[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

FESCo Meeting Minutes 2006-09-21



= 2006 September 21 FESCo Meeting =

Meeting Summaries are posted on the wiki at:
  http://www.fedoraproject.org/wiki/Extras/SteeringCommittee/Meetings

== Attending ==

 * thl
 * bpepple
 * awjb
 * warren
 * c4chris
 * spot
 * rdieter
 * scop
 * abadger1999
 * tibbs
 * jwb
 
== Summary ==

=== Comps.xml ===
 * One more round of nagmails before FC7.
 * comps.xml will start being generated in a subdirectory so people
aren't confused about editing comps.xml.in vs comps.xml
 
=== Approve KMods ===
 * FPB chose to allow kernel modules.
 * kmod guidelines are to be followed.
 * KMod blocker bug to be created so that it's easy to find all KMods.
 * Many members of FESCo want KMod approval to hinge on the kmod being
able to go upstream.  However, this is not yet documented or at full
consensus.
  * Rules for deciding whether a KMod should be approved needs to be
discussed on the list.
 * Kmods built by Extras should already taint the kernel because they
are not signed (which helps kernel maintainers separate bug reports that
involve kmods vs ones that are on the vanilla kernel.)
 
=== FESCo Decisions ===
 * How does FESCo make decisions when things come to a vote?
 * Proposed: voice vote first (ie, no one screams), else majority (of
those present).
 
=== Mass Rebuild ===
 * Packages that aren't rebuilt to be orphaned.
 * Can just follow the unorphan process to make things work.
 
=== EPEL/Enterprise Extras ===
 * Discussion to continue on f-e-l for now (rather than a new list.)
 * spot and mmcgrath to arrange a non-public RHEL 4 repository for the
builders.
 * Opinions solicited for EPEL repositories plan:
https://www.redhat.com/archives/fedora-extras-list/2006-September/msg00672.html
  * People want it as simple as possible to start.
  * Will packages land within Fedora Space:
http://download.fedora.redhat.com/pub/fedora/linux/extras/{el4,el5}
    or Red Hat space:
http://mirrors.kernel.org/redhat/redhat/linux/beta/RHEL5-Beta1/client/i386/
   
=== Packaging Committee Report ===
 * Selecting new members: open nominations followed by voting by current
members.  Currently no open seats to be filled.
 
=== Misc: Remove inactive members ===
 * Inactive members will have cvsextras removed from their account.
 * List of people with cvsextras removed will be kept on the wiki in
case it's needed to readd people.
 * Will also remove all the declined memberships.
 
=== Free Discussion: Competition Packaging ===
 * tor is the specific case that brought this up.
 * How to ensure that the technically "best" package wins?
 * c4chris: I think the idea is: do not encourage competition, but allow
it in some cases
 * No rules for the general case were established.  For tor abadger1999
agreed to take over reviewing.
 
=== Free Discussion: current-branches File ===
 * File to list which branches are currently active (as opposed to
maintenance or EOL).
 * This will be available in the packageDB; do we need the file until
the packageDB is ready?
 * To be discussed next week.
 
