Technical_2005-12-13
08:59 Keybuk ok
08:59 Keybuk let's get this show on the road
08:59 Keybuk sabdfl is at dinner, so won't be coming
08:59 Keybuk and mdz has buggered off on holiday
09:00 mjg59 Ok
09:00 mjg59 So what have we got lined up for us today?
09:00 jdong the crowd ain't as big compared to CC
meetings, huh?
09:00 jbailey Keybuk: So as in, you don't have quorum
anyway?
09:00 Keybuk is Jani Monoses here?
09:00 janimo here
09:00 Keybuk jbailey: we've always considered two a quorum
09:00 jbailey Ah, cool.
09:01 Keybuk janimo: you've proposed yourself for main
upload privileges
09:01 janimo yes, for xfce/xubuntu packages
09:01 janimo mainly for that
09:01 Keybuk would you like to make your case
09:01 Keybuk well, I can certainly say you've been doing a
lot of work on them
09:02 Keybuk have you been doing much work with the
existing MOTU team?
09:02 janimo I'd like xubuntu to be supported and CDs
built for it for dapper
09:02 janimo Keybuk, not much MOTU work recently
09:02 janimo only xfce related uploads
=== Burgwork [n=corey@S010600131016cf6f.gv.shawcable.net] has joined
#ubuntu-meeting
09:02 Keybuk how would you like to improve xfce's
integration with ubuntu?
Janimo: the point is that once you get main
09:02 fabbione upload privileges, you can theoretically
upload everything
09:02 janimo it is well integrated already, but a few more
things to add
09:03 fabbione there is no distinction at the moment
09:03 janimo add gdm and make it a CD
09:03 janimo fabbione I know
I have uploaded an informationa
09:03 janimo XubuntuDapperRoadmap spec today with more
details
09:03 mjg59 janimo: So, what does Xubuntu involve?
09:04 janimo ubuntu standard + x +xfce packages
09:04 janimo + a few other apps
=== ogra would welcome xcfe in main as well as someone who cares for it
like janimo does, in edubuntu the question for a slim desktop comes up a
lot
09:04 janimo right now we have xubuntu-desktop
09:04 janimo the mmain gripe of usres is missing a CD
install
09:04 janimo that would be alot easier if all packages
were in main
09:05 mjg59 janimo: Is the aim to be as well-integrated
as Ubuntu?
09:05 janimo mjg59, rigth
09:05 mjg59 That is, should Xubuntu aim to fulfil the
same specifications?
09:05 janimo easy to use, always free things like that you
mean? sure
09:05 janimo or technical specs?
09:05 fabbione janimo: both
janimo: I was thinking more in terms of
09:05 mjg59 NetworkMagic, power management, that sort of
hting
09:05 janimo well then answer is yes for both
09:06 fabbione it needs to match the same quality and
consistency
09:06 janimo yes, I intedn to work on power mgmt
09:06 Keybuk how would you plan to get others to help with
the packaging work?
09:06 janimo and most of the related specs as I cab
09:06 janimo Keybuk, it is not much packaging involved, as
we use debian packages
I certainly think that the aim for anything
09:06 mjg59 in the *ubuntu namespace should be to provide
roughly the same level of user experience,
but it sounds like you're aiming for that
09:06 janimo I intend to work with them
09:07 Keybuk have you had to make many modifications to
the packages yourself so far?
09:07 mjg59 janimo: Have you been working with them so
far?
mjg59, yes aiming for that but xfce is not
09:07 janimo the same user experience as gnome/kde even
upstream so that transaltes here too
09:07 janimo Keybuk, yes a few
09:07 mjg59 janimo: Sure, I realise that there are
limitations
09:07 janimo to packaging and to code in packages
so you feel you know your way around
09:08 Keybuk Debian/Ubuntu packaging? have you done much
else besides XFCE?
09:08 janimo Keybuk, some of the changes are not yet in
debian, some are not upstream xfce either
09:08 janimo but pushing them actively
there are even heavy technical differences
between edubuntu/ubuntu (i.e. edubuntu
09:08 ogra installs a bunch odf server stuff which
doesnt comply with the "no open ports"
directive)
09:08 ogra so i wouldnt see the binding to tight
09:08 janimo Keybuk, I have been a MOTU since about before
Hoary and worked on some merges back then
09:09 janimo but since then it is mostly xfce related
changes or new packages
09:09 janimo all starting with xubuntu-
09:09 mjg59 janimo: Is there a reasonable xfce userbase
right now?
