:54] <clifford> is that complicated enough? [14:54] <clifford> (that's what the crystal target is doing because not all gems from crystal should go to the cd image) [14:55] <th> sounds nice. [15:00] <daja77_> sent [15:05] Action: daja77_ waiting to be flamed [15:09] <netrunner> some structure would be nice. [15:11] <blindcoder> indeed [15:12] Action: daja77_ adopting the suse structure [15:13] <blindcoder> good idea [15:14] <daja77_> do we want to have announcement ids? [15:15] <daja77_> sth like RL-2006-03-21-01 ? [15:15] <daja77_> or perhaps shorter [15:15] <daja77_> RL-20060321-01 [15:15] <daja77_> would be nice as reference on the wiki too [15:16] <blindcoder> hmm [15:16] <blindcoder> RL-Crystal- ? [15:17] <daja77_> the id would just state that it is one of our announcements, it should be in the announcement which sub distros, targets or whatever are affected [15:18] <blindcoder> okay [15:24] <netrunner> I don't think there would be an announcement that only affects crystal and not generic [15:25] <netrunner> and all generic issues also affect crystal since you can use rock-src to build packages of generic not included in crystal [15:25] <th> 237M for a bbs bootable cd is still pretty big [15:29] <stf^rocklinux> moin [15:29] <stf^rocklinux> th: your last bbs iso was that size, too... [15:34] <stf^rocklinux> th: there's not much to take away from it, though [15:36] <th> stf^rocklinux: ok. perhaps i was only scared by the old size when i had the tars in by accident [15:37] <stf^rocklinux> th: there was a bug in my mkinitrd patch, making it unusable for cross-builds [15:37] <stf^rocklinux> th: I'll send a new patch for it today [15:37] Action: daja77_ checking the wiki for spam [15:38] <th> stf^rocklinux: please make sure that it will be in my next journal [15:38] <th> stf^rocklinux: i'll announce my journal here - as usual [15:39] <stf^rocklinux> th: ok. With it I could successfully create a initrd for QEMUs ARM emulation, and the kernel I build with ROCK boots as well :) [15:42] <stf^rocklinux> I've tried creating an initrd and bootable kernel for other QEMU system emulations too, but so far I've got only x86 to work ... [15:43] <stf^rocklinux> Booting x86 kernel and initrd on x86_64 also works ^^ [15:46] <th> crystal iso will be up in < 10minutes [15:54] #rocklinux: mode change '+o th' by ChanServ!ChanServ@services. [15:54] Topic changed on #rocklinux by th!n=th@montana.hbsn.de: https://iso.rocklinux.de/rock-ftp/official/test-releases/ +++ please test 20060321-crystal_cd1.iso now! [15:54] #rocklinux: mode change '-o th' by th!n=th@montana.hbsn.de [16:32] <th> bbs iso and tar packages is available as well now. [16:32] <th> url is in topic [16:37] <daja77_> template sent [16:43] daja77 (n=daja77@dslb-088-072-040-115.pools.arcor-ip.net) joined #rocklinux. [16:48] blindcoder (n=blindcod@tor/session/x-ecb9fb12e8ef46b6) left irc: Remote closed the connection [16:52] <daja77> >_< [16:55] daja77_ (n=daja77@dslb-088-072-040-115.pools.arcor-ip.net) left irc: Read error: 110 (Connection timed out) [16:58] <netrunner> I like the template, just the contained texts need some toughts. [16:58] <netrunner> thoughts [16:58] <netrunner> e.g. if we tell to sm sync, this means the patch is already in trunk. what if we know about a security issue, but the patch is still in submaster? [17:00] <daja77> netrunner: i'll commit security patches into trunk [17:01] <daja77> if there is an issue pending because of problems, the advisory would of course not state to do sm sync in that case [17:01] <netrunner> ah ok. [17:11] toberkel (n=toberkel@toberkel.net) left irc: Connection reset by peer [17:12] toberkel (n=toberkel@toberkel.net) joined #rocklinux. [17:47] blindcoder (n=blindcod@tor/session/x-79f4d51ed00c2c3e) joined #rocklinux. [17:51] <blindcoder> re [17:52] <stf^rocklinux> wb [17:55] <blindcoder> HAH! [17:55] <blindcoder> got my free harddisk space back [17:55] <blindcoder> fsck sometimes works miracles :) [17:56] <daja77> .oO (even when you don't write it with s) [17:56] Action: daja77 hides [17:57] <blindcoder> hehe [18:20] <th> harr [18:37] <netrunner> https://www.sandmonkey.org/2006/03/19/swastikas-are-trendy-in-cairo/ [18:48] <fake> moin [18:49] <blindcoder> moin fake [18:53] <fake> th: how's your journal ? [18:53] <fake> th: i will start a new build now [19:18] madtux (i=miguel@pf0.hostarica.com) joined #rocklinux. [19:38] <th> fake: if you would wait a little test... [19:38] <fake> th: np [19:39] <th> fake: how now is your now? [19:39] <fake> th: tonight [19:39] <th> fake: the iso (see $topic) just arrived here for a test-install [19:40] <fake> d/l running [19:40] <fake> ~40 min [20:19] <owl> hi fake [20:19] <fake> hi owl [20:19] <fake> damn! [20:19] <th> fake: perhaps you could try a journal with all these avm patches [20:19] <fake> how can i actually do anything with the return value of an asterisk 'application' ? [20:19] <th> or are they mostly public? [20:20] <fake> th: some of them affect personal pkgs from other people [20:20] <th> so we should give them a grace period of 14 days i think [20:21] <th> i think we should wait for current test results [20:21] <th> there might be some work regarding kde prefix change [20:22] <th> fake: perhaps you could try the gcc4 change? [20:22] <th> or start small with kernel update? ;) [20:23] <fake> th: both [20:23] <th> brave [20:23] <fake> th: that's nice, as a basic build will suffice [20:23] <th> basic==bbs? [20:23] <fake> noe, minimal [20:23] <th> or minimal? [20:23] <fake> -e [20:23] <th> that suffices for gcc4 change? [20:23] <th> i doubt that [20:24] <fake> in my opinion, yes *g* [20:24] <th> might cause regressions with core packages not in minimal [20:24] <fake> let's put it this way: if not even a minimal system works, we can skip the big builds [20:26] <th> yea - a successfull minimal build is required. but not sufficient as a journal test [20:26] <fake> i agree. [20:36] <th> current crystal has qt33 AND qt40 [20:36] <th> do we want this? [20:37] <blindcoder> dbus needs qt40 [20:38] <th> ahh i see [20:39] <th> i hope it does not need it for the user interface? [20:41] <blindcoder> no idea, netrunner knows about it [20:44] <th> can't create locak file /etc/mtab~9615 - read only filesystem [20:44] <th> thats when i reboot after installing [20:44] <th> i guess some umount needs the -n flag [20:48] <daja77> hm no comments [20:49] <blindcoder> daja77: be patient, they'll come in like water into the desert [20:49] <blindcoder> rarely, but steadily :) [20:49] <daja77> :) [20:49] <stf^rocklinux> blindcoder, th: dbus uses qt40 by default, but can use qt33 as well afaik [20:49] <daja77> just like to do the second announcement this way [20:50] <blindcoder> daja77: what do you mean with 'cross references'? [20:50] <th> hmmmmm [20:50] <th> the shipped initrd.img still does not work [20:51] <th> first it complains /sbin/init: line {63,68} {/etc/conf/kernel,/dev/fd/62}: No such file or directory [20:51] <th> then about missing /etc/dm/mounts [20:51] <blindcoder> humm [20:51] <th> then it tries to mount the root (from what it found in the fstab) [20:51] <th> and fails to find that device [20:51] <stf^rocklinux> th: my mkinitrd patch is buggy [20:51] <blindcoder> I thought I fixed that latter one [20:51] <th> but the "dumping into emergency shell" just works [20:51] <blindcoder> :) [20:52] <stf^rocklinux> th: but i've fixed it in the meantime [20:52] <th> and i get a broken pipe for the find /sys -name uevent ... [20:52] <blindcoder> th: that's why it's an "emergency" :) [20:52] <daja77> blindcoder: this is for security ids from other organizations, usually the ones who discovered the bug [20:52] Action: blindcoder wonders why suddenly everyone hacks around in the initrd... [20:52] <th> blindcoder: it did not work before because of missing libdl [20:52] <blindcoder> daja77: ah, okay [20:52] <daja77> like [20:52] <daja77> Cross-References: CVE-2006-0745 [20:53] <blindcoder> th: then just call it post installation and it will... I still don't see a problem [20:53] <blindcoder> daja77: I see [20:53] <daja77> just copied that from suse [20:53] <th> blindcoder: it is a problem as long as one has to call mkinitrd manually in post installation [20:53] <blindcoder> th: postinstall hook? [20:53] <th> i just followed the instructions [20:53] <blindcoder> th: we already do this for ssh [20:54] <th> blindcoder: could you add a patch? [20:54] <blindcoder> th: to create the sshd keys [20:54] <blindcoder> th: I can't test it [20:54] <blindcoder> at lesat not in a reasonable timeframe [20:55] <blindcoder> th: besides, didn't fake create a postinstall script already for rockinitrd? [20:55] <stf^rocklinux> th: if you used my mkinitrd patch then the resulting initrd will be incomplete [20:56] <th> stf^rocklinux: so did i? [20:56] <stf^rocklinux> th: I thought you said so [20:56] <stf^rocklinux> th: when you started the build [20:57] <th> https://www.rocklinux.net/people/teha/2006-03-19-journal.txt [20:57] <th> what would i need to rebuild, to create a new iso? [20:59] <stf^rocklinux> rockinitrd should be enough if mkinitrd is called in stone setup [20:59] <blindcoder> th: https://www.rocklinux.net/submaster/data//2006/03/1618512808515.patch [20:59] <blindcoder> th: that one should fix the libdl problem [20:59] <th> blindcoder: yea [20:59] <blindcoder> th: 2006031618512808515 [20:59] <th> blindcoder: i have that. and the libdl problem is gone [21:00] <blindcoder> th: good, problem solved :) [21:00] <th> blindcoder: no. [21:00] <th> blindcoder: you dont listen. [21:00] <blindcoder> I do listen [21:00] <blindcoder> I just don't understand [21:00] <th> blindcoder: it's not the libdl problem any longer [21:00] <stf^rocklinux> th: I think your initrd misses the files from /lib/udev/devices ... [21:00] <th> i do have a bash [21:00] <th> stf^rocklinux: that sounds reasonable [21:00] <th> stf^rocklinux: /dev/fd... [21:01] <th> /lib/udev/devices/ only has the shm and pts dirs [21:02] <stf^rocklinux> th: the other files were not copied because they are special files (symlinks, etc.) [21:02] <th> oh - how smart is that? [21:02] <stf^rocklinux> I've fixed mkinitrd here already [21:02] <th> oh ok [21:02] <th> that was introduced by you patch, right? [21:02] <stf^rocklinux> also makes the .empty file hack obsolete [21:03] <stf^rocklinux> th: I think so [21:03] <th> ok - so when is that initrd built? [21:04] <blindcoder> with fakes patch, post installation [21:04] <stf^rocklinux> th: you mean linux26: boot/initrd.img? [21:04] <stf^rocklinux> th: somewhere in linux.conf... [21:04] <th> how can it be linux26, if it contains our /sbin/init? [21:05] <th> ahh ok [21:05] <blindcoder> linux26 creates an boot/initrd.img compile time [21:05] <blindcoder> and