== Log ==

{{{
(10:00:03) thl has changed the topic to: FESCO Meeting in progress --
init
(10:00:09) thl: hi everyone
(10:00:13) thl: who's around?
(10:00:14) ***bpepple is here.
(10:00:18) ***awjb is here.
(10:00:21) warren: here, but I need to get my car real soon
(10:00:26) warren: (across the street at the shop)
(10:00:27) ***c4chris is here
(10:00:48) ***mmcgrath =true
(10:01:00) che left the room (quit: Read error: 110 (Connection timed
out)).
(10:01:03) ***nirik is in the rabble seats with his cup of coffee. 
(10:01:17) ***thl has some tea
(10:01:26) ***spot is here
(10:01:32) spot: but the packaging meeting is running a bit long
(10:01:33) ***c4chris just had a bear :-)
(10:01:40) c4chris: beer
(10:01:43) c4chris: oops
(10:02:07) scop [n=scop] entered the room.
(10:02:11) ***rdieter is 1/2 here
(10:02:18) ***thl wonders if starting here is wise until the pc meeting
finished
(10:02:22) ***scop ditto
(10:02:51) abadger1999: \me arrives
(10:03:08) thl: well, let's start with some easy stuff
(10:03:19) thl has changed the topic to: FESCO Meeting in progress --
Use comps.xml properly
(10:03:22) warren: I better get my car now... 
(10:03:30) thl: dgilmore: automate comps file during push or via cron;
see also [WWW] this mail on f-e-l
(10:03:58) thl: c4chris in parallel -- is there anything else regarding
comps.xml that needs to be done/discussed?
(10:04:19) spot: thl: maybe someone (not me) could check comps.xml
against FE and send out emails to maintainers?
(10:04:41) thl: spot, c4chris did that already (at least once)
(10:04:56) ***cweyl is here, post coffee-run (rabble)
(10:04:58) c4chris: should be ok
(10:05:05) ***spot is pretty sure most of his stuff isn't in comps, but
doesn't remember getting nagged about it
(10:05:11) c4chris: just planning to go through the list once again
before FC6
(10:05:37) c4chris: talked to jeremy about putting the non .in files in
a subdirectory
(10:05:58) c4chris: need to talk to mschwendt and dgilmore now...
(10:06:04) nirik: can you acl them off so people can't edit them?
(10:06:39) c4chris: spot, I need to check
(10:07:26) thl: okay, so let's move on then
(10:07:38) thl has changed the topic to: FESCO Meeting in progress --
Activate legacy in buildroots
(10:07:57) thl: dgilmore did some stuff -- I at least saw some commits
with new config files
(10:07:58) c4chris: spot, nagged you on Aug 16
(10:08:12) thl: anyway, he seems not be around, so I just skip this
(10:08:15) spot: c4chris: i'll have to go back and look at that
(10:08:20) ***scop 4/5 here
(10:08:25) pygi left the room.
(10:08:32) thl has changed the topic to: FESCO Meeting in progress --
Sponsorship nominations
(10:08:33) c4chris: spot, np
(10:08:39) thl: any new nominations?
(10:09:12) c4chris: pfj is still a candidate afaik
(10:09:16) ***mmcgrath ?
(10:10:09) thl: c4chris, I have problems with all the pauls...
(10:10:23) bpepple: Has mmcgrath done any reviews?
(10:10:28) thl: c4chris, pfj = nodoid?
(10:10:35) c4chris: thl, yes
(10:10:50) thl: c4chris, I told him some days ago to ask his sponsor for
his opinion
(10:11:09) c4chris: thl, ah, ok.  let's wait then
(10:11:45) thl: mmcgrath, reviews?
(10:12:45) mmcgrath: Just MySQL-query-browser in the last month.  I've
laxed quite a bit.  Save my nomination for another time when I have more
reviews :D
(10:12:54) bpepple: ok.
(10:12:58) thl: mmcgrath, k, thx
(10:13:17) thl: mmcgrath, if you need sponsorship status for EPEL stuff
just yell
(10:13:27) mmcgrath: will do.
(10:13:28) thl: k, moving on
(10:13:29) ***mmcgrath brb.
(10:13:39) thl has changed the topic to: FESCO Meeting in progress --
approve kmod's
(10:13:42) thl: just to note:
(10:13:55) thl: FPB chose to allow kmods in general
(10:14:25) thl:
http://www.redhat.com/archives/fedora-advisory-board/2006-September/msg00045.html
(10:14:25) scop: hmm
(10:14:30) jima: with or without fesco approval individually?
(10:14:36) thl: so I think we just go ahed with the old plans?
(10:14:48) thl: jima, they didn't want to get involved into the details
(10:14:49) abadger1999: jima: That's up to fesco (us)
(10:14:50) spot: jima: they said kmods are ok, we say which kmods are
ok.
(10:14:56) scop: thl, are you sure that message is an indication of a
decision?
(10:14:58) cweyl: thl: from my read of the email it was "ok to kmods,
you figure out the implementation details"
(10:15:03) jima: 'k.
(10:15:11) thl: scop, well, rdieter told so on fesco-list
(10:15:21) thl: cweyl, exactly
(10:15:30) rdieter: yay.
(10:15:35) spot: which kmods are pending approval?
(10:15:38) cweyl: very yay :)
(10:15:52) thl: spot, well, there are a lot (ten?) up for review
(10:16:05) spot: again, "which". :)
(10:16:13) awjb: ^^
(10:16:24) thl: spot, non are pending approval
(10:16:32) spot: so, we dont have to approve any from us?
(10:16:36) thl: but some modules up for review were not approved
(10:16:54) thl: spot, ?
(10:17:03) spot: sorry, i know that made absolutely no sense. :)
(10:17:09) cweyl: just for sanity purposes, this approval is the
preliminary "go ahead and package", not "approved to be imported",
right?
(10:17:10) spot: Does FESCO need to approve any kmods?
(10:17:22) tibbs: There have been some new submissions, but so far they