09:09 Keybuk dholbach, ogra: have you worked with janimo
much?
09:09 janimo some in universe some only in REVU
09:09 janimo mjg59, I don;t know how to estimate that :)
09:09 ogra Keybuk, yes, for a while, he's a longstanding
member of MOTU
09:09 mjg59 janimo: Heh. Well, do you get bugs filed?
09:09 janimo I'd say 10%-15% of linux users?
09:10 janimo duno honestly
09:10 janimo yes
09:10 janimo but xfce has a _lot_ less bugs then gnome/kde
even upstream, it is a lot simpler
09:10 mjg59 Sure
09:10 Keybuk anyone else got any questions for janimo ?
But it sounds like there are people using the
09:10 mjg59 packages, and from what ogra says it sounds
like there would be good reason to have them
in main
09:10 janimo mjg59, yes
09:11 fabbione janimo: are you familiar with security?
I get the occasional thank you eamail from
09:11 janimo time to time , with the request for CD images
:)
09:11 fabbione if so how much?
09:11 fabbione how much can you commit for it?
mjg59, absolutely ... we had a lightweight
09:11 ogra target for hoary already which wasnt
completed, xubuntu might fill that gap
09:11 janimo fabbione, not much besides occasional peeks
at bugtrack but I am cautios
09:11 pitti btw, there has been one XFCE related
vulnerability so far, in 2000
09:11 janimo fabbione,commit for security?
09:12 janimo pitti, another code base xfce3
09:12 janimo so 0 actually :)
09:12 pitti janimo: right, what I wanted to say is,
that's damn few
janimo: if xubuntu enters main, it also means
09:12 fabbione providing -security support. are you ready to
commit to it?
09:12 janimo shame on them
09:12 ogra heh
09:12 janimo fabbione, yes both secutiry and updates
09:12 ogra according to pitti that might be a boring job
:)
09:13 janimo indeed :)
janimo: does xfce ship with any suid programs
09:13 pitti for doing hardware magic? does it handle
hardware at all?
09:13 janimo no suid programs
09:13 pitti janimo: like automounting, networking, etc.?
09:13 jdong pitti: I believe it uses g-v-m to handle
those things
my one comment is that we work very much as a
team, and so far you've been mostly working
alone on these packages -- so once in main,
09:13 Keybuk you'll need to both help others work with
your packages and also be ready to sponsor
uploads for the MOTU ... are you happy with
that?
09:13 janimo for automounting we will use ivman now
kubuntu does not
09:13 janimo jdong, no ivman
09:13 jdong ah, ok
09:14 jdong been quite a while since I've used XFCE
09:14 janimo Keybuk, I am happy
09:14 ogra Keybuk, he worked a lot with crimsun
09:14 ogra in a two man team
09:14 ogra (its not the one man show it seems o be)
09:14 Keybuk mjg59: any further questions?
09:14 pitti that's twice as big as our edubuntu and
kubuntu teams :)
Keybuk I have been working alone since MOTUs
09:14 janimo are very busy with merging and their other
teams
09:15 ogra pitti, yeah
09:15 janimo crimsun is swamped with multimedia related
work
janimo: It would be good to see work stick
09:15 mjg59 fairly closely with the Debian packagers, but
I think you've said that you plan to work
with them already
09:15 sivang is this TB meeting?
09:15 janimo mjg59, I just got an alioth acount today to
get svn commit access
09:15 pitti sivang: yes
09:15 janimo to help pushing our changes up
09:15 mjg59 janimo: Cool
09:16 mjg59 Keybuk: Ok, I think I'm done
janimo: xfce has a file manager I
09:16 Riddell believe. does it have a web browser, music
player, console? (to pick the apps I have
open just now)
janimo: that's good, so the Debian
09:16 pitti maintainers acknowledge you as a
co-maintainer then?
