17:00:30 <abadger1999> #startmeeting fpc
17:00:30 <zodbot> Meeting started Wed Mar  6 17:00:30 2013 UTC.  The chair is abadger1999. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:30 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:42 <abadger1999> #topic Roll Call
17:00:49 <abadger1999> Who's here?
17:01:04 * geppetto is here
17:01:45 <abadger1999> spot, limburgher, tibbs, rdieter, SmootherFrOgZ: you all present for an fpc meeting?
17:01:53 <limburgher> More or less.
17:03:34 <abadger1999> If tibbs|w is here that would be four... but still shy of quorum.
17:03:42 <tibbs|w> Yeah, I'm around.
17:03:50 <tibbs|w> Had to be in the office today.
17:03:55 <abadger1999> <nod>
17:04:18 * SmootherFrOgZ is around.
17:04:32 <abadger1999> Cool.  That's 5.  Just enough to pass items :-)
17:04:37 <limburgher> w00t
17:05:13 <tibbs|w> Assuming we could agree, of course.
17:05:29 <limburgher> Pfft.  You're so demanding.
17:06:19 <abadger1999> #topic How long should systemd conversion triggers be kept: https://fedorahosted.org/fpc/ticket/262
17:06:44 <tibbs|w> People appeared to be confused by this.
17:06:58 <abadger1999> I think tibbs's proposal matches what we do elsewhere.
17:07:04 <limburgher> It doesn't look overly complex, modulo release schedule issues. ;)
17:07:16 <tibbs|w> I asked a packager why their triggers are still there when they converted in F15, and they responded that they weren't sure when they were allowed to remove it.
17:07:30 <abadger1999> <nod>
17:07:46 <SmootherFrOgZ> sound good to me
17:07:55 <limburgher> It's sort of like Obsolete/Provides, can can be sunset in a similar way, and this spells it out.
17:08:08 <geppetto> tibbs: One thing to keep in mind is that a bunch of packages might want to keep them for RHEL-6 => RHEL-7 upgrades.
17:08:12 <abadger1999> I guess it doesn't break things to keep it longer.  But making the spec files less complex and filled with old cruft is a good thing.
17:08:34 <tibbs|w> RHEL isn't really my concern.
17:08:36 <limburgher> geppetto:  Is that our problem?
17:08:42 <abadger1999> geppetto: I thought I was told we don't support upgrades between RHEL releases?
17:08:49 <tibbs|w> But I wasn't aware that RHEL even supported any form of upgrading between versions.
17:08:56 <geppetto> abadger1999: Depends what day of the week it is.
17:09:06 <abadger1999> ah, of course :-(
17:09:09 <tibbs|w> I still haven't figured out what that trigger actually gets us, anyway.
17:09:51 <geppetto> limburgher: I don't see the benefit to ACK policy that actively harms RHEL.
17:10:26 <tibbs|w> How can saying "you can remove the trigger if you want after X releases" actively harm something we shouldn't be concerned about anyway?
17:10:29 <limburgher> geppetto:  Does it?  If they support that upgrade path, can't they add the trigger back in?
17:10:36 <abadger1999> geppetto: noe that the change will only allow packagers to remove those pieces.  It doesn't force them to.
17:10:41 <abadger1999> *note that
17:11:06 <geppetto> Yeh
17:11:24 <geppetto> To be fair, I don't mind the proposal as is.
17:11:33 <abadger1999> Cool.  Let's vote then.
17:11:36 <abadger1999> +1
17:11:41 <geppetto> +1
17:11:42 <limburgher> +1
17:11:47 <tibbs|w> +1
17:12:03 <SmootherFrOgZ> +1
17:13:29 <abadger1999> #info update to how long systemd conversion triggers are required t be kept passed (+1: 5, 0:0, -1:0)
17:14:01 <abadger1999> #topic: exception for nmap to bundle lua: https://fedorahosted.org/fpc/ticket/255
17:14:11 <abadger1999> there was some discussion i nthe bug but no new votes.
17:14:26 <abadger1999> tibbs|w: was my last comment sufficient to explain what I was thinking there?
17:14:48 <tibbs|w> Pretty much.
17:14:51 <abadger1999> Cool.
17:15:06 <tibbs|w> I mean, they should try to take an active role, but I didn't think the onus fell on them to actually do the work.
17:15:27 <tibbs|w> I guess this comes down to the general question of what we do when our "temporary exemption" expires and nobody has done anything.
17:15:46 <tibbs|w> Fortunately we've not yet had to deal with that.
17:15:48 <abadger1999> yeah.
17:16:27 * abadger1999 notes that he needs to file a ticket about a temporary samba exception that is still being used but that's for another time....
17:16:51 <tibbs|w> I can't even remember that, but I'm sure there's at least one that's fallen through the cracks.
