17:00:28 <smooge> #startmeeting EPEL
17:00:28 <zodbot> Meeting started Fri Mar  6 17:00:28 2015 UTC.  The chair is smooge. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:28 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:36 <smooge> #meetingname epel
17:00:36 <zodbot> The meeting name has been set to 'epel'
17:00:54 <Evolution> woohoo
17:00:54 <smooge> #chairs bstinson Evolution dgilmore nirik
17:01:02 <smooge> #topic Robot Rollcall
17:01:08 * bstinson is here!
17:01:15 <Evolution> present
17:02:22 <Evolution> guess it's just bstinson and I...
17:02:25 * Evolution looks at smooge
17:03:49 * nirik is here, but also in a git hell. ;)
17:03:57 <smooge> nirik, and dgilmore are working on getting alpha ready
17:04:05 <Evolution> ah
17:04:17 <smooge> I need to let bstinson run these meetings. People show up when he does
17:04:39 <smooge> #topic Status of things
17:04:56 <smooge> Ok so currently RHEL-7.1 just came out and CentOS-7.1 will be soon ...
17:05:28 <smooge> avij posted a bunch of packages that we will need to remove/block from EPEL
17:05:43 <smooge> Evolution, how long before you think the CentOS-7.1 will come out?
17:06:03 <Evolution> really? *that* the question you're going to lead with? :-P
17:06:18 <Evolution> for 7.0 we were right at the 30 day mark.
17:06:48 <Evolution> I would imagine that we'll have the CR repo populated late next week, and then iso testing after.
17:07:11 <Evolution> the goal is still the same "2 weeks for point update media"
17:07:18 <Evolution> that it's always been.
17:07:31 <smooge> Evolution, it was that or "I see you hate chocolate."
17:07:38 <Evolution> heh
17:08:04 <Evolution> some of the debranding patches from .0 don't apply cleanly to .1, so some changes and then testing needed.
17:08:04 <smooge> ok that is cool. the reason is that it leads to the block/remove package timeline
17:08:50 <dgilmore> hola
17:09:00 <smooge> I was going to say we get a complete list of packages in 7.1 core and make a list of packages we need to 'block' in koji
17:09:23 <dgilmore> smooge: we have a json file with all the packages in rhel
17:09:35 <smooge> is it updated to 7.1?
17:09:43 <dgilmore> writing a script to work out what to block shouldn't be hard
17:09:43 <smooge> and who is we :)
17:10:20 <smooge> dgilmore, my main worry was if we want to retire packages now or wait til CentOS-7.1 came out
17:10:26 <dgilmore> smooge: https://infrastructure.fedoraproject.org/repo/json/pkg_el7.json
17:10:36 <dgilmore> its updated daily so should have 7.1 now
17:10:50 <dgilmore> smooge: historically we have waited for CentOS
17:11:45 <smooge> no problem. I just want to make sure we are clear on this. I want to give the guys we are blocking/removing a heads up in case they were going to update nethack-4.0 and it was 3.9 in RHEL
17:12:29 <dgilmore> smooge: bodhi needs to learn to check the json file and block those types of updates
17:12:30 <smooge> dgilmore, can we block and remove later? so that people can't build new packages but it doesn't remove from the repo. Or are they part of the same step in the tools
17:12:41 <dgilmore> smooge: no
17:12:48 <dgilmore> its blocked and removed all at once
17:13:07 <smooge> ok then I will make a warning email this weekend to send out
17:13:39 <smooge> I will play with the json and python also
17:14:04 <smooge> alright Evolution what was I supposed to lead with :)?
17:15:51 <Evolution> no clue. we could always go back down the python3 rabbit hole
17:15:53 <Evolution> :-P
17:16:03 <Evolution> (dear god please no)
17:16:10 <dgilmore> how about ppc64le
17:16:10 <smooge> #topic How far does this rabbit hole go? Python3
17:16:18 <smooge> dgilmore, next up
17:16:46 <smooge> OK all I want to do on the python3 is say I have looked over the proposal and I think its good enough for any approval we need to give.
17:17:08 <smooge> however if there are parts that I am overlooking I would be happy to learn them.