don't have the required information and need to be pinged.
(10:17:27) thl: spot, well, I think we should proceed as planed in the
past
(10:17:32) nirik: perhaps a KMOD blocker would be nice?
(10:17:34) thl: spot, so yes, we should approve them
(10:17:37) rdieter: how does it work, approve first, then review, or
vice-versa (or it doesn't matter)?
(10:17:52) thl: rdieter, ask first, then review normally
(10:17:56) rdieter: +1 to kmod blocker (it would help finding them at
least)
(10:18:04) thl: +1 to kmod blocker
(10:18:07) scop: +1
(10:18:08) awjb: +1
(10:18:11) c4chris: +1
(10:18:12) bpepple: +1
(10:18:12) tibbs: +1
(10:18:21) abadger1999: +1
(10:18:26) thl: just to be sure:
(10:18:28) rdieter: blocker for "kmod needs approval" or just "kmod" in
general?
(10:18:29) nirik: so whats the status on zaptel kmod then? is it
approved? or rejected because they don't want to submit upstream?
(10:18:43) thl: rdieter, maybe both?
(10:18:57) bpepple: nirik: I'd lean toward reject.
(10:19:02) scop: I don't think we've actually voted on that
(10:19:06) cweyl: nirik: it is GPL.
(10:19:16) thl: nirik, I'd also still lean toward reject
(10:19:22) scop: (nor am I sure if we have defined what determines the
outcome of a vote)
(10:19:23) rdieter: submission to upstream is required, so if that's not
in the plans, then no/reject.
(10:19:28) Rathann: which reminds me, I'll withdraw my acx100 kmod
because I'm told there's some effort to get it upstream
(10:19:34) cweyl: just because it has a commercial fork shouldn't be a
blocker IMHO, mysql does the same thing and it's in fedora
(10:19:45) tibbs: Rathann: that's backwards.
(10:19:47) awjb: rdieter: but why I do not see and never saw the point
in it...
(10:19:50) spot: if they have no plan to put code upstream, i'll vote
against it.
(10:19:55) thl: Rathann, see the  annoucment of the last mm-kernel --
might get dropped
(10:19:59) tibbs: If there's effort to get it upstream then it's
definitely wanted.
(10:20:12) thl: tibbs, +1
(10:20:22) Rathann: tibbs: I just saw a mail in acx100-users saying
something to the contrary
(10:20:23) bpepple: tibbs: agreed.
(10:20:24) rdieter: awjb: if you haven't figured it out by now...
(10:20:33) thl: but saying "we don't want it upstream" is IMHO a no-go
(10:20:38) thl: I don#t want such modules
(10:20:49) cweyl: thl: from a support issue?
(10:20:53) tibbs: Of course, other issues could keep it out of
upstream...
(10:20:54) thl: it might happen that someone tells us that they want it
upstream but it never happens
(10:20:57) thl: but that's life
(10:21:37) thl: cweyl, more a "be a nice open-source citizen" 
(10:21:48) ***jwb is here now
(10:21:55) nirik: ok, so is zaptel-kmod ready for a formal +/- vote? or
do the criteria need more discussion on list or the like?
(10:22:07) thl: nirik, did the zaptel guys ever get back in the
bugreport
(10:22:17) thl: we should tell them why we don#t want there module
(10:22:21) spot: i -1 zaptel if they're not even trying to take the code
upstream.
(10:22:22) thl: maybe they change there ming
(10:22:23) nirik: They just posted the one thing on the bug... 
(10:22:31) scop: if the consensus is that "we don't want it upstream" is
a blocker, then it needs to be documented
(10:22:32) nirik: I can ping them/the bug again 
(10:22:36) Rathann: tibbs:
http://sourceforge.net/mailarchive/message.php?msg_id=36862141
(10:22:39) awjb: rdieter: maybe I am to stupid... but for me this is a
strange view on kmods ... only thing I can think of is gpl issues but
thats about it... nothing from the technical point that I could find
against it
(10:22:40) thl: nirik, please do so
(10:23:04) thl: and let's get back to zaptel later
(10:23:08) nirik: ok, let me do that... and bring up results next week?
(10:23:10) thl: that okay for everybody?
(10:23:13) spot: sure
(10:23:18) thl: nirik, if possible: yes
(10:23:20) bpepple: works for me.
(10:23:25) c4chris: thl, yup
(10:23:31) nirik: who is going to make the FE_KMOD / FE_KMOD_NEEDS_FESCO
blockers?
(10:23:38) thl: spot, and yes, we should document it
(10:23:43) thl: s/spot/scop/
(10:23:50) scop: but I don't agree with that
(10:24:09) thl: spot, then let's discuss it on the list during the next
week
(10:24:18) thl: that's probably better 
(10:24:21) thl: s/spot/scop/
(10:24:23) thl: :-/
(10:24:25) c4chris: scop, "that" = ?
(10:24:27) ***spot scratches his head
(10:24:36) thl: sorry spot
(10:24:44) scop: "that" == "we don't want it upstream is a blocker"
(10:24:44) rdieter: maybe we clarify the policy on kmods in the wiki, to
help out folks understand the "why" (like awjb)?
(10:24:56) c4chris: scop, ah, ok
(10:25:00) thl: rdieter, exactly
(10:25:05) abadger1999: awjb: The argument is that it's very hard to
maintain kmods outside of the kernel due to the fast moving nature of
the kernel-module ABI.
(10:25:07) c4chris: scop, and how about the blocker bugs ?
(10:25:15) awjb: scop: i don't agree with this either ...
(10:25:36) warren: back
(10:25:39) ***scop is confused ;)
(10:25:47) thl: we all are a bit confused now
(10:25:48) scop: blocker bugs are definitely good
(10:25:53) thl: let's discuss it on the list
(10:26:03) awjb: abadger1999: thought someone would bring this up... but