the postinstall script overwrites it runtime [21:06] <th> so my current initrd (see journal) is an overwritten one? [21:06] <blindcoder> this is so that a known-to-work initrd with current configuration is always present for new kernel version [21:06] <th> or are you talking of another patch, that's not in yet? [21:06] <th> (in as in "in my journal") [21:07] <blindcoder> with fakes patch, yes, it will be overwritten postinstall [21:07] <blindcoder> yes, it's already in [21:07] <blindcoder> 2006031618512808515 [21:09] <th> blindcoder: well so i have a postinstall version. and it's still broken. so that's because of stf's patch and has nothing to do with fake's patch. [21:09] <blindcoder> probably. I haven't had time to study each initrd patch closely [21:09] <blindcoder> the postinstall version fixes the libdl thing [21:10] <blindcoder> and at least the one I have running still creates the /dev/fd symlink [21:10] <blindcoder> the one in trunk doesn't anymore [21:10] <blindcoder> and relies on /lib/udev/devices/fd to be present [21:10] <blindcoder> which it seems not to be [21:13] <th> if we do it postinstall - so we can remove all mkinitrd relevant stuff from linux.conf? [21:14] <blindcoder> no [21:14] <blindcoder> we still need a file to look at to decide if we need to recreate a new initrd or not [21:14] <blindcoder> currently that's boot/initrd.img [21:15] <blindcoder> which comes from linux26 [21:16] <stf^rocklinux> System.map and /lib/modules would be good triggers to [21:17] <blindcoder> suggest it to fake [21:17] <blindcoder> it's his patch after all [21:18] <blindcoder> lib modules probably won't work [21:18] <blindcoder> and I'm unfamiliar with system.map [21:18] <blindcoder> it's just a symlink isn't it? [21:19] <stf^rocklinux> yes [21:20] <blindcoder> sounds good [21:20] <blindcoder> even more so as initrd.img probably won't get overwritten anyway [21:20] <blindcoder> since its checksum will have changed [21:23] <th> ( i just want to reduce unused sleeping code ) [21:23] <blindcoder> how about getting stuff working first? [21:24] <th> i'd prefer getting stuff working first. [21:24] <th> i was only asking that question to clarify things and give us a reminder for later [21:24] <blindcoder> good :) [21:25] <th> can i recreate the initrd from the emergency shell? [21:25] <blindcoder> hardly [21:26] <blindcoder> hmm [21:26] <blindcoder> no [21:26] <blindcoder> you should now have a bash with pid 1 [21:27] <th> ack [21:27] <th> i do [21:27] <th> well ok - i can reboot with install cd again [21:27] <blindcoder> you can just boot without initrd [21:28] <th> hmmmmmmm [21:29] <th> i think i'll get initrd working first. [21:29] <blindcoder> well, I don't know who removed the creation of the /data/fd symlink out of initrd [21:29] <blindcoder> ask that person to fix it [21:30] <th> we dont need it if we get it from /lib/udev/device [21:30] <th> s [21:30] <th> we dont need two solutions for one mechanism [21:31] <blindcoder> well, obviously we don't get it from there for one reason or another [21:32] <th> yea that's because stf's patch was unfinished, if i understood that right [21:32] <th> stf? [21:33] <stf^rocklinux> th: that's right [21:34] <th> ahh and current `mkinitrd` does fail with syntax error [21:34] <th> ok - i think i need your fix now, stf [21:35] <stf^rocklinux> th: mom [21:36] <th> done < <( mount ) | sort | uniq >/etc/conf/initrd/initrd_fsck [21:36] <th> that's a simple syntax error and can be resolved by s/< // ? [21:37] <blindcoder> no [21:37] <blindcoder> that's not a syntax error [21:37] <blindcoder> which