13:00 < anibal> Meeting Starts
13:00 < joshk> alright
13:00 < kmuto> sure
13:01 < Bill_Jennings> ok
13:01 -!- svenl is now known as luther
13:01 < GyrosGeier> affirmative.
13:01 -!- djen [~djen@67.136.145.212] has joined #debian-boot
13:01 < luther> Please everyone renqme yourself
13:01 -!- Mithrandir [~tfheen@vawad.raw.no] has joined #debian-boot
13:01 < joshk> to?
13:01 -!- luther is now known as SvenLuther
13:02 -!- Loevborg [foobar@p508BBEA1.dip.t-dialin.net] has joined #debian-boot
13:02 -!- mellum [~hueffner@joseki.Informatik.Uni-Tuebingen.De] has joined #debian-boot
13:02 < mellum> hi
13:02 < anibal> Topics:
13:02 < anibal>     * Preparation of the beta release and any issues that might be blocking it
13:02 < anibal>     * When to string freeze
13:02 < anibal>     * Architecture status rollcall
13:02 < anibal>     * Whatever else that needs to be discussed
13:02 < SvenLuther> joshk something readable in irc logs.
13:02 -!- guus [guus@xar.sliepen.eu.org] has joined #debian-boot
13:02 -!- mentor [~matthew@mentor.registered] has joined #debian-boot
13:02 < joshk> I think this is good enough
13:02 < anibal> Who is logging?
13:02 < joshk> nw
13:02 < joshk> mw
13:02 < SvenLuther> please add 2.6 migration too
13:02 < joshk> er, me
13:02  * joshk curses this Sun keyboard
13:03 < anibal> joshk is logging, thanks
13:03 -!- Md [md@md.staff.freenode] has joined #debian-boot
13:03 -!- huhlig [~huhlig@68-189-119-139.ca.charter.com] has joined #debian-boot
13:03 -!- lool [lool@joule.via.ecp.fr] has joined #debian-boot
13:03 -!- Igloo [~igloo@i-194-106-34-144.freedom2surf.net] has joined #debian-boot
13:03 < anibal>     * 2.6 migration
13:03 -!- GyrosGeier is now known as SimonRichter
13:03 < anibal> Any other new items?
13:04 < joshk> huhm, what about d-i automation? i'm wondering how much people have tried to figure this out?
13:04 < Mithrandir> joshk: like, automated installations?
13:04 < joshk> yes
13:04 < joshk> completely automated
13:04 -!- lieb [~lieb@adsl-67-117-70-170.dsl.sntc01.pacbell.net] has joined #debian-boot
13:04 -!- xdsg [~xsdg@3ffe:2900:2002:f00d:20e:a6ff:fe8f:3b0] has joined #debian-boot
13:04 < joshk> configure the network, and let 'er rip
13:04 < Mithrandir> I've thought a lot about it, but haven't had any time to actually test it yet.
13:05 < joshk> you have code?
13:05 < SvenLuther> joshk i also have profesional interest in this stuff.
13:05 < Mithrandir> http db, nothing more.
13:05 < anibal>     * d-i automation
13:05 < joshk> SvenLuther: me too
13:05  * xdsg looks at the topic... is this a bad time to ask a question?
13:05 < joshk> if it's quick, maybe not
13:05 < anibal> let's start then
13:05 < joeyh> yes, we're in the midle of a meeting
13:05 < SvenLuther> joshk it may be a separate issue though, let's speak about it later ok ?
13:05 < joshk> heh ok
13:05 < joshk> yeah
13:06 < anibal>     * Preparation of the beta release and any issues that might be blocking it
13:06 < joeyh> so, it's come to my attention that it's been 2.5 months since beta4.
13:06 < SvenLuther> yes, aned it is still what the web site recomends.
13:06 < joshk> do you think what we have in sid_d-i represents something close to a release candidate?
13:07 < joeyh> I'd like to set a date for the next release, does the end of this month sound doable?
13:07 < joeyh> it seems pretty good to me
13:07 < joshk> yeh
13:07 < SvenLuther> that' 2-3 weeks from now.
13:07 < xdsg> I'm using the nightly dated 20040716; I'm at the GRUB install step and GRUB is sucking CPU, but apparently doing nothing.  I'm using expert26, and I did select XFS as the root partition format; Is this what the XFS warning was about, and if so, can I just kill grub-install, chroot /target, and install grub myself?
13:07 -!- Irssi: T-D: Removed: Meeting on Saturday 17 July at 20:00 UTC
13:07 -!- Irssi: T-D: Added: Meeting right now | Topic: Preparation of the beta release and any issues that might be blocking it
13:07 -!- SimonRichter changed the topic of #debian-boot to: Welcome to the Debian-Installer development channel -- http://www.debian.org/devel/debian-installer | For help mail debian-boot@lists.debian.org | Meeting right now | Topic: Preparation of the beta release and any issues that might be blocking it
13:07  * huhlig nods
13:07 < joshk> xdsg: please wait until the end of the meeting, i suppose
13:07 < SvenLuther> joeyh: i am _NOT_ interested in supporting 2.4 kernels during the sarge lifetime.
13:07 < xdsg> joshk: k
13:08 < joshk> me either, but it's the only thing viable on sparc32
13:08 < SimonRichter> APUS and Amiga seem to have issues with 2.6 as well.
13:09 -!- eugen [~eugen@news.univ.kiev.ua] has joined #debian-boot
13:09 < lieb> the end of this month, is that a new try at tc1?  we'd like to get ia64 boot + 2.6 done before final
13:09 < joeyh> end of month is 2 weeks, that's a rather tight schedule since doing the release will take at least 1 week
13:09 < SvenLuther> joshk apus is different, but there will be no upstream support of 2.4 in the generic powerpc kernel.
13:10 -!- AstinusUK [~ahowells@cyberhowells.demon.co.uk] has joined #debian-boot
13:10 < joshk> that's what happened with sparc, too - the maintainer just quit, possibly citing reasons of sanity :)
13:10 < AstinusUK> How can I change the kernel on an install CD so its compatible with my USB keyboard/
13:10 < joshk> sparc32, that is
13:10 < joshk> AstinusUK: sorry, we're in a meeting right now, short answer: it should be, long answer: please wait unti the end of the meeting
13:10 -!- radiojoe [~joe@c-24-20-187-242.client.comcast.net] has joined #debian-boot
13:10 < AstinusUK> Sorry about that.
13:11  * AstinusUK didn't know.
13:11 < SimonRichter> AstinusUK, no problem
13:11 < lieb> wrt ia64, we'd be more interested in 2.6 as well
13:12 < joshk> joeyh: well, assuming we slow down feature developments, and concentrate on bugfixes it will not take that long
13:12 < joshk> lieb: dannf has already uploaded ia64 2.6 debs/udebs already i think.
13:12 < joeyh> no, 1 week is the minumum time to get a d-i release out. Really
13:12 < joeyh> refer back to previous timelines
13:12 < SvenLuther> joeyh: i am interested in making 2.6 the default on powerpc, with fallback to 2.4 possible.