that the nice thing about os even people who don't work on the kernel
can read changelogs and read lkml
(10:26:03) c4chris: thl, yes
(10:26:10) scop: before moving on:
(10:26:13) thl: warren, can you create two new tracker bug assiged to
nobody fp org?
(10:26:27) thl: FE_KMOD and FE_KMOD_NEEDS_APPROVAL ?
(10:26:35) nirik: I can go thru and add them to all the pending kmod
reviews after they are created. 
(10:26:49) scop: when something comes up for a vote, how many votes /
what percentage do we need for something to pass?
(10:26:54) nirik: unless someone else wants to. ;) 
(10:26:57) thl: warren, or can you tell me the pw for nobody fp org ?
(10:27:03) warren: oh
(10:27:03) warren: ok
(10:27:12) cweyl: scop: I think it's the "if no screams..." rule :)
(10:27:13) ***warren finds it...
(10:27:32) scop: 100%?
(10:27:34) scop: eek
(10:27:35) thl: cweyl, normally yes
(10:27:53) jwb: scop, i think it's majority vote
(10:27:58) thl: but sooner or later we have to make decisions where
somebody yells
(10:28:09) thl: then at least 51% of the members should attend
(10:28:09) c4chris: scop, I'd think 51% of the quorum...
(10:28:12) scop: yes, and we need to define the rules *beforehand*
(10:28:22) thl: and at least 51% of those have to vote +1
(10:28:23) warren: thl, sent you mail
(10:28:26) thl: warren, thx
(10:28:26) rdieter: thl: voice vote first (ie, no one screams), else
majority (of those present).
(10:28:27) abadger1999: awjb: A second argument is that the kernel
maintainers don't want to wade through kernel bug reports where external
kernel modules are involved.
(10:28:34) thl: rdieter, exactly
(10:28:39) jwb: abadger1999, that's a weak argument
(10:28:40) abadger1999: awjb: There may be other arguments as well.
(10:28:41) bpepple: rdieter: +1
(10:28:47) jwb: abadger1999, they should already be screening for that
(10:28:54) abadger1999: awjb: I've lost track :-)
(10:28:56) cweyl: abadger1999: didn't FPB settle that with "kmods are
ok"?
(10:29:03) jwb: cweyl, yes
(10:29:05) awjb: abadger1999: ^^ lets take it to the list
(10:29:09) warren: Don't they already have "tainted" messages?
(10:29:18) jwb: warren, not sure
(10:29:23) spot: warren: yeah, but nothing we approve should taint.
(10:29:24) cweyl: I think we're past the do we or don't we decision, we
have that.  we're at the implementation details stage, no?
(10:29:34) jwb: cweyl, yes
(10:29:34) rdieter: kernel maintainers are likely to insist bugs be
reproduced without extra kmods.
(10:29:42) thl has changed the topic to: FESCO Meeting in progress --
M{ae}ss-Rebuild
(10:29:45) jwb: rdieter, that is fine
(10:29:49) thl: I'd like to move on now
(10:30:01) thl: let's stop talking about kmods for today ;-)
(10:30:06) jwb: but they
(10:30:09) jwb: 're so fun
(10:30:10) jwb: ;)
(10:30:18) thl: :)
(10:30:19) rdieter: kmod, kmod, kmod, kmod!  OK, out of my system.
(10:30:22) spot: jwb: don't make me get Axel. ;)
(10:30:24) warren: spot, yes they should
(10:30:30) thl: scop, status update for M{ae}ss-Rebuild?
(10:30:37) thl: anything we need to discuss`
(10:30:39) jwb: warren, no they shouldn't.  and off to the list with
that
(10:30:41) warren: spot, our modules aren't gpg signed
(10:30:42) scop: see wiki
(10:30:47) spot: warren: ah.
(10:30:47) scop: another push coming up tonight
(10:30:57) scop: and status update after it
(10:31:14) rdieter: fyi, saved libpqxx/koffice from being dropped
yesterday. (:
(10:31:33) cweyl: scop: if you've flagged something as "orphaned due to
no rebuild", all it takes to get it back in the repos is to claim &
rebuild, right?
(10:31:41) ***nirik is trying to get fontforge happy, but it's not
building right in devel. ;( 
(10:31:54) jwb: cweyl, yes
(10:31:54) c4chris: scop, good job, I'd liek to say
(10:32:01) awjb: rdieter: yes thanks again for it... wonder why libpqxx
was bitching around so much...
(10:32:02) jwb: yes, good job scop
(10:32:12) thl: scop, yeah, thx for all your work
(10:32:15) scop: thanks
(10:32:33) thl: anything else regarding the mass-build?
(10:32:40) scop: but IMHO grabbing something from the orphaned page
should go through the normal unorphan process
(10:32:43) rdieter: awjb: a combination of -fvisibility generating
warnings, and -Werror
(10:32:45) Rathann: do I have to wait for a reply from former
maintainers if I want to take something over?
(10:33:06) scop: except perhaps if the old maintainer of a package grabs
it due to having missed the deadline by a few days
(10:33:11) thl: scop, well, I think we don't have time to wait a whole
week 
(10:33:12) awjb: rdieter: I will look at it later... still have 3
packages to go \o/
(10:33:17) cweyl: Rathann: they've had several nags, and weeks to
rebuild their packages
(10:33:22) Rathann: ok
(10:33:31) cweyl: if they were going to reply, I think they would have
done so by now :)
(10:33:32) thl: scop, well, people also should check if the maintainer
is still around
(10:33:32) scop: thl, no problem, my opinion is a humble one  ;)
(10:33:36) Rathann: I'll rebuild the two I want tonight
(10:33:42) thl: if he is -> talk to him
(10:33:48) cweyl: thl: of course
(10:33:56) Rathann: I Cc'd them in my mail to -extras
(10:34:03) cweyl: and then berate them soundly for not rebuilding ;)
(10:34:06) thl: that's why I send out the "maitainers that didn#t