17:17:36 <abadger1999> So SmootherFrOgZ, we could use your vote on that ticket. limburgher, if you change your mind about being +0, you could update it too.
17:17:56 <SmootherFrOgZ> is there so much pain to update lua in fedora?
17:18:08 <abadger1999> We don't have anyone else here today who hasn't already voted. so we're unlikely to get to quorum on this.
17:18:19 <abadger1999> SmootherFrOgZ: The lua maintainers seem to have just dragged their feet.
17:18:39 <abadger1999> SmootherFrOgZ: and there are about 80 packages that require lua (iirc) so it needs some testing post-update.
17:18:46 <tibbs|w> Yes, the problem here is with how lua is being maintained.
17:18:53 <geppetto> Well, AIUI it's a big change and they wanted to make sure the core apps. would work with it first (like rpm).
17:18:58 <SmootherFrOgZ> so it's more a time's issue
17:19:04 <SmootherFrOgZ> and man power
17:19:05 <limburgher> I'm with tibbs|w
17:19:21 <tibbs|w> But rpm isn't the concern here since it's been compatible for some time.
17:19:56 <tibbs|w> In the context of this ticket, the best solution appears to be to let the base lua package move forward and add lua51 for compatibility.
17:20:08 <limburgher> That would be my preference.
17:20:23 <limburgher> It's just that something needs to occur to get that to happen.
17:20:23 <geppetto> yeh
17:20:26 <tibbs|w> But, again, can't do anything about it unless the lua maintainers do something.
17:20:46 <abadger1999> yep.
17:20:49 <SmootherFrOgZ> tibbs|w: yeah :/
17:20:52 <limburgher> Yeah, it would be nice if it was short of the full-on NRM process, but there you are.
17:21:10 <abadger1999> okay, deferring ticket -- please vote in the fpc ticket if you haven't already.
17:21:30 <SmootherFrOgZ> so i'm +1 for temporary exception.
17:21:40 * SmootherFrOgZ will do
17:21:44 <abadger1999> #info please vote on nmap temporary bundling exception if you haven't yet: https://fedorahosted.org/fpc/ticket/255
17:22:01 <abadger1999> #topic  Update to the filtering guidelines: https://fedorahosted.org/fpc/ticket/189
17:22:03 <tibbs|w> We should also ping on the open bug.
17:22:39 <abadger1999> I'll add a note to ping the lua bug once we have votes on the temporary bundling exception.
17:22:59 <abadger1999> This has been sitting for a while.  The new draft is here: https://fedoraproject.org/wiki/User:Toshio/AutoProvidesAndRequiresFilteringDraft
17:23:10 <limburgher> I don't see why we can't ping it now. . .
17:23:54 <abadger1999> limburgher: Sure.  I'll ping it after the meeting.
17:23:56 <limburgher> abadger1999:  Is there a diff against current?  My wiki-fu is weak.
17:24:13 <limburgher> abadger1999:  Cool.  If lua gets updates in the mean time, the point is moot.
17:24:30 <abadger1999> limburgher: History tab: https://fedoraproject.org/w/index.php?title=User%3AToshio%2FAutoProvidesAndRequiresFilteringDraft&diff=325338&oldid=324523
17:24:46 * limburgher goggles at the wow-ness
17:25:02 <abadger1999> limburgher: This is a pretty big update though -- rpm added a lot of support for filtering.
17:25:03 <tibbs|w> It's pretty much a rewrite.
17:25:43 <limburgher> Yeah, I was coming to that realization.
17:26:09 <tibbs|w> The fallout from the new RPM macro substitution stuff is interesting.
17:27:03 <abadger1999> tibbs|w: What was that part?
17:27:45 <tibbs|w> There was a link to an email on the rpm-dev list.
17:28:00 <tibbs|w> http://lists.rpm.org/pipermail/rpm-list/2013-January/001359.html
17:28:14 <abadger1999> oh yeah.
17:28:19 <abadger1999> Yes that is interesting.
17:28:46 <abadger1999> Lots of ways that people have mistyped a ternary expression.
17:29:50 <limburgher> Upper primates are an endless source of entertainment.
17:30:49 <tibbs|w> So,I think we need this, and this draft seems reasonable.
17:30:50 <geppetto> abadger1999: The warning about "this one will only use the regex defined last" needs to be more obvious IMO.
17:31:08 <tibbs|w> But I haven't gone all the way through it.
17:31:28 <geppetto> Also maybe a warning that %{blah} is going to be passed as plain text to the regex engine (so anything "insecure" in it needs to be escaped)
17:31:52 <limburgher> Yeah, overall I'm happy.
17:32:00 <abadger1999> geppetto: how about an {{admon/warning}} at the top of Usage?
17:32:19 <geppetto> abadger1999: Yeh, that's fine.
17:32:33 <abadger1999> geppetto: do you have some wording for that second warning?