line? [21:38] <th> 78 [21:38] <blindcoder> and file? [21:38] <th> /sbin/mkinitrd [21:38] <blindcoder> [ -f $file ] || continue [21:38] <blindcoder> that's line 78 here [21:38] <th> well you need to apply stf's broken patch i think [21:39] <blindcoder> hmm [21:39] <blindcoder> at least it's no syntax error [21:39] <blindcoder> that's valid syntax [21:39] <th> but my bash tells me: checking necessary fsck programs ... //sbin/mkinitrd: line 78: syntax error near unexpected token `<' [21:39] <blindcoder> done < <( mount ) | sort | uniq >/etc/conf/initrd/initrd_fsck [21:39] <blindcoder> line 49 here [21:39] <stf^rocklinux> th: https://www.rocklinux.net/submaster/smadm.cgi?i=2006032121380929687 [21:40] <blindcoder> th: that while loop works if I run it on an interactive shell by copy and paste [21:41] <th> stf^rocklinux: so what was the issue with that while .. <( mount) ? [21:42] <th> -. [ -e "/sbin/fsck.${fs}" ] && echo "/sbin/fsck.${fs} /sbin/fsck.${fs}" [21:42] <th> +. [ -e "${rootdir}/sbin/fsck.${fs}" ] && echo "/sbin/fsck.${fs} /sbin/fsck.${fs}" [21:42] <th> hmm [21:42] <stf^rocklinux> th: that while loop is one of the few lines that I haven't really changed [21:42] <th> yea [21:43] <th> i'd love to understand why i get a syntax error here [21:43] <stf^rocklinux> th: $rootdir="" unless you explicitly set it... [21:43] <stf^rocklinux> what error message? [21:43] <th> 21:42:00 < th> but my bash tells me: checking necessary fsck programs ... //sbin/mkinitrd: line 78: [21:43] <th> syntax error near unexpected token `<' [21:44] <blindcoder> th: do you by any chance run the mkinitrd with "sh mkinitrd"? [21:44] <th> the $rootdir="" should do no harm [21:45] <th> DAMN [21:45] <th> the first line is broken [21:45] <th> first line reads: [21:45] <blindcoder> or is your /bin/bash no real bash? [21:45] <th> # #!/bin/bash [21:45] <blindcoder> well, yes [21:45] <blindcoder> of course [21:45] <blindcoder> what else? [21:45] <th> but my shell is a real bash [21:45] <th> so it should be bash parsed [21:45] <stf^rocklinux> blindcoder: #!/bin/bash [21:46] <blindcoder> #!/bin/bash [21:46] <blindcoder> that's what it reads here [21:46] <th> but the problem vanishes if i fix the first line! [21:46] <th> the first line was broken [21:46] <stf^rocklinux> lol [21:46] <blindcoder> th: but why? [21:46] <th> now this gets interesting [21:46] <th> good question [21:46] <blindcoder> it isn't here if I apply stfs patch [21:47] <th> it's broken as well in my build/crystal-TRUNK-x86-pentium2-32-crystal-expert directory on the build machine [21:47] <th> let's do a big `svn diff` [21:47] <stf^rocklinux> blindcoder: I've fixed the first line in my latest patch, btw. my old patch is still in SM [21:47] <blindcoder> stf^rocklinux: id? [21:47] <stf^rocklinux> while read dev a mnt b fs c ; do [21:47] <stf^rocklinux> [ -e "${rootdir}/sbin/fsck.${fs}" ] && echo "/sbin/fsck.${fs} /sbin/fsck.${fs}" [21:47] <stf^rocklinux> done < <( mount ) | sort | uniq >/etc/conf/initrd/initrd_fsck [21:47] <th> stf^rocklinux: you KNOW that "first-line-bug"? [21:48] User646 (n=User@ip72-197-13-134.sd.sd.cox.net) joined #rocklinux. [21:48] <stf^rocklinux> th: I though it's not a problem, just stumbled over it... [21:48] User646 (n=User@ip72-197-13-134.sd.sd.cox.net) left irc: Client Quit [21:48] <stf^rocklinux> th: https://www.rocklinux.net/submaster/smadm.cgi?i=2006032121380929687 [21:48] <stf^rocklinux> id: 2006032121380929687 [21:49] <blindcoder> 2006031919444902232 [21:49] <blindcoder> there it's in :) [21:50] <blindcoder> hmm [21:50] <blindcoder> stf^rocklinux: hope you don't mind me reworking half of that patch after it's applied? [21:50] <stf^rocklinux> blindcoder: silly typo ^^ [21:51] <stf^rocklinux> blindcoder: do as you like :) [21:51] <stf^rocklinux> blindcoder: any special reason for reworking the patch? [21:52] <blindcoder> modprobeopt=`echo $kernel | sed '/2.4/ { s,.