rebuild a single package report"
(10:34:32) cweyl: thl: I liked that.  helped differentiate packages that
just "slipped" from people who may well have vanished
(10:34:41) scop: I plan to list packages that are broken due to some of
their deps being dropped tonight/tomorrow
(10:35:06) scop: then maybe drop them early next week?
(10:35:22) thl: scop, yeah, might be okay
(10:35:26) thl: send that to the list
(10:35:26) cweyl: is there a way to take packages and assign them to
sigs?  (after talking to that sig, of course)
(10:35:35) thl: if to many people yell wait a bit longer
(10:35:47) scop: ok
(10:35:51) Rathann: should I leave the former maintainer's email in the
last field in owners.list?
(10:35:54) scop: also, comps.xml needs to be cleaned up from dropped
packages, planning to do that in the weekend
(10:36:06) rdieter: Rathann: imo, yeah.
(10:36:10) cweyl: e.g. there are a number of perl modules orphaned up
there...  that sig may want to take them over en-mass.
(10:36:12) Rathann: will do
(10:36:12) scop: Rathann, I think that makes sense if there was no
response
(10:36:25) thl: c4chris, might be a good idea to check for that
(comps.xml needs to be cleaned up from dropped packages) in your scripts
as well
(10:36:25) xover left the room (quit: Read error: 110 (Connection timed
out)).
(10:36:39) c4chris: thl, it does in principle
(10:36:47) thl: ohh? nice :-))
(10:37:04) c4chris: there's a list near the bottom
(10:37:09) ***thl will jump to the next item on the schedule soon
(10:37:19) c4chris: (I think I like your plitting idea more and more...)
(10:37:34) c4chris: splitting
(10:37:40) thl has changed the topic to: FESCO Meeting in progress --
Enterprise Extras
(10:37:48) thl: c4chris, no need to hurry with that
(10:37:53) thl: so, EPEL now
(10:38:00) thl: create a public mailinglist -- what name?
(10:38:06) thl: fedora-extras-el-list ?
(10:38:19) quaid [n=quaid] entered the room.
(10:38:29) jwb: that sounds fine with me
(10:38:36) scop: no fedora in "epel", no fedora in list name?
(10:38:37) thl: or can we/do we want to live without a special
mailinglist for now
(10:38:37) cweyl: do we need a separate list yet?
(10:38:38) errr: is it possible for me to package for fedora 4,5,6 all
from one system (running fc5) its getting kind of old to fire off vmware
to build for fc4 and 6
(10:38:43) ***nirik groans at yet another list. 
(10:38:51) cweyl: thl: I'd say hash it out on extras-list first
(10:39:02) c4chris: extras for now I'd say
(10:39:15) ***mmcgrath doesn't care what the list is called.
(10:39:24) jwb: errr, yes use mock
(10:39:26) thl: cweyl, maybe; other requested a sepecial mailing list
already "because the traffic on f-e-l is so high"
(10:39:31) errr: thanks jwb 
(10:39:45) c4chris: thl, ah, right
(10:39:55) cweyl: thl: yah...  but I'd think something like this you'd
want a lot of eyeballs on it to start with :)
(10:40:05) rdieter: cweyl: +1
(10:40:14) thl: well, then let's use f-e-l until people complain :)
(10:40:27) mmcgrath: thats fine.
(10:40:35) c4chris: thl, yup.  20 msg a day is not so high in my book...
(10:40:35) thl: mmcgrath, spot , what about RHEL4 + updates for the
builders?
(10:40:47) thl: any process?
(10:40:53) RTLM [n=RTLM] entered the room.
(10:40:55) mmcgrath: To my knowledge no progress has been made in that
area as of yet.  
(10:41:01) spot: no one ever told me where to put the files.
(10:41:06) ***spot coughs
(10:41:13) mmcgrath: spot: one sec.
(10:41:45) mmcgrath: spot: somewhere on
ntap-fedora1.fedora.phx.redhat.com?
(10:41:57) spot: mmcgrath: email me and tell me exactly where.
(10:42:08) thl: mmcgrath, spot, thx
(10:42:15) mmcgrath: K, I might have to check with Stacy about some of
it...
(10:42:17) thl: mmcgrath, can you prepare the builder after that?
(10:42:29) mmcgrath: Yeah, myself or dgilmore.
(10:43:05) thl: mmcgrath, just go ahead then when you have everything
together; 
(10:43:23) thl: did anybody read
https://www.redhat.com/archives/fedora-extras-list/2006-September/msg00672.html ?
(10:43:36) thl: Am I just stupid? is that much to complicated?
(10:43:54) rdieter: +1 too complicated (and I said as much on the list
too)
(10:44:11) abadger1999: smooge definitely went overboard there.
(10:44:21) jwb: i ignored it for now
(10:44:26) c4chris: I'd like to start real simple
(10:44:38) rdieter: c4chris: +1
(10:44:39) abadger1999: At least for a first (or second) iteration.
(10:44:44) thl: abadger1999, that's not smooge's mail
(10:44:46) ***spot is unsure why we're not just doing el4, el5, devel
(10:44:51) c4chris: evolution has good sides...
(10:45:08) jwb: spot, i thought it was just el5, devel?
(10:45:18) ***cweyl likes simple
(10:45:22) thl: el4 also
(10:45:25) thl: just start now
(10:45:29) spot: if so, why am i populating a RHEL 4 builder? :)
(10:45:31) mmcgrath: We can always make it more complicated later :D
(10:45:32) jwb: oh, yes makes sense
(10:45:34) thl: or "soon"
(10:45:37) jwb: el5 isn't out :)
(10:45:39) abadger1999: thl: Oh sorry.
(10:45:59) nirik: since el5 isn't out, wouldn't it be just el4, devel?
(10:46:01) thl: abadger1999, np :)
(10:46:12) c4chris: nirik, yes
(10:46:19) thl: spot, what would devel be in your scheme?