13:12 < lieb> joshk, some of them. we have 2.6 stuff for boot yet
13:12 < joshk> that's when you had to revert lots of broken changes
13:13 < joeyh> 1 week assuming everything goes peachy.
13:13 < joshk> mmkay
13:14 -!- dan2 [~beta3@beta3.registered] has joined #debian-boot
13:14 < dan2> did I miss something
13:14 < joshk> joeyh: the fact is sid_d-i seems to work fine as-is
13:14 < joeyh> well, on arches other than m68k perhaps
13:14 < joshk> anyway, i think we've exhausted this topic
13:14 < joeyh> maybe we should go through the arches
13:14 < joshk> yes
13:14 < joshk> let's go in real alphabetical order this time
13:14 < joshk> alpha?
13:15 < joshk> bueller?
13:15 < SvenLuther> joeyh: when do you plan to move the sid_d-i .udebs to sarge ?
13:15 < joshk> SvenLuther: there will be no more sid->sarge transitions, we'll be releasing from sid
13:15 < joshk> for the better anyway, because automatic testing handling for udebs isn't implemented anyway
13:15 < anibal> Before the arches:
13:15 < anibal>     * When to string freeze
13:16 -!- mephisto_bsd [mephisto_b@200-100-111-9.dial-up.telesp.net.br] has joined #debian-boot
13:16  * bdale notices the activity and raises this window...
13:16 < SvenLuther> joshk: This mean we need to change the netboot so it will always take the .udebs from sid.
13:16 < SvenLuther> even if you choose sarge as the .debs.
13:16 -!- Irssi: T-D: Removed: Topic: Preparation of the beta release and any issues that might be blocking it
13:16 -!- Irssi: T-D: Added: Topic: When to string freeze
13:16 -!- SimonRichter changed the topic of #debian-boot to: Welcome to the Debian-Installer development channel -- http://www.debian.org/devel/debian-installer | For help mail debian-boot@lists.debian.org | Meeting right now | Topic: When to string freeze
13:16 < vorlon> would be nice to have alpha on 2.6, but I'm simply not in a position to do this until there's some sign of the PCI fixes making it into the kernel-images.
13:16 < joeyh> it's a pity Christian Perrier is not here.
13:16 < joshk> SvenLuther: i guess that's debatable. i'm not really sure of the new scheme, joeyh is better prepared to talk about that
13:16 < mephisto_bsd> Hello! Somebody could help me with booting debian sarge installation on a Sparc Sun blade 1000?
13:16 < joshk> anibal: yeah, bubulle would know the most about when to freeze..
13:16 < joeyh> anyway, we have something like 25 languages at 100%
13:16 < anibal> he's on leave in spain
13:17 < SimonRichter> mephisto_bsd, after the d-i meeting, sure.
13:17 < joshk> hmm. request to move this meeting to #debian-boot-meeting?
13:17 < huhlig> joshk, I was wondering when someone was going to suggest that
13:17 < SvenLuther> joshk : current sarge powerpc 2.6 is not installable because the 2.6 .udebs are not in sarge. Since we want to install sarge ...
13:17 < Mithrandir> I could add a join message
13:17 < mephisto_bsd> SimonRichter: sorry! What time?
13:17 < vorlon> Is seppy here?
13:18 -!- anton [~anton@212.39.88.133] has joined #debian-boot
13:18 < joeyh> er, they _were_ at 100% earlier today, now they're all at 99%
13:18 < joeyh> anyway, we seem to be close
13:18 < anibal> I would suggest to stay here and maybe announce the next meeting at the new channel
13:18 < joeyh> 29 languages, which seems like enough to me. So do we need a string freeze?
13:18 < joeyh> anibal: works for me
13:18 < joshk> anibal: hmm ok
13:18 -!- mephisto_bsd [mephisto_b@200-100-111-9.dial-up.telesp.net.br] has left #debian-boot []
13:18 < SimonRichter> mephisto_bsd, I'd say anywhere from 30 minutes to an hour.
13:18 < SvenLuther> joeyh: what if we need to modify a string until the release ?
13:19  * joeyh suspects choose-mirror added a new country BTW, translators.
13:19 < vorlon> joeyh: I think we need a string freeze on general principle, so that people don't go changing strings *after* everybody thinks they've got it all translated.
13:19 < joeyh> vorlon: fair enough
13:19 < joeyh> we also have quite a lot of translated but un-uploaded strings
13:19 < joshk> we can just make a translation upload run like last time
13:19 < Mithrandir> entry-msg added, so we should be left a bit more alone now.  FYI.
13:19 < kmuto> I, as translator, am little tired to follow message changes
13:19 < joeyh> if we want to release by the end of the month, I'd suggest a string freeze starting in the next few days
13:20 < vorlon> agreed.
13:20 < SvenLuther> joeyh: what about next WE ?
13:20 < lieb> ia64 has partman-efi (to be checked) and manual additions in the pipeline
13:20 < CIA-3> debian-installer: pelle-guest * r17826 packages/netcfg/ (debian/changelog netcfg-common.c): Generate hotplug mapping entries for 16-bit PCMCIA cards as well
13:20 < joeyh> there are also maybe three or 4 languages not at 99% (100%) that could easily make it in a week
13:20 < joeyh> next weekend is a bit late
13:21 < SvenLuther> i have a minor nobootloader change.
13:21 < anibal> so string freeze early next week and beta release by the end of the month?
13:21 < joeyh> is there anyone who can coordinate a string freeze (contact translators, etc), while bubulle is out of town?
13:21 < joshk> when will he be back?
13:21 < joshk> i think seppy is qualified
13:21 < anibal> kmuto?
13:21 < kmuto> seppy?
13:21 < joshk> or bouz
13:21 < joeyh> yes, I think he suggested seppy
13:21 < SvenLuther> joeyh: 1 week til string freeze, 1 week til freeze, 1 week of the preparation of the release ?
13:21 < vorlon> joeyh: which languages "could easily make it"? (so they can be listed in the meeting notes)
13:21 < joshk> (denis barbier)
13:22 < joeyh> SvenLuther: your month seems to have one more week in it than mine
13:22 < joeyh> vorlon: bs, zh_TW, it, cy, nb
13:22 < SvenLuther> joeyh: if you release 1 week later, it is not the end of the world, is it ?
13:22 < joeyh> all well above 90%
13:23 < joeyh> do we really need a whole week to make template changes?
13:23 < joeyh> there have not been many lately
13:23 < vorlon> I think the string freeze should start Monday at the latest.
13:24 < joeyh> I was thinking more tuesday-ish, but somewhere in there
13:24  * SimonRichter suggests the night between monday and tuesday as a compromise
13:24 < anibal> don't you need to announce a string freeze and allow translators some time?
13:24 < SvenLuther> joeyh: tuesday 24:00 UTC time ?
13:24 < vorlon> anibal: the date of the string freeze is the date that developers stop uploading string changes.
13:24 < joeyh> anibal: well, they can catch up all during the freeze. Yes, it needs to be announced
13:25 < joeyh> I'd want to do final uploads of translations by next weekend
13:25 < SvenLuther> joeyh fine with me.