17:33:12 <geppetto> uh …
17:34:56 <geppetto> "Warning the regex engine is only passed the final string, after rpm macro expansion, so you can't use unescaped data via. rpm macros."
17:35:14 <geppetto> feel free to point and laugh :)
17:36:40 <geppetto> Anyway … +1
17:37:58 <geppetto> abadger1999: On that note … %global _privatelibs libprivate.so.* … should be %global _privatelibs libprivate[.]so.* … :)
17:39:15 <geppetto> dito. the __provides_exclude example
17:40:13 <geppetto> And __provides_filter_from is missing the 2nd \ (maybe just use the [.] form again?)
17:40:52 <geppetto> Anyway … +1
17:43:07 <abadger1999> https://fedoraproject.org/wiki/User:Toshio/AutoProvidesAndRequiresFilteringDraft#Usage <= first warning added
17:43:18 <abadger1999> geppetto: Is this what you meant for the second warning? https://fedoraproject.org/wiki/User:Toshio/AutoProvidesAndRequiresFilteringDraft#Regular_Expression_Variant
17:43:29 * abadger1999 updates the exaples
17:45:35 <tibbs|w> Folks, I have an emergency so I'm going to be away now.
17:45:39 <tibbs|w> I'll vote in the ticket later.
17:45:52 <abadger1999> tibbs|w: thanks
17:45:54 <geppetto> abadger1999: Kind of, mainly to point out that %define foo_bar foo.bar … will have the unescaped . in the final string too.
17:47:18 <geppetto> abadger1999: __provides_exclude here: https://fedoraproject.org/wiki/User:Toshio/AutoProvidesAndRequiresFilteringDraft#Regular_Expression_Variant
17:47:36 <geppetto> then all the examples are good, I think :)
17:50:55 <limburgher> So I suppose that's pretty much a wrap then?
17:51:28 <abadger1999> geppetto: okay, Take a look.
17:51:37 <abadger1999> limburgher: yeah, we've lost quorum.
17:51:44 <geppetto> ok, +1
17:51:49 <abadger1999> +1
17:52:10 <abadger1999> let's vote here, I'll record where we stand in the ticket and we can finish voting in the ticket this week.
17:52:18 <limburgher> +
17:52:19 <SmootherFrOgZ> +1
17:52:20 <limburgher> 1
17:52:26 <limburgher> +1 (for clarify)
17:52:29 <abadger1999> and then I have one last things to bring up in open floor.
17:52:36 <limburgher> which is like clarity.
17:52:43 <abadger1999> Okay, +4 which is everyone that's still present
17:53:07 <abadger1999> #info Current vote on filtering update is (+1: 4, 0:0, -1:0)  will continue vote in ticket.
17:53:11 <abadger1999> #topic Open floor
17:53:15 <Viking-Ice> has the cron stuff been dealt with ?
17:53:19 * Viking-Ice arrived late
17:53:22 <abadger1999> So... we're getting behind in the ticket queue.
17:53:44 <abadger1999> Viking-Ice: We've just lost quorum (someone had to leave for a work emergency) so we won't get to vote on that today.
17:53:57 <abadger1999> But we could discuss it as an open floor item in just a minute.
17:54:18 <abadger1999> My open floor item -- we're getting behind in the ticket queue.  I sent an email asking what we'd like to do about that to the list.
17:54:32 <Viking-Ice> abadger1999, really I have to wait another week to start working on that
17:54:34 <abadger1999> please reply if you have a preference...
17:55:39 <limburgher> abadger1999:  Which list?  I can't find it. :(
17:55:44 <abadger1999> hmm... actually that might have been a repyly to a devel list post or something.
17:55:52 <abadger1999> I'll dig it out and forward to the packaging list
17:55:57 <abadger1999> then you can reply there :-)
17:56:06 <limburgher> abadger1999:  Perfect.
17:56:27 <abadger1999> Viking-Ice: Would you like to discuss the cron ticket a bit?  We might be able to vote during the week.
17:56:30 <abadger1999> (on the ticket)
17:57:14 <Viking-Ice> well we just need to decide a hard dependency on cronie for packages containing cron jobs or use virtual provide instead
17:58:02 <abadger1999> Okay... if it's just hard dep vs virtual provide, I'd vote for a virtual provide.  Does cronie already provide one?
17:58:05 <geppetto> Viking-Ice: I think a generic provide is going to be much better, esp. as we've already changed it once and systemd is likely to subsume it eventually.
17:58:06 <Viking-Ice> I personally dont see the point in using virtual provide since we only ship one cron binary daemon ( cronie ) not vixi-cron or dcron
17:58:41 <Viking-Ice> geppetto, systemd will only subsume partial of those jobs
17:58:56 <geppetto> Uh huh.