09:16 Keybuk ok, your vote?
09:16 janimo Riddell, I think we'll use thunar as the
default fimemanager
09:16 mjg59 +1 from me
09:16 janimo we'll have firefox, xfmedai, xfce4-teminal
09:16 Keybuk +1 from me also
09:16 Keybuk Janimo: welcome aboard
09:16 pitti congrats janimo
09:16 janimo thanks all :)
09:17 ogra YAY, welcome janimo
09:17 janimo nice to join the team
09:17 ogra :)
09:17 mjg59 Ok, moving on? How many developer candidates
have we got today?
09:17 janimo I'll probably help out with other main work
if needed
09:17 Keybuk Johnny Mast, are you here?
09:17 ajmitch janimo: good work :)
09:17 Keybuk Brandon Hale, are you here?
09:17 janimo pitti, yes they are ok with me working close
woth them
09:18 ogra Keybuk, tseng is already in ubuntu core
09:18 Keybuk you know, I thought he was
09:18 ogra (brandon hale)
09:18 Keybuk didn't mentally match the name
09:18 sivang welcome janimo , to main? :)
09:18 Keybuk jdong: I can see you're here
09:18 jdong :)
09:18 janimo sivang, yes :)
09:18 Keybuk ok
09:18 Keybuk jdong: you've applied to join the MOTU team
09:18 sivang janimo: big congrets :)
09:18 ogra is vuntz here ?
09:18 jdong Keybuk: correct
so, I guess, first off, what have you been
09:19 Keybuk working on so far and what would you like to
work on?
Well, my work is primarily concentrated in
09:19 jdong Backports, and Backports does the most work
in Universe
as a result, sometimes the Backports team
09:19 jdong needs to have changes done in Universe and
such
09:20 jdong currently, the Team has Mez and slomo who are
on MOTU
09:20 mjg59 jdong: Is that "the most work in Universe" or
"does most of its work in Universe"?
09:20 jdong and at times that's not enough
09:20 vuntz ogra: I'm here, but I'm going away for ~30
minutes right now
09:20 ogra vuntz, just because you are on the list
mjg59: a combination of both: Most requested
09:20 jdong packages are in Universe, and most "issues"
we find are also stemming from Universe
packages
09:21 Keybuk right, but you don't upload the resulting
backports to universe?
09:21 jdong no, I don't
jdong: Your phrasing seemed to suggest that
09:21 mjg59 backports did more work in universe than any
other group
09:21 ogra Keybuk, the process for backports is a bit
different ...
09:21 jdong mjg59: maybe I'm having a bad day with
phrasing things :)
09:21 mjg59 jdong: Ok, as long as that's not what you
meant, then fine :)
09:22 ogra Keybuk, elmo just triggeres them if jdong
tested locally ...
09:22 jdong and ogra has been nagging for me to join MOTU
too :-D
09:22 jdong ever since day one
09:22 Keybuk ogra: so you'd support jdong's application?
09:22 mjg59 jdong: So how would being an MOTU change
things for you?
09:22 ogra jdong, but i can hardly judge your packaging
skills ...
mjg59, if changes have to be made to a
09:23 ogra package to backport fine, someone has to
touch the package ...
mjg59: Well, first off I hope that it would
09:23 jdong mean less of me bitching at others... and I
can do work without relying on others
mjg59, so it would be a requirement to have
09:24 ogra one of the backporters being able to do that,
Mez is rarely here recently, he wanted to
take that role
ogra: I'm pretty experienced with packaging,
09:24 jdong but not with the specific Debian tools used
to upload and such
09:24 mjg59 jdong: Well, we'd prefer there to be no
bitching regardless of the situation...
09:24 jdong mjg59: I suppose _nagging_ would be a better
word :)
09:24 mjg59 Heh. Sure.
09:24 jdong mjg59: I'm known as the one who walks in to
#ubuntu-motu just to ask for stuff to be done
the thing is that if a package needs minor
09:25 ogra changes, they have to be done (uploaded) by
someone with upload rights ....
09:25 ajmitch jdong: who has worked with you on packaging
changes that can vouch for those skills?