13:25 < SimonRichter> joeyh, "final" being final for the release candidate, I presume?
13:26 < joeyh> yes, final for the release, whatever it's called
13:26 < joeyh> something I've thought about doing is we could run a test candidate release concurrent with the string freeze
13:26 < anibal> joeyh: sat 24th of july?
13:26 < joeyh> anibal: yes
13:27 < joshk> oooh, my birthday :)
13:27 < joeyh> just to get a feel for where everything is
13:27 -!- kebil [kestas@ctv-213-164-127-69.init.lt] has joined #debian-boot
13:27 -!- lool [lool@joule.via.ecp.fr] has left #debian-boot []
13:27 < vorlon> sounds like a good idea to me.
13:27 < anibal> everybody agrees?
13:27 < SvenLuther> joeyh: fine with me
13:27 < seppy> kmuto: hey
13:27 < kmuto> seppy: hi
13:27 < joeyh> I made a images upload to the autobuilders 2 days ago. That's not been accepted yet, but it could be such a test candidate
13:27 < anibal> Let's move on then
13:28 < vorlon> We just tell testers "we know it's not 100% translated, don't report bugs".
13:28 < joeyh> right
13:28 < anibal>     * Architecture status rollcall
13:28 < joshk> alpha?
13:28 < SvenLuther> alpha first.
13:28 < anibal>    * Alpha
13:28 < joeyh> I guess it will depend on how quickly and well it builds, amoung other things
13:29 < vorlon> Right.  Alpha is in pretty solid shape; the two big outstanding issues for us are that automatic partitioning doesn't yet account for the space aboot needs, and that kernel images in the archive are still in need of a major PCI fix.
13:29 < joeyh> is that 2.6 kernels?
13:29 < vorlon> I'm working on the first of these this week, and maybe I'll get lucky and get aboot-installer working and uploaded in time for the freeze.
13:29 < joeyh> has the aboot thing at least been documented yet? It was mentioned last week
13:30 < vorlon> That's 2.4 kernels -- I haven't even touched 2.6 kernels in d-i yet, the last time I looked at them they had the same bug.  I don't know if the new kernel team has fixed it for 2.6 yet, though.
13:30 < joeyh> how bad is it?
13:30 -!- anton [~anton@212.39.88.133] has left #debian-boot []
13:31 < vorlon> I'd guesstimate it probably affects close to 1/4 of the alpha hardware I've heard from, including my own systems.
13:31 < vorlon> I was really hoping for a fix to the 2.4 kernel source package in June...
13:32 -!- Irssi: T-D: Removed: Topic: When to string freeze
13:32 -!- Irssi: T-D: Added: Topic: Architecture status rollcall
13:32 -!- SimonRichter changed the topic of #debian-boot to: Welcome to the Debian-Installer development channel -- http://www.debian.org/devel/debian-installer | For help mail debian-boot@lists.debian.org | Meeting right now | Topic: Architecture status rollcall
13:32 < SvenLuther> vorlon is the alpha kernel in the debian-kernel svn archive ?
13:32 < joeyh> it's the same problem with just lots of hardware not working that we've had for a long time
13:32 < vorlon> SvenLuther: yes.
13:32 < joeyh> right?
13:32 < vorlon> joeyh: right.
13:32 < joeyh> ok
13:32 < joeyh> well, it didn't stop us from releasing beta4, so I guess it shouldn't stop at least another beta
13:32 < joshk> vorlon: i can't find it there
13:32 < joeyh> sucks though :-(
13:32 < anibal> next arch:
13:33 < SvenLuther> vorlon but kernel 2.4 works on those ?
13:33 < vorlon> I consider it release critical for sarge; I wouldn't let it hold up the beta.
13:33 < vorlon> SvenLuther: no.
13:33 < SvenLuther> vorlon: kernel 2.2 ?
13:33 < anibal>     * amd64
13:33 < vorlon> But I have enough experience with 2.4 that I work around the breakage.
13:33 < vorlon> Why would I care about 2.2? :P
13:33 -!- anton_ [~anton@212.39.88.133] has joined #debian-boot
13:33 < joeyh> heh, the status is we have a huge flamewar in debian-vote
13:33 < Q_> amd64 currently doesn't have a netinst image.  I've been trying to get it working without much success.
13:34 < SvenLuther> vorlon: well, if we never supported those subarches, it is not RC.
13:34 < Q_> I've mail the list about it without a reply.
13:34 < joeyh> Q_ have you asked manty for help?
13:34 < vorlon> SvenLuther: we've moved on.
13:34 < joshk> Q_: you can probably get debian-cd from newly self-appointed debian-cd guru joeyh ;)
13:34 < joeyh> pfeh, not for netinsts
13:34 < Q_> Basicly I have no idea how to make them.
13:34 < SvenLuther> vorlon: anyway, this is mostly kernel work, and not d-i related right ?
13:35 < joeyh> you really need to talk to manty then, he's the guy who does
13:35 -!- eddyp [~eddy@82.77.146.149] has quit [Read error: 60 (Operation timed out)]
13:35 < Q_> We've also had some problems with changes in sid base.
13:35 < SvenLuther> does the d-i in sarge/sid actually is the one usable, or is CVS debian-cd still needed.
13:36 < vorlon> SvenLuther: it's kernel work that blocks d-i.  If the kernel team doesn't take care of it by the end of the month, I'm going to have to figure out NMU.  And these are not "subarchs", this is a Debian-specific kernel bug introduced in 2.4.2x that breaks handling of lots of PCI chipsets.
13:36 < joshk> i think CVS is the only thing worth using
13:36 < joeyh> Q_: http://debian-amd64.alioth.debian.org/install-images/sid-amd64-netinst.iso  <-- what's this?
13:36 < joeyh> SvenLuther: you need cvs
13:36 < Q_> joeyh: A very outdated version used with some custom script.
13:36 < joshk> joeyh: it was hand made by a crackheaded mrvn script :)
13:36 < kmuto> will we support amd64 for officially? or reference implementation only?
13:37  * huhlig would hope officially
13:37 < SvenLuther> vorlon: but fixing the kernel will fix d-i too:
13:37 < joshk> amd64 - the name is newly debatable, anyone familiar with intel's new Nocona offerings?
13:37 < vorlon> SvenLuther: correct.
13:37 < SvenLuther> vorlon: and you are welcome to the debian-kernel team :)
13:37 < Mithrandir> Q_: have you talked to maswan?  He was working on a bit on debian-cd on bach.
13:37 < anibal> So, what's the status of amd64 d-i?
13:37 < joshk> as we compile it, the amd64 distribution is fully compatible with IA32-E / EM64T (other names for nocona
13:37 < Mithrandir> joshk: the kernel images are so far called amd64-nocona, I think.
13:38 < SvenLuther> joeyh: why we odn't upload a working debian-cd to the archive ?
13:38 < joshk> i guess that's kind of awkward
13:38 < vorlon> SvenLuther: well, I was rather hoping that someone who was already getting up-to-speed on kernel maintenance would have taken care of it by now...
13:38 < joeyh> SvenLuther: good question.