17:59:17 <geppetto> Which is to say … I have $5 that says otherwise.
17:59:36 <Viking-Ice> geppetto, only for components ( if relevant ) that already have a dependency on systemd
18:00:08 <Viking-Ice> geppetto, draft of my page is here https://fedoraproject.org/wiki/User:Johannbg/Systemd/cron-migration#Cronto_Systemd_Timer_migration
18:00:45 <Viking-Ice> that page contains an list that *might* be migrated to timer units
18:01:21 <geppetto> Uh huh.
18:03:26 <abadger1999> If we use a virtual provide we need to know what functionality that represents.
18:03:44 <Viking-Ice> yup
18:03:52 <abadger1999> /etc/cron.d is processed?  /etc/cron.daily?  provides crontab command?
18:05:44 <geppetto> One provide for the first two, I'd assume.
18:06:16 <abadger1999> <nod>
18:06:16 <geppetto> If they need access to the crontab command etc. … then I think it' valid that they depend on cronie, at least until systemd ships something to replace that too.
18:06:31 <Viking-Ice> again systemd will not be doing that
18:06:42 <geppetto> sure, cool story.
18:06:51 <Viking-Ice> systemd will not be venturing into users cron
18:07:07 <geppetto> anyway...
18:07:25 <Viking-Ice> geppetto, we decided that in brno and I have not been informed that that has been changed
18:07:43 <limburgher> eval(we)?
18:08:19 <Viking-Ice> limburgher, everyone present at the systemd hackfest systemd fad day
18:08:25 <geppetto> it doesn't matter … even if it turns out we'll keep two cron systems forever, it works the same way.
18:08:54 <limburgher> Viking-Ice: which, since I wasn't there, doesn't help me much.
18:09:29 <limburgher> Viking-Ice: Is there something posted somewhere where I can read about that hackfest?
18:10:35 <Viking-Ice> limburgher, kay,lennart,harald,myself the brno systemd guys Tom, Zbigniew and more
18:10:50 <limburgher> Viking-Ice: <nods>, thanks.
18:10:58 <Viking-Ice> Lennart was going to post something I believe dont think he has done so yet
18:11:50 <Viking-Ice> both days where broadcasted via hangout live on G+ perhaps there are recordings of those days on the tube
18:13:04 <limburgher> Anyway, back on topic (sorry). . .to abadger1999's question on functionality, what do we know?
18:16:55 <geppetto> limburgher: On the provide? Do you disagree with what I wrote? Not sure about what the name should be, but should be fairly easy to come up with something before the next meeting when we can vote on a real proposal.
18:17:31 <limburgher> geppetto:  No, not at all, just wondering if there were additional thoughts.
18:17:39 * geppetto nods
18:18:01 <Viking-Ice> afaikt only crontab actually creates those directories http://pkgs.fedoraproject.org/cgit/crontabs.git/tree/crontabs.spec, and cronie requires http://pkgs.fedoraproject.org/cgit/cronie.git/tree/cronie.spec while fcron does not require crontab nor cronie http://pkgs.fedoraproject.org/cgit/fcron.git/tree/fcron.spec
18:18:15 <Viking-Ice> cronie requires crontabs
18:19:15 <Viking-Ice> so it might be enough to have those packages just require crontabs and fix fcron to require it as well ?
18:21:31 <geppetto> yeh, having "crontabs" be the requirement is fine by me.
18:24:00 <Viking-Ice> sounds to be the most logical solution to me
18:25:21 <tibbs|w> Ugh.  Sorry for running out.
18:25:21 <Viking-Ice> limburgher, abadger1999 ^^ ?
18:25:58 <limburgher> tibbs|w no worries.
18:26:08 <limburgher> crontabs seems OK.
18:26:26 <tibbs|w> Air conditioner water supply broke and flooded a mechanical room and my office.
18:26:35 <limburgher> Yow.
18:26:41 <SmootherFrOgZ> We actually over time, can we go ahead on that item on fpc's trac?
18:26:50 <abadger1999> wfm
18:27:02 <abadger1999> both the reuire crontab and the vote on ticket.
18:27:18 <abadger1999> We have five present still..
18:27:46 <abadger1999> but maybe we need a written proposal...
18:28:21 <abadger1999> I'm a little lost about whether Require: crontabs will actually drag in a cron daemon and if that's what we're actually trying to achieve.
18:28:46 <abadger1999> Viking-Ice: want to write the proposal?
18:29:07 <abadger1999> And I'll email the list to see if we can vote in ticket.
18:29:30 <abadger1999> limburgher, The email was the haskell Guidelines reply.  You and tibbs both replied to it already.
18:29:48 <abadger1999> limburgher: (email about what to do about our ticket backlog)
18:30:09 * abadger1999 closes out meeting since he's trying to double up with fesco right now.
18:30:14 <abadger1999> #endmeeting