09:25 ogra so that will either put a burden on MOTU or
we get and uploader from the bp team
ajmitch: that's the thing -- I haven't really
09:25 jdong needed to do anything of that magnitude with
MOTU before
09:26 jdong ajmitch: usually the types of changes
Backports needs are very minor
09:26 jdong i.e. changing a version on b-d's, etc
and i'd like to see jdong in that role since
09:26 ogra he merely leads backports ... but you would
have to work a bit with the motu to prove
your skills i public
09:26 ogra s/i/in
09:26 jdong ogra: yes, absolutely :)
09:27 Keybuk ogra: would you preferred he worked with you
before being granted upload rights?
09:27 jdong ogra: and I'll try to spend time working on
other stuff in MOTU too
09:27 \sh g
Keybuk, not particulary with me personally
09:27 ogra but with the MOTU team so he can get some
votes for his knowledge
09:27 \sh evening...too late
09:28 ajmitch hi \sh
ok, I tend to concur; I'd prefer jdong to
09:28 Keybuk work a bit more with you guys so you feel
confident about his skills first
09:28 ogra Keybuk, i simply cant judge his skills ince i
never used any backport in ubuntu yet
09:28 Keybuk mjg59: what do you think?
09:28 ogra jdong, that would be rad :)
I'm broadly in favour, but I'd agree with
09:29 mjg59 Keybuk - some more time doing general MOTU
work would be good
I agree with ogra...I think it's better to
09:29 \sh have an overview about his skills in
common...
09:29 \sh for this, he has to come more often to -motu
jdong: would you mind working a bit more
09:29 Keybuk closely with the MOTU team, and come back at
the next TB meeting when they should be able
to praise your skills
09:29 jdong Keybuk: sure
=== jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca] has
joined #ubuntu-meeting
09:30 jdong I'll try to be in #ubuntu-motu more often
09:30 jdong and follow what the guys there want me to do
:)
ok, we'll postpone to then; your work looks
great so far to me, but we're very much a
09:30 Keybuk team distro, so come to the next TB meeting
with praise from the MOTU team for your work
so far there
09:30 jdong ok, cool
=== Hirion [n=hirion@draugr.de] has joined #ubuntu-meeting
ok ... now is anyone else here waiting for
09:31 Keybuk universe upload privileges who hasn't made it
to a meeting since they proposed themselves?
jdong: work with us on the rest of the
09:31 \sh merges...and bug fixes has to be done soon. I
think we agree to help you out with the
uploads..so we can see your work
09:31 Keybuk jbailey: are you fit and rebooted, or do you
need a bit more time?
09:32 jbailey Keybuk: I'm rebooted.
09:32 jdong \sh: absolutely I need much more packaging
experience for MOTU :)
09:32 Keybuk ok, Mirror Status, the floor is yours
09:32 jdong I just want to take the first step today :)
09:32 ogra jdong, thats really great, i'm happy you do
that :)
09:32 mjg59 jbailey: Can you explain what the problem is,
and how we should deal with it?
09:32 ogra (you know how log i nagged you for it ;) )
=== jbailey looks to see which order he put them in.
09:33 jdong :)
mjg59: The mirrors occasionally have trouble,
09:33 jbailey running out of space, or failing to update,
etc.
mjg59: Right now the only way to get this
09:33 jbailey information is by pinging James or Karl
directly.
09:33 jdong more than occasionally
09:33 ogra yes, the us mirror seems to be a PITA
09:34 elmo jbailey: you realise there's LP specs about
this, right?
That means that getting this information is
09:34 jbailey limited by timezones / availability of two
people.
09:34 jbailey elmo: Right. But that means that it's 6
months to a year away.
09:34 elmo jbailey: says who?
09:34 jbailey elmo: I don't remember seeing it on the
priority list for dapper.
The problem is right now when we know there
09:35 jbailey are mirror problems, that information is not
exposed in a useful way.
09:35 Keybuk so what would you propose instead of the LP
spec?
09:36 elmo when was the last time US mirror was broken
just out of interest?
09:36 elmo because AFAIK, it hasn't broken since we
started pushing it with a two stage push
09:36 jbailey elmo: Mmm. Right before ubz I think?