13:38 < Q_> We have kernel images for "general", and specific for K8 and "xeon".
13:38 < joshk> yeah, but then amd64 is a misnomer
13:38 < Q_> xeon being the nocoda
13:38 < joshk> it's not important
13:38 < SvenLuther> vorlon: someone with the hardware needs to do the work.
13:39 < vorlon> I think working debian-cd packages would be good, but isn't there a fair amount of divergence between what manty actually runs, and what sits in CVS?
13:39 < joshk> did the ctte know about nocona when debatinng x86-64?
13:39 < vorlon> SvenLuther: I thought there already was a designated alpha kernel maintainer?
13:39 < SvenLuther> joshk: why ? its the official name.
13:39 < huhlig> joshk, you mean the one that intel refuses to admit exists
13:39 < joeyh> vorlon: not really, he has a wrapper around it, but the core cvs stuff is little changed
13:39 < fs> joshk: it was not released by then
13:39 < vorlon> SvenLuther: note that I've been told the PCI fix will be included in the next upload of kernel-source, and even *that* hasn't happened yet.
13:39 < joshk> huhlig: no, it's released, so now people do know
13:39 < fs> joshk: afaik, amd tried to agree with intel on a common name, but intel was not interested
13:40 < Q_> anibal: It seems to work for people while other say it failed without much detail of why it failed.  Some of those failures are explained by sid base breakage because we only have sid.
13:40 < joeyh> can we pick up the pace, I have an appointment in an hour..
13:40 < SvenLuther> vorlon is the patch in the svn ? i haven't seen it in the changelog, and hch is in vacations.
13:40 < joshk> ok, next arch then
13:40 < joshk> is that i386?
13:40 < vorlon> SvenLuther: I was told by wli that it was already in svn.
13:40 < anibal> could someone state the status of amd64 for me, please?
13:40 < fs> joshk: but this discussion is over, please don't beat a dead horse again
13:40 < vorlon> joshk: hppa
13:40 < joshk> ah
13:40 < huhlig> joshk, odd, I spoke with one of the intel reps yesterday when they came thru and they had "no plans to reimplement a 64 bit processor"
13:40 < vorlon> ... or arm, even
13:40 < joshk> oh, quite
13:40 < SvenLuther> vorlon: i will look at it tomorrow.
13:40 < vorlon> SvenLuther: thank you.
13:40  * huhlig hmms
13:40 < SvenLuther> vorlon: can you build a kernel out of svn and test it ?
13:40 < anibal>     * arm
13:40 < huhlig> good to know my reps dont know squat
13:41 < anibal> let's move on to arm
13:41 -!- YeTr2 [~rotoroote@wbar99.tmp1-4.16.208.26.tmp1.elnk.dsl.genuity.net] has joined #debian-boot
13:41 < joeyh> kyllikki: here?
13:41 < vorlon> SvenLuther: I can try.
13:41 < SvenLuther> vorlon ok, let's coordinate tomorrow about this.
13:41 < anibal> arm status?
13:42 < joeyh> going once, going twice..
13:42 < kyllikki> joeyh: hello
13:42 < joeyh> hey, you are here.
13:42  * kyllikki was nose in kernel guts
13:42 < joeyh> any arm changes since last week?
13:42 < anibal> it was reported in good shape last meeting
13:42 < kyllikki> not as far as I know and as no other bugger touches it :-)
13:43 < anibal> ok
13:43 -!- mephisto_bsd [mephisto_b@200-100-111-9.dial-up.telesp.net.br] has joined #debian-boot
13:43 < anibal>     * hppa
13:43 < anibal> bdale?
13:43 < seppy> I apologize for being that late. About the string-freeze, we have 32 langs good over 90% - quite nice. Am I right that you discussed it from thuesday 20th to sat. 24th? I think that's ok. And yes - it should be announced to translators, I can do that.
13:43 < bdale> I'm here, but I haven't tried d-i on hppa since my rc4 (?) reports
13:44 < joeyh> seppy: yes, correct and thank you
13:44 < joshk> urgh. ENOLUNCH.
13:44 < anibal> seppy: would you like to co-ordinate the string freeze?
13:44 < bdale> I can schedule some time Mon/Tue to test hppa remotely if needed
13:45 < joeyh> jbailey also has a hppa, and was hoping to test, but he's not here
13:46 < AstinusUK> I have a HPPA :)
13:46 < joshk> you're welcome to test then, and please file an installation report
13:46 < SvenLuther> AstinusUK: did you try d-i on it ?
13:46 < AstinusUK> SvenLuther: Planning to
13:46 < AstinusUK> Need to fix this kernel issue first
13:46 < AstinusUK> Since my HPPA only support USB KB and MOUSE
13:46 < AstinusUK> and 2.4.17 doesn't work with them
13:47 < SvenLuther> AstinusUK: what is the 2.6 status on it ?
13:47 < joshk> don't we use a newer kernel on hppa nowadays?
13:47 < SimonRichter> joshk, you need to suffer, like I do. I burned our tomato soup, and my flatmates have managed to talk me into a real good depression now.
13:47 < AstinusUK> JoshK: Yeah, but the jigdo files are so far out of date you can't download them
13:47 < joshk> AstinusUK: use images from...
13:47 < joshk> er
13:47 < anibal> AstinusUK: you may ask cjwatson for help
13:47 < AstinusUK> JoshK: Which leaves you with Woody 3.0r2 ISOs
13:47 < joshk> someone in X, paste him a gluck URL
13:48 < AstinusUK> My 3.0r2 isos have 2.4.17 kernel
13:48 < AstinusUK> which doesn't support USB KB :(
13:48 -!- mephisto_bsd [mephisto_b@200-100-111-9.dial-up.telesp.net.br] has quit []
13:48 < seppy> anibal: I'm sorry - what is there to coordinate? I can send a announce, yes.
13:48 < SvenLuther> AstinusUK: hppa should be able to netboot and thus use the netboot images, no ?
13:48 < AstinusUK> SvenLuther: Not sure, I'm on a 64k ISDN connection though
13:48 < AstinusUK> SvenLuther: Hence buying the ISOs and having them shipped
13:49 < anibal> seppy: names were asked to coordinate the string freeze before
13:49 < SvenLuther> AstinusUK: where do you live ?
13:49 < AstinusUK> SvenLuther: UK
13:49 < bdale> SvenLuther: yes, depends on which hppa as to which boot protocol, but I netboot routinely
13:49 < AstinusUK> HP C3600
13:49 < SvenLuther> bdale notice that the netboot image can also be launched from a disk.
13:49 < bdale> that one's tftp'able
13:50 < AstinusUK> Well I was asking how to replace a kernel on a install disk because I'm somewhat a newbie
13:50 < AstinusUK> Willing to be used as a guinea pig though :)
13:50 < joshk> AstinusUK: were you using a CD?