*,-n,; q; }; s,.*,--show-depends,'` [21:52] <blindcoder> that's the first [21:54] <blindcoder> functions defined mid-script is the next [21:54] <th> https://www.rocklinux.net/people/teha/2006-03-19-journal.patch [21:54] <th> that's the current build [21:54] <th> agains r7267 [21:54] <th> stop [21:55] <th> against r7257 [21:55] <blindcoder> for f in `find $file` ; do [21:55] <blindcoder> what the? [21:56] <stf^rocklinux> blindcoder: this handles files and directories in one loop ;) [21:56] <th> stf^rocklinux: was that the wrong patch id? where is the fix for the 1st line in 2006032121380929687? [21:56] <stf^rocklinux> btw. empty directories are copied now [21:56] <blindcoder> stf^rocklinux: what about files with spaces in them? [21:57] <stf^rocklinux> th: no, I introduced that bug in my first patch [21:57] Action: netrunner becoming ill :( [21:57] <th> stf^rocklinux: your first patch being 2006032121380929687? [21:57] <th> -#!/bin/bash [21:57] <th> +# #!/bin/bash [21:57] <th> that's 2006031919444902232 [21:58] <th> ok - i just hotfix the first line [21:58] <th> then i fix it in my test-install [21:58] <th> and call mkinitrd after install [21:58] <stf^rocklinux> blindcoder: still won't work (since we read from a whitespace seperated list) [21:58] <th> that's what needed to be done in the last install as well [21:58] <th> so it's only better [21:59] <blindcoder> stf^rocklinux: also, what's the difference between initrd_base and initrd_essential? [21:59] <blindcoder> stf^rocklinux: example: [21:59] <blindcoder> /etc/conf/initrd/initrd_foo contains: [21:59] <blindcoder> /foo/ /foo/ [21:59] <blindcoder> /foo/ contains: [21:59] <blindcoder> "/foo/abc def" [22:00] <blindcoder> now, that for f in `find $file` will barf about not finding /foo/abc and not finding def [22:00] <stf^rocklinux> th: first line bug was introduced in 2006031919444902232, so simply use 2006032121380929687 instead [22:00] <stf^rocklinux> blindcoder: I see... [22:00] <blindcoder> will become even funnier if "abc def" is a directory [22:00] <blindcoder> stf^rocklinux: that's why I'm gonna clean it up a bit :) [22:01] <stf^rocklinux> blindcoder: the problem with the approach find $file | while read f ; is that it's a subshell with it's own local $lib variable o_O [22:01] <th> stf^rocklinux: so 2006032121380929687 is an alternative to 2006031919444902232? and not an incremental? [22:01] <stf^rocklinux> th: yes [22:01] <blindcoder> while read f ; do ... ; done < <( find $file ) [22:01] <blindcoder> that's the solution [22:02] <th> diff -Naur 2006031919444902232 2006032121380929687|wc [22:02] <th> 419 2039 14103 [22:02] <th> that's a big incremental [22:02] <th> shall i cancel that whole journal because of it? [22:02] <blindcoder> hell, even this might work: [22:02] <blindcoder> OLDIFS=${IFS} ; IFS='\n' ; for f in `find $file` ; do ... ; done ; IFS=${OLDIFS} ; unset OLDIFS [22:04] <th> could you two work out an increment to 2006031919444902232? [22:04] <blindcoder> also, I'm quite sure that for f in `find $file` will choke on a line > 65535 characters [22:04] <th> one which you both agree to? [22:04] <stf^rocklinux> blindcoder: yes [22:04] <blindcoder> th: use his newer patch [22:04] <blindcoder> th: I'll clean it up _after_ it's been applied [22:04] <blindcoder> I don't like fiddling around in half applied patches [22:04] <blindcoder> it usually becomes a mess [22:04] <stf^rocklinux> th: revert the old, apply the new? [22:05] <th> that would mean a full journal for nothing [22:05] <th> there are some kde ongoings in it [22:05] <th> so i'd prefer to apply the first [22:05] <blindcoder> huh? [22:05] <th> first the 1st-line thing fixed [22:05] <blindcoder> discard the first, apply the new one [22:05] <th> i'm currently testing such a build [22:05] <blindcoder> what's the problem? [22:05] <blindcoder> that patch only changes rockinitrd files [22:06] <th> i can't apply the new one without doing a full build [22:06] <blindcoder> interrupt it, patch around, pick it up again [22:06] <th> there is nothing to interrupt [22:06] <th> the iso's are online [22:06] <blindcoder> apparently I'm too simple minded [22:06] <stf^rocklinux> th: I'll create an incremental patch [22:07] <blindcoder> rsync? [22:07] <th> now it's the question wether to apply the journal (with some hotfixes perhaps) or drop it. [22:07] <th> rsync? [22:07] <th> atd? [22:07] <th> named? [22:08] <blindcoder> th: revert the old patch, apply the new one, create-errlist -newdelete, build-target, create-iso, rsync [22:08] <th> ahh [22:08] <th> ok i'll try [22:08] <th> with that incremental patch [22:09] <th> so it's still a hotfix [22:09] <th> stf^rocklinux: ok waiting for it. [22:09] <blindcoder> apparently I'm really too simple minded [22:09] <th> blindcoder: why? [22:10] <blindcoder> and with thta simple mind I'm now going to bed [22:10] <blindcoder> good night! [22:10] <th> good night [22:13] <stf^rocklinux> blindcoder: gn8 [22:13] <stf^rocklinux> th: 2006_03_2122130112045 [22:16] <fake> puh [22:16] <fake> how long was i away? you're talking too much ;) [22:16] <th> stf^rocklinux: that would not apply to my current journal [22:17] <th> patching file package/base/rockinitrd/mkinitrd.sh [22:17] <th> Hunk #1 FAILED at 1. [22:17] <th> Hunk #4 FAILED at 96. [22:17] <th> (dry-run) [22:17] <th> hmmmmm [22:17] <th> perhaps some other patch i have [22:18] <th> that's a mess [22:18] <fake> .iso is here, let me fix one thing with the asterisk, then i'll install it on my workstation hete [22:18] <stf^rocklinux> th: I applied my patches against svn trunk [22:19] <th> yea and i applied the whole journal before [22:19] <fake> th: fixdiff ed <sm patch> package/base/rockinitrd/mkinitrd.sh [22:20] <th> bash: fixdiff: command not found [22:20] <fake> rocket emerge submaster [22:21] <th> /usr/sbin/rocket: line 195: cd: /usr/src/rock-src: No such file or directory [22:22] <th> that's a bbs ;) [22:22] <th> fake: can you give me the fixdiff only? [22:23] <th> i've no trunk install anywhere yet [22:23] <th> it's all stable ;-) [22:23] <fake> https://www.rocklinux.net/svn/submaster/trunk/fixdiff.sh [22:23] <stf^rocklinux> th: I've diffed my patch against your journal: 2006_03_2122231812651 [22:24] <th> stf^rocklinux: cool [22:24] <stf^rocklinux> th: hope I didn't mix it all up ^^ [22:25] <th> that confuses me [22:26] <th> how can this be against my journal if it patches an existing package/base/rockinitrd/initrd_essential.txt which does not exist in my journal? [22:29] <stf^rocklinux> th: grml... a left-over file ... [22:32] <stf^rocklinux> th: 2006032122314413122 [22:35] <stf^rocklinux> th: PLEASE say it applies :) [22:36] <th> stf^rocklinux: https://www.rocklinux.net/people/teha/mkinitrd.sh [22:37] <th> stf^rocklinux: 2006032122314413122 still patches my non-existing initrd_essential.txt [22:37] <th> ah yes [22:37] <th> it's