09:36 elmo and that was ? weeks? ago
=== seb128 [n=seb128@ubuntu/member/seb128] has joined #ubuntu-meeting
09:36 elmo jbailey: ...
09:36 jdong elmo: I got some md5sum mismatches fairly
recently (2 weeks?)
09:36 jbailey elmo: This was intended to come up at the
last TB meeting, which I couldn't make.
09:36 Keybuk tbh, I don't see anything here for the TB to
decide
09:36 Keybuk there's a LP spec for it, which is the
long-term solution
09:36 jbailey Keybuk: mdz refered me to the TB
09:37 Keybuk and the short-term solution appears to be
"ask elmo"
09:37 jbailey Right.
09:37 elmo jdong: dude, you tried to convince me you had
md5sum mismatches for breezy on archive.u.c
09:37 jbailey I'm saying that short term solution isn't
always useful.
09:37 Keybuk which is true for most of our archive
management right now
09:37 jdong elmo: Maybe my weeks are off :)
Right now that request is that we ask the
09:37 jbailey archive administrators to keep a status board
of some sort on mirror status so that known
issues can be looked at right away.
09:37 jdong elmo: but the first few weeks of Breezy's
release
09:38 elmo jbailey: why on earth are you going to the TB
in the first instance with this?
09:38 ogra which would be pre ubz
09:38 jbailey elmo: mdz refered me to the tb.
09:38 Keybuk imo, mdz should have referred you to elmo
09:39 jbailey Who ultimately makes the decisions about how
these issues are handled?
09:40 elmo the same person who ultimately makes all
decisions?
09:40 Keybuk mdz, sabdfl, elmo, etc.
So it was probably refered here for
09:40 jbailey discussion at a time when all the parties are
generally available.
jbailey: in any event my point is that you
09:40 elmo didn't even try asking the people who
currently run the mirrors
09:40 elmo which is, umm, nice
09:40 elmo really
09:41 jdong I gotta run, be back in like 15-20 minutes,
hopefully in time :)
09:41 Keybuk mjg59: anything from your corner?
elmo: My contact for distro things when
09:41 jbailey driven by customers asking me questions is
always mdz.
09:41 elmo jbailey: don't be obtuse
Without knowing why mdz thought it should be
09:41 mjg59 brought up here, I'm not sure we can do too
much
09:41 elmo if a customer asked you about gcc, you'd go
to doko, not mdz
09:41 mjg59 Can we punt this until mdz is around?
09:41 jbailey elmo: Actually, no. I'd email mdz.
09:41 jbailey Unless it was somthing I thought could be
answered very quickly.
yeah, that's my best suggestion right now ...
09:41 Keybuk wait until mdz is back, to see whether he had
any reason for bring it here
09:41 elmo well maybe the TB should look at how you
communicate with the distro team then ...
09:42 Keybuk until then, jbailey: 1) make sure the LP spec
meets your needs
presumably the idea of stuff going through
09:42 Kamion mdz is so that mdz can manage how much distro
team time is taken up by escalated support
09:42 Keybuk and 2) buy elmo some flowers, and ask him
nicely?
Kamion: Yes, as well as tracking and
09:42 jbailey stuff. Before he was a bit surprised that I
had been asking the distro team for things
and that he hadn't been aware of it.
09:43 Keybuk ok, so the advisory list thing
you're suggesting some kind of process that
09:43 Keybuk involves compiling a list of packages in
*-updates and what they chance?
09:43 Keybuk uh, change?
09:44 Keybuk similar to the existing security stuff?
Right - not necessarily a list, but an
09:44 jbailey explanation or advisory by email as to why
updates are coming out.
09:44 pitti oh, interesting idea
The notification applet stuff isn't always
09:44 jbailey sufficient to explain what's going on, or
that it's coming.
09:44 pitti but isn't apt-listchanges something that
should come close?
09:44 Kamion Could amber be modified to do -updates?
09:44 jbailey So in some ways it can look like a security
update that showed up without an advisory.
(speaking as an occasional -updates approver,
09:45 Kamion I don't want to have to construct those
things by hand)
09:45 jbailey Kamion: Right. I don't have any suggestions
as to the machanics of this.