13:50 < bdale> SvenLuther: I know that, but I'm a couple of thousand miles from my hppa boxes right now...  ;-)
13:50 < AstinusUK> JoshK: I have the full Woody set for HPPA
13:50 < AstinusUK> JoshK: 3.0r2 + an update CD
13:50 < joshk> AstinusUK: try this
13:51 < joshk> http://gluck.debian.org/cdimage/testing/sid_d-i/hppa/current
13:51 < anibal> ok, move on
13:51 < AstinusUK> Can I get the netinst one?
13:51 < kmuto> well, time is limited
13:51 < AstinusUK> then use my CDs for .debs?
13:51 < anibal>     * i386
13:51  * AstinusUK nods and will wait.
13:51 < joshk> no, this is a new release
13:51 < joshk> yeah, hold on
13:51 < joeyh> we seem to be in good shape
13:52 < joshk> i agree
13:52 < kmuto> I tried i386 yesterday
13:52 < kmuto> good shape
13:52 < joshk> what's this whole firmware-modules stuff
13:52 < joeyh> pcmcia seems to work again (but per is making it better)
13:52 < anibal> ok, next arch
13:52 < joeyh> the 2.6.7 kernel is not in testing yet, and until it is, we have framebuffer problems in base-config there
13:52 < joshk> waitawte
13:52 < joshk> wait
13:52 < kmuto> but I dunno sid debootstrap status
13:52 < joshk> if you have a tg3 on i386 or amd64
13:52 < joshk> it takes a LONG time waiting for firmware
13:52 < joshk> which it never gets
13:52 < joeyh> the lowmem patch has bugs, I don't know if zboob will get it fixed in time
13:52 < joshk> just an annoyance, wondering if anyone had seen it
13:53 < joeyh> haven't seen it
13:53 < joeyh> also, we should have full CD#1 for i386 now with all the tasks and a usable desktop on it
13:53 < joeyh> so I'm hoping to include the full cds in the release
13:53 < anibal>     * ia64
13:53 < kmuto> kernel 2.6.7 will be installed after 2 days
13:54 < SvenLuther> joshk that is the non-free firmware we removed for DFSG compliance thingy ?
13:54 < kmuto> op
13:54 < joshk> SvenLuther: no idea
13:54 < bdale> tg3 had non-free firmware, yes
13:54 < Mithrandir> joshk: I've seen it with dfs.
13:54 < SvenLuther> joshk : probably, we do that for tg3.
13:54 < joeyh> dannf: around?
13:54 < joshk> is there a way to stop the delay?
13:54 < vorlon> joeyh: how much of your CD work for i386 carries over to the other ports?
13:54 < lieb> ia64 efi partman patches were submitted thsi week.  also re-writes of install manual ia64 portions
13:54 < bdale> I haven't tried ia64 recently either, but it's easier for me to test remotely than hppa (faster, better serial console support)
13:54 < joshk> hmm, maybe we should echo /bin/false > /proc/sys/kernel/hotplug
13:54 < joshk> or something
13:54 < joeyh> vorlon: to some extent, it depends on size issues of course
13:55 < joeyh> we probably have _more_ tasks on the first cd for other ports than before
13:55 < bdale> lieb: partman changes allow auto-partitioning to work right?
13:55 < SvenLuther> maybe we would need a set of non-free .udebs which include this frimware or something ?
13:55 < lieb> those patches need to be uploaded and manual translation
13:55 < joshk> they would never be included on stock media anyway
13:55 < vorlon> ok.  BTW, was a solution ever found to the problem with debian-cd using the wrong debootstrap package list to compile the netinst image contents?
13:55 < maswan> btw, I should probably say this in here too, I'm working on debian-cd stuff for amd64, but ran out of time on friday and haven't had time to poke it (and proably won't have until monday/tuesday or so).
13:55 < joeyh> lieb: is that in svn already?
13:55 < lieb> bdale, yes. a new partman-efi which both checks for correct efi and has recipes w/ efi
13:55 < SimonRichter> SvenLuther, most of them need to be distributed separately, probably even by different entities
13:56 < lieb> partman wsa not in svn as of yesterday
13:56 < vorlon> We're getting quite a few lib packages & other cruft in base by now that could be trimmed to our advantage.
13:56 < Md> joshk: please don'
13:56 < Md> joshk: please don't mess with hotplug...
13:56 < bdale> lieb: there's an HP-specific service partion guid that we should teach it to know about and leave alone, too...
13:56 < fs> joshk: tg3 is broken on amd64? the kernel is build with debian 2.6.7  kernel-source and -patch, the problem will arise on i386-2.6 too as sppn as 2.6.7 hits sarge
13:56 < joeyh> lieb: it would be good to get that in for the string freeze
13:56 < joeyh> and uploaded, in general
13:56 < fs> s/sppn/soon/
13:56 < lieb> patches for manual have been sent in against the bug reports
13:56 < Dr_O> SvenLuther:  non-free .udebs with the tg3 firmware would be nice!
13:57 < SvenLuther> Dr_O: patches are welcome.
13:57 < fs> a lot of amd64 boards have tg3 nics, we support is crucial for this arch
13:57 < anibal> next arch
13:57 < anibal>     * m68k
13:57  * joeyh cringes
13:58 < SimonRichter> Amiga bootloader sucks
13:58 < joeyh> we got confirmation today that the genext2fs bug is still there
13:58 -!- nwp [~nwp@210-55-222-75.adsl.ihug.co.nz] has joined #debian-boot
13:58 < SvenLuther> fs: it is non-free, debian doesn`t compromise on that.
13:58 < lieb> anyway, 2.6 and cleaned up (menu boot a la i386) in process but 2.6 is just starting to get in no status on it from me assume not working (yet)
13:58 < fs> is the tg3-firmware-removal patch for 2.6.7 broken? afaik 2.6.6 worked
13:58 < vorlon> did ia64 get skipped?
13:58 < joeyh> I'm really worried about m68k. Is anyone able to work on it?
13:58 < Mithrandir> fs: tg3 works just fine without firmware, IME.
13:58 < vorlon> ah, no, I was just sleeping.
13:58 < fs> Mithrandir: yes, that was my last status too
13:59 < SimonRichter> joeyh, I'm working around bugs in the bootloader
13:59 < SimonRichter> joeyh, working on the bootloader itself needs to wait until after sarge
13:59 < joshk> fs: it's not broken, it just waits a long time for firmware
13:59 < SvenLuther> SimonRichter: it is not possible to put it in main anyway.
13:59 < SimonRichter> joeyh, because we'd need a toolchain for that.
13:59 < Md> Mithrandir: it works slower, because it will lack hardware checksumming. if we ship the driver it should be renamed...
13:59 < SvenLuther> since it needs the non-free amigaos to build.
14:00 < joeyh> Florian Lohoff posted to -boot today and reported the ext filesystem error problem on boot
14:00 < SimonRichter> SvenLuther, yep, and some sources seem to be lost
14:00 < joeyh> we truely need to track that down
14:00 < SvenLuther> SimonRichter: have you investigated amililo ?
14:00 < maswan> Md: *shrug*, it works at line speed
14:00 < SimonRichter> SvenLuther, not yet
14:00 < Md> maswan: using more CPU
14:00 < SvenLuther> Roman and co should have it. I know Geert runs it.