adding it [22:37] <th> disregard that [22:38] duck_chris (n=root@pf0.hostarica.com) joined #rocklinux. [22:38] <stf^rocklinux> th: you have fixed the first line manually? [22:39] <th> ah damn [22:39] <th> is that the only reject? [22:40] <th> yes it is [22:40] <th> then i undo that s/# #/#/ hotfix and it should apply [22:40] <th> patching file package/base/rockinitrd/initrd_essential.txt [22:40] <th> patching file package/base/rockinitrd/mkinitrd.sh [22:40] <th> patching file package/base/rockinitrd/rockinitrd.conf [22:40] <th> patching file package/base/rockinitrd/rockinitrd.desc [22:40] <th> applies now. [22:41] <stf^rocklinux> th: thanks :) [22:41] <th> sorry for troubling you [22:42] <stf^rocklinux> np. juggling with patch files is a delicate business ^^ [22:42] <th> Scheduling package 'rockinitrd' for rebuild ... [22:43] <th> fake: you might want to delay the install [22:43] <fake> th: y? [22:43] <th> hotfixing the iso [22:43] <fake> aye [22:46] <th> i'll try to supply xdelta [22:49] <th> xdelta needs glib? [22:49] <stf^rocklinux> th: yes glib12 afaik [22:51] <th> ahh ok [22:51] <th> built. [22:52] <th> 30568 root 18 0 359m 356m 237m R 93.0 40.3 0:53.18 xdelta [22:53] <th> rsync is faster [22:53] <th> rsync is already transferring the incrementals [22:54] <th> sent 1562651 bytes received 182162 bytes 14479.78 bytes/sec [22:54] <th> total size is 676655104 speedup is 387.81 [22:54] <th> nice speedup [22:55] <th> damn you can't do rsync if you have http access only [22:55] <th> well - perhaps xdelta finishes somewhen [22:56] <th> it did [22:56] <th> hahahaha 561MB of xdelta [23:00] <daja77> re [23:00] <stf^rocklinux> wb [23:05] #rocklinux: mode change '+o th' by ChanServ!ChanServ@services. [23:05] Topic changed on #rocklinux by th!n=th@montana.hbsn.de: https://iso.rocklinux.de/rock-ftp/official/test-releases/ +++ please test 20060321-crystal_PL1_cd1.iso now! [23:05] #rocklinux: mode change '-o th' by th!n=th@montana.hbsn.de [23:16] <fake> so it seems a return value <> 0 makes the extension leave ... hrm. [23:19] <fake> duck_chris: you're logged in as root on your machine... [23:20] <th> fake: that might be an incorrect identd response from some bouncer [23:20] <fake> th: yes, maybe - or maybe not ;) [23:36] <th> stf^rocklinux: cooooooool [23:36] <th> stf^rocklinux: thanks for PL1 [23:37] <th> boots without manually recreating initrd now [23:37] <stf^rocklinux> th: nice :) [23:40] <stf^rocklinux> pfft, with current speed my download of the ISO will take 30 days ... [23:40] <th> stf^rocklinux: i hope you're downloading PL1? [23:40] <stf^rocklinux> th: yes [23:41] <th> stf^rocklinux: do you want me to ship it to you? [23:41] <fake> i would be faster [23:41] <stf^rocklinux> th: np, I think speed will increase overnight [23:41] <th> k [23:42] <th> stf^rocklinux: btw - you gave some kde-prefix==/usr reports on rock-devel, no? [23:42] <stf^rocklinux> yes, I wrote one [23:43] duck_chris (n=root@pf0.hostarica.com) left #rocklinux. [23:43] <stf^rocklinux> My running system is build with this patch [23:43] <th> i was not (yet) asking for a patch, was i? [23:44] <stf^rocklinux> th: no, but I used the kde prefix=/usr patch for the system I'm running [23:44] <stf^rocklinux> th: there are 2 minor problems with it, both have to do with .desktop files [23:45] <th> ahh ok [23:45] <th> ok [23:45] <th> yea - i just reread your mail [23:49] <th> wow even `chsh` works. [23:49] <th> never tried that thing before [23:49] <th> works as user [23:57] <th> oh [23:58] <th> we have no kernel sources in crystal? [00:00] --- Wed Mar 22 2006