Certainly I'd be willing to construct these
09:45 jbailey for the -updates that I do and add them to
the mix somehow.
from my corner, yes something like this would
09:45 Keybuk be nice, propose it on #ubuntu-devel and
rally people round to figure out how to do it
well, for updates we maybe don't need md5sums
09:46 pitti and such; but still some template would be
nice for version numbers etc.
09:46 mjg59 Right. I think it's reasonable to tell people
why we're giving them an update.
pitti: I'm not sure how a -update is
09:46 jbailey different than a security patch in that
sense.
09:46 Keybuk so +1 from me, sounds like a nice thing to do
09:47 pitti jbailey: well, as long as amber generates
these, I don't particularly care
pitti: In any event we're delivering them a
09:47 jbailey package, so it would be nice to have the
usual trappings. md5sum is the easiest part
to generate. =)
09:47 Keybuk mjg59: ?
=== seb128_ [n=seb128@ANancy-151-1-45-177.w83-196.abo.wanadoo.fr] has
joined #ubuntu-meeting
=== loogaroo [n=manuel@u-124-094.adsl.univie.ac.at] has joined
#ubuntu-meeting
09:48 mjg59 +1
09:48 Keybuk ok
09:48 Keybuk jbailey: if you could lead discussion and
implementation for it, go for it
09:48 jbailey No problem, i'll follow that through.
09:48 pitti Kamion: how are -updates uploads published?
09:49 Keybuk next thing on the agenda is me ...
Kamion: certainly there is a girl for it
09:49 pitti which could be modified to spit out template
emails?
this is probably a no-brainer, but I didn't
09:49 Keybuk want to go ahead and make such a change
without at least checking with some other
people first
=== seb128_ [n=seb128@ANancy-151-1-45-177.w83-196.abo.wanadoo.fr] has left
#ubuntu-meeting ["Ex-Chat"]
09:49 ogra Keybuk, breaking the wrold again ?
09:49 jbailey Mmmm. Modified girl.
09:49 jbailey (sounds ike an ingrediant)
=== seb128_ [n=seb128@ANancy-151-1-45-177.w83-196.abo.wanadoo.fr] has
joined #ubuntu-meeting
currently there's no useful place during boot
09:49 Keybuk to put things like pid files, control
sockets, dhcp leases, etc.
09:50 pitti /var/run isn't useful?
/var/run should be that useful place, but
09:50 Keybuk it's either on the root filesystem and not
yet r/w or on another filesystem altogether
09:50 mjg59 (See NFS-root cases)
09:50 Keybuk it's fine once in rc2.d, but in rcS.d it's
sub-optimal
09:50 jbailey Keybuk: Is this one beast to add to your
movemount collection?
09:50 jbailey +more
09:50 Keybuk so my proposal is to make /var/run a tmpfs
that's always available
09:51 pitti Keybuk: when would you mount that?
which would mean anything writing into a
09:51 Keybuk sub-directory of it would need to make that
sub-directory if it doesn't already exist
09:51 Kamion pitti: no it's just kelly -z
09:51 Kamion pitti: which is just the thing that does
accepted->installed
09:51 Keybuk pitti: probably in the initramfs, and moved
to /root/var/run
Keybuk: if you mount it in initramfs, then a
09:51 pitti later mount of /var in fstab would shadow it,
right? and mouting it late would defeat the
purpose?
09:51 Keybuk then if not already mounted, in the first
init script
and in the mountall script, if we mount /var
09:52 Keybuk as something else, so a move dance to keep it
there
09:52 pitti heh
09:53 Keybuk so that's it really, is this sane or should I
go find some other solution?
for the records, we already fixed one or two
09:53 pitti packages to create /var/run/<package subdir>,
but it certainly needs some more fixes to
have complete coverage
LSB does allow for packages to rely on the
09:53 pitti existence of /var/run/directory, so it could
give us some hard time with keeping up
09:53 pitti (especially for universe)
09:53 pitti but I didn't actually count the number of
affected packages
09:54 pitti the fix is trivial, though
09:54 Keybuk *nods*
Keybuk: So the main downside is that we may
09:55 mjg59 need to alter some packages that expect
subdirectories to be persistant?