(10:46:32) thl: there is normally now rawhide rhel to build for
(10:46:38) thl: s/now/no/
(10:47:01) spot: devel would be el5 betas
(10:47:11) jwb: are those public?
(10:47:14) spot: yep.
(10:47:29) thl: spot, well, there are noemally no "betas"
(10:47:37) jwb: erm
(10:47:45) thl: only in those two or three months were in now
(10:47:55) spot: thl: devel is composed of the fc6 test releases on the
fedora side
(10:48:10) spot: devel is composed of the el5 beta releases on the el
side
(10:48:15) nirik: I would think the latest publicly released beta... 
(10:48:16) thl: why not just crate a el5 branch now
(10:48:20) thl: and build for the betas?
(10:48:28) ***spot is fine with either
(10:48:33) jwb: i like that better
(10:48:34) thl: s/build/& there/
(10:48:54) c4chris: thl, fine too
(10:49:08) thl: k
(10:49:27) thl: another thing: where shall we upload the stuff to?
(10:49:27) c4chris: avoids branching later...
(10:49:34) thl:
http://download.fedora.redhat.com/pub/fedora/linux/extras/{el4,el5} ?
(10:49:44) thl:
http://download.fedora.redhat.com/pub/fedora/linux/extras/el/{45} ?
(10:50:11) mmcgrath: jbw:
http://mirrors.kernel.org/redhat/redhat/linux/beta/RHEL5-Beta1/client/i386/
(10:50:41) nirik: oh yeah, should extras just be one tree, or a
client/server branches?
(10:50:48) c4chris: extras/{el4,el5}
(10:51:07) mmcgrath: Is it the opinion of most people here that this is
a branch or a fork?
(10:51:10) spot: one tree seems reasonable.
(10:51:21) thl: one tree +1
(10:51:28) ***jwb regrettably has to go
(10:51:31) thl: mmcgrath, well, it's a branch afaics
(10:51:32) c4chris: one tree +1
(10:51:40) c4chris: branch
(10:51:47) jwb: there can be only one
(10:51:52) jwb: +1
(10:52:03) mmcgrath: my concern is that people will think its a fedora
branch.
(10:52:07) rdieter: branch/one_tree: +1
(10:52:19) cweyl: this rabble likes branch/one_tree...
(10:52:42) mmcgrath: IE they go to the fedora site, click on fedora
mirrors, click on extras and start downloading.
(10:52:49) rdieter: mmcgrath: who cares?  Or are you just concerned
about perception?
(10:53:20) thl: mmcgrath, we can move it somewhere else later if we
really want to
(10:53:20) tibbs: I would hope that they could go to a fedora site,
click around and start downloading.
(10:54:10) thl: we're slowing down... shall we stop on this topic for
today?
(10:54:15) mmcgrath: I still believe this is a fork.   After all, none
of this software will end up on any fedora machines, ever.
(10:54:25) mmcgrath: Yeah.  we can go on.  This discussion would
probably be better on the list.
(10:54:31) bpepple: thl: +1
(10:54:37) c4chris: thl, +1
(10:54:42) thl has changed the topic to: FESCO Meeting in progress --
Packaging Committee Report
(10:54:54) thl: scop, spot, rdieter, tibbs, abadger1999, ... ?
(10:54:59) tibbs: Again, not much to report.
(10:55:09) spot: we didn't decide anything other than that we will be
taking open nominations for new seats on the committee
(10:55:29) tibbs: Not that there are currently any open seats...
(10:55:30) spot: those nominations will be voted on by the existing
members to determine new membership
(10:55:49) spot: Yep. Right now, there are no open seats.
(10:56:27) thl: k, anything else?
(10:56:29) spot: nope.
(10:56:56) thl has changed the topic to: FESCO Meeting in progress --
MISC -- remove inactive extras contributors from cvsextras
(10:57:06) thl:
https://www.redhat.com/archives/fedora-extras-list/2006-September/msg00612.html
(10:57:09) thl: remove them?
(10:57:28) bpepple: thl: +1
(10:57:29) c4chris: sure
(10:57:43) cweyl: should we keep a list of people removed for inactivity
on the wiki somewhere?
(10:57:44) thl: sopwith or other that might need access can be readded
later easily
(10:57:52) thl: cweyl, yes
(10:57:55) cweyl: might come in handy for those who later decide they
want back in :)
(10:58:02) awjb: cweyl, +1
(10:58:16) thl: I'll probably also will remove all those taht have the
status "declined"
(10:58:25) scop: I suppose such a page should have an ACL on it
(10:58:31) thl: (most of them applied for cvsextras accidentally)
(10:58:38) xover [n=link] entered the room.
(10:58:39) thl: scop, well, maybe
(10:58:48) thl: scop, but we can always look into the history
(10:58:54) thl: isn't that enough?
(10:59:03) nirik: could you just add a 'inactive' status? then they
could be readded and you could see that they were once there?
(10:59:04) c4chris: thl, watch it like the schedule...
(10:59:07) ***thl is not sure himselfs
(10:59:18) ***scop loathes wiki's history interface
(10:59:30) thl: well, I'll look at it
(10:59:31) abadger1999: One or two of the inactivity people are active.
(10:59:39) c4chris: easier to search in the ml archive
(10:59:40) thl: maybe ACls are the best for this job
(10:59:51) thl: abadger1999, I'll recreate the list
(10:59:55) abadger1999: jspaleta, for instance has been having someone
else do commits for him because of his internet situation.
(11:00:01) thl: abadger1999, and check it before i actually remove those
poeple
(11:00:10) thl: abadger1999, k, noted
(11:00:17) green_ left the room (quit: Read error: 110 (Connection timed
out)).
(11:00:26) scop: thl, did you take owners.list into account when looking
for inactive people?