14:01 < SimonRichter> SvenLuther, I'm putting it on my "to check out" list
14:01 < maswan> Md: Ok, I'll remember to take a lok at that this week to see if there is any significant cpu usage
14:01 < anibal> next arch:
14:01 < SvenLuther> that's me :)
14:01 < fs> hm
14:01 < anibal>     * mips,mipsl
14:02 < maswan> Md: If so, well, you can run off and get the firmware in a deb somewhere or so.
14:02 < SvenLuther> ah, no, not yet.
14:02 < joeyh> someone reported on the list that the lowmem problem with mips seems to not be a problem now
14:03 < SimonRichter> okay, I'm going to look after m68k/Amiga a bit, but I'd need support for the other subarches
14:03 < joeyh> SimonRichter: are you familiar with the genext2fs problem? (We can discuss this later)
14:03 -!- bbc [~bbc@CM-vina5-165-4.cm.vtr.net] has quit ["Leaving"]
14:03 < anibal> next arch:
14:03 < anibal>     * powerpc
14:03 < SvenLuther> ok, that's me, and i will let simon comment on apus.
14:03 < SimonRichter> joeyh, not yet
14:04 -!- KA9Q [~chatzilla@65.115.140.226] has joined #debian-boot
14:04 < SvenLuther> Ok. Powerpc is in pretty good shape. The real problem is that we really want 2.6 to be the default kernel on it, and this is right now problematic.
14:05 < joeyh> so, I think the main holdup on making 2.6 the default is lack of wide testing
14:05 < SvenLuther> Well, not problematic, but sarge doens't have the 2.6 udebs, so things like filesystem are broken.
14:05 < vorlon> data point: for my iBook, I can't get X to run using the 2.4 kernel, but installing 2.6 worked great.
14:05 < joeyh> we had good luck with offering 2.6 as an option for i386.Lots of people still tried it
14:05 < SvenLuther> joeyh: you said so already, but i don't believe so.
14:06 < SvenLuther> joeyh: i believe that most powerpc people already run 2.6.
14:06 < joeyh> at a minimum, you'll need to offer both, I can't see dropping 2.4 with 2.6 as of yet unproven
14:06 < joeyh> which is the defasult, <shrug>
14:06 < SvenLuther> joeyh: and i _DONT_ want to d o lot of 2.4 support during sarge's lifetime.
14:07 < SvenLuther> joeyh: i would make the 2.6 kernel the default, and have 2.4 as fallback situation (linux24).
14:07 < joeyh> manty asked you to email him about the CD stuff btw
14:07 < SvenLuther> We need to make this change before the rc release.
14:07 < SvenLuther> yes, i have seen it, i will write to him after the meeting.
14:07 < anibal> next arch:
14:07 < SimonRichter> anibal, sec..
14:07 < anibal> ok
14:08 < SimonRichter> APUS status: 2.4 doesn't boot, 2.6 doesn't build
14:08 < SimonRichter> both seem trivial actually
14:08 < SvenLuther> Kamion mentioned the discover problem also, since there seem to be discover 2 needed for 2.6, but i am notknoweldgeable about this problem.
14:08 < SimonRichter> the 2.4 issue seemed to be debug sections which confused the bootloader, but after stripping those sections, the problem persists.
14:09 < joeyh> yes well, that's a good example of why you can't just expect 2.6 to magically work everywhere before lots of people have tried it
14:09 < SvenLuther> notice that apus would be nice to have but is not RC quality. There are hardly more than 10 or so users left :)
14:09 < CIA-3> debian-installer: athornton-guest * r17827 packages/arch/s390/ (3 files in 3 dirs):
14:09 < SimonRichter> the 2.6 issue seems to be an error in the kernel makefiles, fortunately at make install time.
14:09 < CIA-3> debian-installer: Adam Thornton 2004-07-17
14:09 < SvenLuther> joeyh: the apus and the powerpc kernels are two different beast. Not even the same package.
14:09 < CIA-3> debian-installer: Changed CTC Protocol debconf priority to critical. Changed default protocol to
14:09 < CIA-3> debian-installer: "0" (S/390). Changed module probing to reflect the fact that S/390 needs the
14:09 < CIA-3> debian-installer: "netiucv" module, not just the "iucv" module, to do TCP/IP over IUCV.
14:10 < joeyh> I was referring to discover
14:10 < joeyh> feh, athorthon should be here
14:10 < SvenLuther> joeyh: ok. What i don't understand is why it is working right now if there is a problem with discover though.
14:10 < SimonRichter> plan is to get both issues fixed by tuesday
14:10 < SvenLuther> SimonRichter: cool.
14:10 < anibal> next arch:
14:10 < anibal>     * s390
14:11 < SvenLuther> anibal: wait, i want to know about this discover issue.
14:11 < anibal> okay. back to ppc
14:11 < joeyh> SvenLuther: discover 1 works with 2.6, but not ideally. Some module names hav changed.
14:11 < SvenLuther> joeyh: ok, does these modules affect us ?
14:11 < joeyh> see vast discussions of this for i386 2.6, unless kamion has some new issue
14:12 < SvenLuther> also, i am confused, since hotplug comes after discover qnd covers up for anything discover missed.
14:12 < SimonRichter> SvenLuther, the same thing bit me on the Pegasos, it seems
14:12 < joeyh> hotplug does not run in d-i
14:12 < joeyh> waldi: so, here?
14:13 < SvenLuther> joeyh: so it probably means that those ,odules don't affect us.
14:13 < SimonRichter> SvenLuther, I needed to manually insert eepro100 since discover loaded ee100 or something such, and also wrote that into /etc/modules
14:13 < Md> SvenLuther: after the system has been installed discover should not try to discover PCI/USB/etc
14:13 < SvenLuther> Ok.
14:13 -!- Loevborg [foobar@p508BBEA1.dip.t-dialin.net] has quit ["Leaving"]
14:13 < SvenLuther> Let's go forward and discuss this elsewhere.
14:13 < anibal> let's move on
14:13 < anibal> next arch:
14:13 < anibal>     * s390
14:13 < SvenLuther> joeyh: still, the fact that d-i hardware detection and post reboot module loading is not the same order is bothering me.
14:14 < joeyh> it is in the same order.
14:14 < joshk> s390 twice?
14:14 < anibal> s390 status, please
14:14 < joeyh> well, if no s390 people hare here, I know a tiny bit
14:14 < joeyh> waldi and athornton are workin gon it.
14:14 < joshk> someone filed a report
14:14 < joshk> yes
14:14 < SvenLuther> joeyh: well, not. i usually get ethernet an ip-over-firewire exchanged post reboot.
14:14 < joeyh> I think it mostly works first stage, still need ssh for second stage
14:14 < joeyh> SvenLuther: I fixed that 2 weeks ago.
14:15 < SimonRichter> SvenLuther, d-i detection makes a list of 2.4 module names to load. When using 2.6, this doesn't work.
14:15 < joeyh> perhaps not on powerpc, shrug
14:15 < anibal> last arch:
14:15 < anibal>     * sparc
14:15 < SvenLuther> joeyh: oh, that is why it is different when choosing sarge and sid ?