09:55 \sh pitti: if this can be done after UVF...no
problem with me ,)
09:55 pitti \sh: it's a trivial bug fix, sure
09:55 Keybuk mjg59: right
09:55 pitti \sh: just a mkdir -p and a chmod/chown in the
init script
09:55 mjg59 Keybuk: Hm. I worry a bit about it being a
semantic change.
09:56 \sh pitti: so we need a list of all packages who
are bringing in some init.d scripts...
Keybuk: Do we have any idea how many packages
09:56 mjg59 it would break, and have you discussed this
with Debian at all?
09:56 pitti \sh: no, packages which put stuff into
/var/run
09:56 pitti \sh: and either ship a directory in /var/run,
or create it in postinst
in main, I've found only one or two that it
09:56 Keybuk breaks, and the Debian maintainer of both
seemed happy to accept the patch
09:56 fabbione there are also packages that have /var/run/
in the code...
09:56 Keybuk most of them already have patches to make it
work
09:57 mjg59 Keybuk: Ok, sounds good
09:57 pitti sounds like 'early breakage'
09:57 \sh I need an aspirin, a tea, some drugs, and a
list of packages for later..
\sh: this can add up to the list of remaining
09:58 sivang merges, or just add this when working on a
merge :)
09:58 sivang s/add/patch/
09:59 Keybuk ok, any further business?
09:59 Keybuk jdong: are you here?
sivang: if you have such a package to merge
09:59 \sh ... do it now...or after uvf and when all the
merges are done
10:00 ogra <jdong> I gotta run, be back in like 15-20
minutes, hopefully in time :)
10:00 ogra that was 20min ago
10:00 Keybuk yeah, sadly I've got to run in a minute too
:-/
10:00 Keybuk the answer to jdong is the same as jbailey,
make sure the LP spec is right
10:01 Keybuk several already exist, from what I can see
10:01 ogra i thought Mez worked on one at ubz
10:01 Keybuk ok, end of meeting then
10:01 Keybuk the next TB meeting will be in 4 weeks time I
guess
10:01 Keybuk as two-weeks is the 27th
10:01 mjg59 Wow. That was swift.
well...I think there is no need for a new
10:01 \sh process...because elmo is processing all the
request when he has time...and thats pretty
often..
10:02 Kamion oh, is there time for me to slot in one
question?
10:02 ogra yup... and LP is near (somehow)
10:02 Kamion forgot to put it on the agenda
10:02 Keybuk meh, go on then
10:02 jdong were my minutes too long?
10:02 Kamion if not, no worries
10:02 Keybuk is it quick?
10:02 Kamion so Debian is in the middle of killing off
base-config
10:02 Kamion and moving the entire installation to the
first stage
10:02 Kamion this would kill off a fair few of my bugs, so
I'm inclined to follow suit
10:02 pitti that sounds nice actually
10:03 Kamion my question is whether this is an utterly
insane thing to do for dapper?
10:03 Keybuk sounds nice to me, we've already done some of
that?
10:03 Kamion yes, but not the big long package
installation step
10:03 Keybuk ultimately I think that's possibly your call
;)
10:03 jdong Keybuk: you can drop my item from the agenda,
I don't think it'll be necessary as of now
10:03 Keybuk you'd judge better than anyone else whether
it's doable
the downside is that it means that it takes
10:03 Kamion longer to find out whether your
newly-installed system actually boots
10:03 jdong btw elmo thx very much for pushing those
backports through today!
10:04 ogra and it will slow down the installation, no ?
since you do everything from CD ?
Keybuk: ok, I thought it was a big enough
10:04 Kamion change that I ought to bring it up, but if
that's your feeling then fine, I'll judge by
how quickly the code matures upstream
10:04 Kamion ogra: no, actually it'll speed it up because
we can throw away archive-copier
10:04 ogra ah, k
10:04 ogra yes, thats a 10min step less ...
10:04 Kamion there'll be some extra seeking back and
forward, but I think it'll be dwarfed
=== sivang wil be waiting to test those images :)
10:05 Kamion righto, thanks
MeetingLogs/Technical_2005-12-13 (last edited 2008-08-06 16:19:32 by localhost)