(11:00:42) thl: scop, no
(11:00:59) scop: ok
(11:01:02) Tjikkun left the room (quit: Read error: 113 (No route to
host)).
(11:01:03) thl: scop, I did only that rough check that's described in
the mail
(11:01:14) thl: I'll take a closer look
(11:01:30) scop: actually, I think your method should be ok
(11:01:31) thl: and as I said: people that got removed accidentally can
be readded later easily
(11:01:50) scop: those folks' bugzilla account won't go anywhere anyway
(11:01:55) ***thl considers this settled
(11:01:59) warren left the room ("Leaving").
(11:02:00) abadger1999: Would htey have to resign cla and so on?
(11:02:01) cweyl: so long as there's a record to establish "yes, you
were removed accidentally", I like it
(11:02:06) Tjikkun [n=tjikkun] entered the room.
(11:02:08) thl: abadger1999, no
(11:02:22) abadger1999: Sounds easily reversible then.
(11:02:24) thl: another group in the ac handles the cla status
(11:02:37) thl: that "cladone" iirc
(11:02:39) nirik: they would likely just have to re-request cvsextras...
(11:02:48) thl: nirik, yes, I think so
(11:02:59) thl: getting late -- moving on
(11:03:12) thl has changed the topic to: FESCO Meeting in progress --
MISC -- enhance AWOL
(11:03:19) thl: did anybody look into it
(11:03:24) ***thl did not :-/
(11:03:38) ***c4chris neither
(11:04:15) thl: k, then let's skip this and look at it in  the next
meeting
(11:04:26) thl has changed the topic to: FESCO Meeting in progress --
free discussion around extrs
(11:04:34) thl: anything else we should discuss?
(11:04:48) cweyl: thl: I'd like to add a "current-branches" file to
common
(11:04:49) c4chris: you wanted the competition packaging...
(11:05:02) warren [n=warren] entered the room.
(11:05:04) warren: oops
(11:05:17) cweyl: so scripties/people can automagically determine what's
a current fedora release...
(11:05:23) thl has changed the topic to: FESCO Meeting in progress --
competition packaging (aka tor)
(11:05:34) thl: well, what do we want to do about it
(11:05:47) thl: wait another week and hope that it all works out
somehow?
(11:05:51) abadger1999: I don't really like it.
(11:06:12) thl: abadger1999, "it" = competition packaging | wait another
week" ?
(11:06:19) abadger1999: competition packaging.
(11:06:22) c4chris: thl, yea maybe work out a proposal first...
(11:06:25) ***scop is off to take care of today's push+massrebuild
things
(11:06:34) thl: abadger1999, well, I agree in general
(11:06:38) rdieter: I think I like competition packaging, but I'm not
sure how best to actually do it.
(11:06:45) scop: btw, there's a precedent in the smart package
(11:06:46) tibbs: In this case, though, someone just needs to submit
another package and get it over with.
(11:06:47) c4chris: scop, may the force be with you :-)
(11:06:58) thl: abadger1999, but now and then "competition packaging"
might be the best
(11:07:08) rdieter: to ensure the "best" package wins, that is.
(11:07:22) abadger1999: tibbs: But I think Enrico's packaging choices
may be valid in this case....
(11:07:34) rdieter: best doesn't necessarily equate to fastest review.
(11:07:44) c4chris: abadger1999, so why doesn't his package get
approved ?
(11:07:44) abadger1999: tibbs: it's a poor situation where both choices
are less than ideal.
(11:07:49) ***Rathann looks at ctorrent package he's just taken over
(11:07:58) abadger1999: rdieter: I agree.
(11:08:03) tibbs: Tor isn't really a complicated package, though.
(11:08:07) rdieter: The reviewer disagrees with Enrico's choice(s).  And
neither will back down.
(11:08:23) ***Rathann wonders how on earth did this pass review
(11:08:29) ***nirik also doesn't agree with the tor packageing.... so I
dropped the review and let someone else try. 
(11:08:43) thl: well then let's just note something like this: "normally
packagers should work together in one bug getting a certain software
packaged; but htere is no rule yet that disallows opening another bug
with and alternative package of the same software"
(11:08:50) awjb: bugnumber plz
(11:09:06) rdieter: IMO, if the submitter/reviewer reach an impasse, the
reviewer should step away.
(11:09:08) thl: awjb, 175433
(11:09:11) nirik:
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=175433
(11:09:11) awjb: ty
(11:09:24) c4chris: rdieter, yes
(11:09:30) nirik: rdieter: thats what I did many months ago... extras
still has no tor. 
(11:09:50) c4chris: thl, I find it ok.
(11:09:52) rdieter: nirik: good, imo, you did the right thing.
(11:10:03) abadger1999: nirik: You were close to approving it a couple
times.  Was the split of packages the only remaining issue?
(11:10:16) c4chris: the caveat is that such review requests tend to get
closed/duplicate
(11:10:20) abadger1999: or were there other things that combined to give
you a general bad feeling?
(11:10:29) c4chris: so a clear marking should be used in cases like
this...
(11:10:38) nirik: yes. It's complicated and confusing, and only to make
things nicer for a non fedora env...
(11:11:02) rdieter: nirik: like el4? (:
(11:11:20) bakers [n=bakers] entered the room.
(11:11:21) nirik: no, like ensc's minimal/embeded boxes with no
SysVinit. 
(11:11:55) tibbs: Some of the subpackages depend on things not in
Fedora.