14:15 < joshk> ok
14:15 < joshk> first of all
14:15 < joshk> the best news is that SILO has purportedly been fixed
14:15 < kmuto> nice
14:15 < joshk> brian larson is going to give a new netinst a runthrough on one of his previously busted machines
14:15 < joeyh> joshk: I wanted to suggest that you go and mail everyone who reported it failed
14:15 < SvenLuther> BTW, why don't we use discover2 then ?
14:16 < joshk> joeyh: i'll get to it,heh, work has been hell :)
14:16 < joeyh> there are a lot of install reports you can grep for by sparc.*boot.*fail
14:16 < joshk> k
14:16 -!- anton_ [~anton@212.39.88.133] has left #debian-boot []
14:16 < joshk> and then there's the possiblity of having 2.6 for sparc64 only
14:16 < joeyh> SvenLuther: jesus, go read the threads. This is old news.
14:16 < joshk> (because the sparc32 tree for 2.6 is busted)
14:16 < joshk> for like, 70% of machines
14:17 < joshk> like sven i wouldn't like to hang on to 2.4 for much longer
14:17 < joeyh> did you work around the smp kernel problem you mentioned last time?
14:17 < joshk> but it's still usable
14:17 < fs> does 2.6 work on ultrasparc 1 smp now?
14:17 < joshk> yes
14:17 < joshk> should
14:17 < fs> ok, giving it a try next week when I have physical acces to my U2 again
14:17 < joshk> that's basically it. 2.4 might be moving to a well tested initrd setup if makedev gets fixed
14:17 -!- anton [~anton@212.39.88.133] has joined #debian-boot
14:18 < joshk> and 2.4.27 gets released
14:18 < SvenLuther> joeyh: about discover, i mean this was old news month back, has nothing changed since then ?
14:18 < joeyh> SvenLuther: correct
14:18 < joshk> it makes life a lot easier for me because of the 1.5MB kernel size upper limit on sparc32
14:18 < joshk> anyway - i think taht's it fo rme
14:18 < joshk> sparc is largely golden
14:18 < SvenLuther> joeyh: and nothing is planned to change before the sarge release ?
14:18 < waldi> joeyh: what?
14:18 < joeyh> waldi: s390 summary, what did I miss?
14:18 < SvenLuther> joeyh: nobody is working on it and nobody shows interest.
14:19 < waldi> i did some work on ssh but still not finished
14:19 < joeyh> think it will be working in 2 weeks?
14:20 < anibal> next topic:
14:20 < joeyh> it would be rather nice to finally release s390
14:20 < waldi> joeyh: should be possible
14:20 < joeyh> please try, even if it's not perfect
14:20 < CIA-3> debian-installer: athornton-guest * r17828 packages/arch/s390/ (netdevice/debian/changelog zipl-installer/debian/changelog): Documented changes and referenced numbers of fixed bugs.
14:20 < CIA-3> discover1: joshk * discover1-data/debian/changelog: typo
14:21 < anibal>     * d-i automation
14:21 < waldi> okay
14:21 -!- Irssi: T-D: Removed: Topic: Architecture status rollcall
14:21 -!- Irssi: T-D: Added: Topic: d-i automation
14:21 -!- SimonRichter changed the topic of #debian-boot to: Welcome to the Debian-Installer development channel -- http://www.debian.org/devel/debian-installer | For help mail debian-boot@lists.debian.org | Meeting right now | Topic: d-i automation
14:21 < SvenLuther> anibal: what is missing after this topic ?
14:21 < joshk> ok. so, i've been giving automation a bit of thought
14:21 < joshk> the dumb way is to create a fictitious debconf priority above critical
14:22 < SvenLuther> joshk and ?
14:22 < anibal> SvenLuther: it's in the agenda
14:22 < joshk> find a way to retrieve a debconf database 
14:22 < joshk> and then raise priority to that
14:22 < Mithrandir> joshk: why not just use a custom backend?
14:22 < SvenLuther> anibal: sorry, but i can have one windows only here.
14:22 < joshk> the elegance of the priority system (which, in the case of automation, could be considered a bug) is that if there's an error you'll return to interactive mode
14:22 < SimonRichter> joshk, if all questions are "seen", there shouldn't be an issue.
14:22 < SvenLuther> anibal : and my battery is running low.
14:22 < anton> Cant we use just critical instead of s.t. higher?
14:22 < joshk> Mithrandir: i don't know much about debconf frontends?
14:23 < Mithrandir> joshk: say, http backend, all answers are already seen, so they'll use the default answer.
14:23 < anibal> will finish soon, this is the last agenda item
14:23 < joshk> Mithrandir: that would be fine, i don't nkow how to do it
14:23 < SvenLuther> how does skolelinux do it ? I heard they already do fully automates installs.
14:23 < joshk> Mithrandir: however 
14:23 < joshk> anaconda is very good at fault tolerance
14:23 < Mithrandir> joshk: change deconf.conf to use the http backend, basically.
14:23 < joshk> partitioning is very vague if you look at a kickstart file
14:23 < joeyh> skolelinux installs at critical, and asks 3 questions
14:23 < joshk> it just says 'create a 4 gig / somewhere'
14:24 < maswan> joshk: as a user of that, what I'd prefer is interactive but an informative error message and hit return to reboot (and retry).
14:24 < joshk> so that you can install perfectly on a bunch of slightly different machines
14:24 < Mithrandir> joeyh: you don't want to ask even three questions if you are installing a 192 node cluster.
14:24 < maswan> isn't intreactive
14:24 < SvenLuther> joeyh: but it uses autopartkit, and not partman.
14:24 < Mithrandir> joshk: could autopartkit be told to do the same?
14:24 < Mithrandir> more or less?
14:25 < joshk> i don't know, autopartkit seems to be in the partitioner graveyard
14:25 < joeyh> joshk: partman-auto is close to that
14:25 < joshk> oh, also, i forgot to mention something about sparc and partitioning
14:25 < joeyh> the only reason I can see to use autopartkit is that it can do lvm
14:25 < SvenLuther> joshk: isn`t partman auto using debconf to fill partition choices and such ?
14:25 < joshk> what's a good way to make sure that either /boot or / is in the first gig of the disk?
14:25 < maswan> Mithrandir: well, if you could eliminate those 3 questions and add a resonable way to customize what is installed and where or so.
14:25 < joshk> it bit trave11er on one of his installs. also /boot or / has to be ext3.
14:25 < joshk> or ext2
14:25 < Md> joshk: there is an ioctl you can use
14:25 < joshk> Md: isn't there a way to ask partman?
14:26 < Md> joshk: I don't know
14:26 < joshk> hmmm
14:26 < joshk> i thought vorlon was just talking about this sort of thing
14:26 < joshk> vorlon?
14:26 < SvenLuther> what is the status of partman maintenance anyway ?
14:26 < bdale> lieb: is partman-efi where I can stare at the sources?
14:26 < Md> joshk: but yes, looking at the partitions table is probably easier
14:26 < maswan> btw, our method of reinstalling a node today is "bootsector -d /dev/hda; reboot" and n minutes later you have a fresh install just as we want it.