(11:12:12) nirik: tibbs: yeah, he had those commented. 
(11:12:18) poolshark: rdieter: i don't mind stepping down as reviewer,
but then still no tor in fedora...
(11:12:24) c4chris: I think the idea is: do not encourage competition,
but allow it in some cases
(11:12:35) thl: c4chris, +1
(11:12:44) bpepple: c4chris: +1
(11:12:47) nirik: I also had email from the upstream tor that they
didn't like that they would get support emails about the complexity of
the package
(11:13:09) nirik: I think they posted in the bug too. 
(11:13:09) rdieter: poolshark: so either someone else submit a better
one, or someone else reviews the current submission.  it's the only way.
(11:13:22) tibbs: The "tons of subpackages" approach hasn't really
worked well with clamav.
(11:13:43) ***nirik really dislikes the extras clamav package as
well. ;( 
(11:13:45) rdieter: tibbs: mostly because it (the packagine) was buggy
(apparently).
(11:13:46) poolshark: rdieter: one was already submitted, but closed as
duplicate. can it be reopened ?
(11:13:57) thl: poolshark, sure
(11:14:02) rdieter: poolshark: imo, yes, that's what we're saying here.
(11:14:09) poolshark: ty
(11:14:14) c4chris: poolshark, yup and add a small explanation
(11:14:18) nirik: poolshark: does the submitter still want to submit it?
might ping them?
(11:14:31) tibbs: I think the submitter is here.
(11:14:39) abadger1999: ugh.  But what happens if someone starts
reviewing ensc's version again?
(11:14:39) poolshark: nirik: will do (Paul Wouters)
(11:14:45) abadger1999: And seems to agree with him?
(11:14:56) abadger1999: First through the post wins does not seem very
ideal...
(11:15:07) tibbs: No-rules cage match.
(11:15:14) c4chris: abadger1999, ah well... in this case we'll have to
vote here
(11:15:16) nirik: poolshark: might also cc the tor upstream? their rpm
guy wanted to submit, but didn't because it was already there?
(11:15:18) poolshark: after 10 monts, what are the odds ?
(11:15:26) poolshark: nirik: sure
(11:15:47) nirik: poolshark: thanks. 
(11:15:50) rdieter: imo, if upstream wants to maintain it, that would be
ideal.
(11:16:00) abadger1999: Hell, if we just want tor in Extras, I'll review
tor tonight.
(11:16:11) thl: abadger1999, then do that
(11:16:19) thl: poolshark, can you wait one more week?
(11:16:28) abadger1999: I don't mind ensc's split package aproach --
don't know if there's other things I'd dislike though...
(11:16:32) thl: give abadger1999 and ensc some time for it
(11:16:33) poolshark: thl: sure
(11:16:38) nirik: poolshark: Paul Wouters is the upstream rpm
maintainer. 
(11:16:56) thl: poolshark, tia
(11:17:01) thl: k, so let's move on
(11:17:13) abadger1999: For the future, I think we need to figure out
how to evaluate something that's better, technically, though...
(11:17:24) warren: ensc is being unreasonable
(11:17:27) thl has changed the topic to: FESCO Meeting in progress --
"current-branches"
(11:17:40) c4chris: just add the file
(11:17:42) abadger1999: poolshark: Which person are you?  I'll send you
an email if I think the review's going to hang up on something else.
(11:17:49) ensc: warren: I hate such metadiscussions which does not
provide a single argument
(11:18:08) warren: ensc your arguments are based on inconveniencing you
vs. everyone else.
(11:18:31) thl: c4chris, --verbose ?
(11:18:39) warren: That is my only thought.  I don't care about this
package personally.
(11:18:40) ensc: there is no policy regarding it and I do not accept
"everybody else it doing it so"
(11:18:48) ensc: I have good reason not to do it so
(11:18:52) c4chris: thl, just create the current-branches file
(11:19:12) thl: c4chris, for what purpose?
(11:19:16) c4chris: (I thought that was the topic)
(11:19:33) c4chris: <cweyl> thl: I'd like to add a "current-branches"
file to common
(11:19:59) ***thl confused cweyl and c4chris -- sorry
(11:20:03) c4chris: to list which of the branches are current
(11:20:16) c4chris: thl, np
(11:20:38) thl: I still fail to see the purpose for it -- sorry
(11:20:48) ***thl seems to lack context
(11:20:49) warren: "current" meaning what?
(11:20:51) c4chris: cweyl, ?
(11:21:08) thl: cweyl seems to be afk
(11:21:15) c4chris: I think he wanted an automatic way to get the
currently maintained FC branches
(11:21:27) c4chris: currently FC5 and devel
(11:21:28) thl: I think we can wait another week with this
(11:21:36) c4chris: thl, right
(11:21:44) thl has changed the topic to: FESCO Meeting in progress --
anything else?
(11:21:47) bakers: Does anyone here use xchat?
(11:22:08) nirik: bakers: yes, but you might wait to ask questions until
the meeting is over. ;) 
(11:22:08) c4chris: thl, nope
(11:22:09) bakers: I'm trying to script it to auto send my password when
I connect
(11:22:13) ***thl will close the meeting in 45
(11:22:44) thl: bakers, see nirik's comment (or just look into the
freenode faq)
(11:22:52) ***thl will close the meeting in 15
(11:22:58) nirik: bakers: yeah, see the freenode faq. It has
instructions. 
(11:23:09) thl: -- MARK -- Meeting over
(11:23:27) thl: thx everyone
(11:23:31) c4chris: thl, thx
}}}

Attachment: signature.asc
Description: This is a digitally signed message part


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]