14:26 < vorlon> about what?
14:26 < lieb> bdale, it is an attach to the bugreport info
14:26 < joeyh> anton is here, but I don't think he's listening
14:27 < lieb> I can also send the patches direct
14:27 < joshk> vorlon: doesn't alpha need some special partitioning setup for aboot to work?
14:27 < bdale> lieb: please.
14:27 < vorlon> joshk: yes.
14:27 < anton> He is.
14:27 < Mithrandir> maswan: where does that get its config from?
14:27 < vorlon> It needs reserved space at the front of the disk, and the only way I've come up with to ensure this is by allocating a small special partition as partition #1 at the front of the boot disk.
14:27 < maswan> Mithrandir: This is FAI, so it is doing clunky magic and some brutal rewriting/copying of conf files.
14:27 < anton> Can you see this?
14:27 < SvenLuther> anton: what is your impression of partman ? Is it ready ?
14:27 < Dr_O> so most of the automation can be done using debconf db via http what can NOT use this?
14:28 < joeyh> anton: hi
14:28 < anton> SvenLuther: look at bugs with grave priority.
14:28 < SimonRichter> would it be possible to have an optional script for automatic installs that determines the disk layout?
14:28 < SvenLuther> anton: no web browser here, but i will. What is your impression though ?
14:28 < Mithrandir> Dr_O: I'm not sure, as I've only written the code for http db, but it's never been tested in an actual install (to the best of my knowledge).
14:29 < SimonRichter> the questions for that script would be Priority: low
14:29 < maswan> Mithrandir: See why I'd like a the nice d-i to do the installation and then a postinst.pl that reads netgroups and do class-specific customization.
14:29 < anton> At present - many easy to fix bugs.  The real impression will come when they are fixed.
14:29 < SimonRichter> and people wishing to set up a cluster would preset them to use a script that fits their needs.
14:29 < joshk> maswan: heh, anaconda does this with kickstart, indeed
14:29 < SvenLuther> anto: no big problems remainig though; right ?
14:29 < joeyh> anton: do you have much time to work on it lately?
14:30 < Mithrandir> maswan: you know, the answers to the debconf questions could come from a CGI script. ;)
14:30 < Mithrandir> maswan: but you probably want a postinst of some sort anyhow, I agree.
14:30 < maswan> Mithrandir: hehe, yeah, that'd work. :)
14:30 < maswan> Mithrandir: Yeah, and we have a tree of locally added packages that are in afs and not debs anyway, so..
14:30 < anton> The coming week I will have enough time to work but mostly offline.
14:31 < maswan> Mithrandir: Hmm.. You'd need to feed it a list of packages to install too via a cgi, I guess.
14:31 < kmuto> hmm, I have a question. d-i auto installer feature is an action item of next d-i release?
14:31 < SvenLuther> anton: how  does that play with the string freeze joeyh called for tuesday ? Any UI changes needed ?
14:32 < maswan> Mithrandir: well, via http anywya.
14:32 < Mithrandir> maswan: a local apt repository and a task, perhaps.
14:32 < SvenLuther> kmuto: i would say no. There is no time for it. We hope to have it in time for sarge though.
14:32 < maswan> Mithrandir: yeah, sounds resonable
14:32 < kmuto> SvenLuther: OK
14:32 < anton> SvenLuther: I will try not do do string changes after tuesday.
14:33 < SvenLuther> anton: Ok.
14:33 < anibal> closing meeting, objections?
14:33 < SvenLuther> fine with me.
14:33 < anibal> joeyh?
14:33 < kmuto> our release target date is end of month, right?
14:34 < SvenLuther> Mithrandir, maswan : BTW automatic install is only half the problem though, automated upgrades/instaklled system reconfiguration is also needed.
14:34 < joeyh> sorry?
14:34 -!- anton [~anton@212.39.88.133] has quit ["[BX] This BitchX's for you"]
14:34 < anibal> beta release is end of month, right?
14:34 < joeyh> yes
14:35 < maswan> SvenLuther: Well, for us, upgrade = reinstall. system reconfiguration = reinstall.
14:35 < Mithrandir> SvenLuther: well, you could just reinstall; I imagine that is what maswan might want to do.
14:35 < SvenLuther> joeyh beta5 or rc2 or ?
14:35 -!- Irssi: T-D: Removed: Meeting right now | Topic: d-i automation
14:35 -!- Irssi: T-D: Added: Meeting just over
14:35 -!- SimonRichter changed the topic of #debian-boot to: Welcome to the Debian-Installer development channel -- http://www.debian.org/devel/debian-installer | For help mail debian-boot@lists.debian.org | Meeting just over
14:35 < AstinusUK> Okay, anyone free to help me for 5 minutes??  Still need to get this damned HPPA running somehow - might involve reburning a new install disk with different kernel on, but you guys are the experts here -- I'm just the n00bie!
14:35 < SvenLuther> maswan: you are from FAI ?
14:35 < maswan> Mithrandir: We already do that fairly often, more often on a quick-booting machine.
14:35 -!- anton [~anton@212.39.88.133] has joined #debian-boot
14:35 -!- eugen [~eugen@news.univ.kiev.ua] has left #debian-boot []
14:35 < maswan> SvenLuther: I use FAI currently at work. It is quite a bit of work to setup though.
14:35 < joeyh> SvenLuther: I suppose what we call it will depend on how good it is
14:36 < anibal> meeting closed. thanks eveyone.
14:36 < joeyh> thanks anibal
14:36 < SvenLuther> joeyh: hehe.
14:36 < xdsg> I'm using the nightly dated 20040716; I'm at the GRUB install step and GRUB is sucking CPU, but apparently doing nothing.  I'm using expert26 to install SID, and I did select XFS as the root partition format; Is this what the XFS warning was about, and if so, can I just kill grub-install, chroot /target, and install grub myself?
14:36 < anton> I was dropped.
14:36 < maswan> The thing is, the freshly reinstalled system is the how all machines _should_ work.
14:36 < Mithrandir> removed entry msg.
14:36 < kmuto> anibal: thanks, good job
14:36 < SvenLuther> see you all.
14:36 -!- rollingbits [rbits@200.213.105.134] has joined #debian-boot
14:36 < maswan> not a reinstall with half a dozen upgrades on top
14:36  * SvenLuther happy that my ibook battery lasted the whole meeting :)
14:37 < vorlon> xdsg: it should work if you re-run grub-install from the chrooted directory.  And yes, you only get that warning if you've configured your partitions in a way that's not supported.
14:37 -!- lieb [~lieb@adsl-67-117-70-170.dsl.sntc01.pacbell.net] has quit [Remote closed the connection]
14:37 < anibal> joshk: please post the URL of the log file.
14:37 < joshk> oh, right, hold up
14:37 < SvenLuther> maswan: well; i have to leave, but am interested in this kind of stuff, maybe we should start a little sub-di project or something to discuss this kind of stuff ?
14:38 < SvenLuther> Ok, really need to go now.
14:38 < SvenLuther> bye