17:17:13 * maxamillion is here now
17:17:17 <maxamillion> that upgrade went terribly
17:17:34 <bstinson> smooge: +1, the conversations on the list seemed to handle some of the clarifications needed
17:18:19 <smooge> dgilmore, nirik have you guys had anything to say on it? I can move it to an on-list vote if you hvan'te had time because of alpha but could look at it over the weekend
17:19:20 <nirik> last I looked at it it was good. I can look again this weekend.
17:19:35 <dgilmore> smooge: my head has been in Fedora Alpha land all week
17:19:48 * dgilmore is good with what he has seen
17:21:06 <bstinson> looks like ownership is being handed off to mstuchli, i can reach out to see what support we can give
17:21:39 <Evolution> I'm +1 for python3 as list-clarified
17:22:06 <smooge> okie dokie
17:22:24 <smooge> I will put this to a list vote with finalization on next Friday.
17:22:36 <maxamillion> +1 for python3 also (for what it's worth, I can never remember who all's +1's technically "count")
17:22:49 <smooge> #info Vote on Python3 proposal will be done on the list with finalization at next meeting.
17:23:00 <smooge> #topic More Architectures
17:23:30 <smooge> dgilmore, this is your part
17:24:26 <maxamillion> oh man
17:24:27 <dgilmore> so rhel will be supporting ppc64le as an official arch and I think we should also
17:24:41 <dgilmore> we have the hardware
17:24:45 <Evolution> agree.
17:25:00 <dgilmore> they have java which is a pain point in epel6
17:25:14 <dgilmore> I think it will be a win for us
17:25:22 <maxamillion> I'm all for it ... does CentOS have the hardware? (since we use CentOS in mock/copr configs for building, or is that not the case in koji?)
17:25:44 <dgilmore> maxamillion: they do not have ppc64 builds from memory
17:25:55 <dgilmore> Evolution: is that still true?
17:26:38 <Evolution> yes, but we have access to hardware now via 3rd party. so after x86_64 and i686 we'll start on that
17:26:43 <dgilmore> Evolution: and what is the CentOS plans in regards to LE?
17:27:04 <maxamillion> granted I don't see copr/mock being a real consideration since there's not likely a way to run that arch in openstack
17:27:06 <Evolution> we're going to put some effort into both ppc64le and aarch64.
17:27:27 <maxamillion> brave
17:27:41 <Evolution> in theory, ppc64le will be easier since that can be done from actual rhel, where aarch64 is done from the old f19stage4
17:27:58 <dgilmore> maxamillion: well you can run ppc64 be and le in openstack
17:28:15 <dgilmore> Evolution: :( yeah
17:28:19 <Evolution> from what I've seen, the ppc64le stuff looks fairly badass.
17:28:27 <dgilmore> it is nice
17:29:02 <dgilmore> I do not know what aarch64 will look like in rhel land
17:29:16 <dgilmore> but I think it is safe to assume that at some point it will be there
17:29:29 <dgilmore> and I think we should support it at that point
17:29:39 <maxamillion> dgilmore: I didn't know that ... qemu emulation?
17:29:40 <Evolution> I'm probably ~1month out from having something installable for cent on aarch64
17:29:45 <dgilmore> is anyone against supporting ppc64le in epel?
17:30:17 <bstinson> no objection here
17:30:20 <Evolution> dgilmore: I'm doing my rebuild as sig work at present, because what i'm doing may end up being different from what actually shows up if/when it arrives.
17:30:20 <smooge> OK I am going to be the negative nanny here. We have ~100 users of PPC. All of them pretty much behind redhat.com and ibm.com
17:30:38 <dgilmore> Evolution: cool
17:31:29 <dgilmore> smooge: I do believe openpower is starting to get there
17:31:41 <dgilmore> smooge: so I feell that the number of users will increase
17:32:24 <dgilmore> http://www.zdnet.com/article/rackspace-joins-openpower/
17:32:42 <smooge> My main issue is that Power is a secondary architecture in Fedora. It has people who focus on there. Are they going to help out on the EPEL side or is it just oging to be you and Evolution ?
17:33:03 <dgilmore> http://www.tyan.com/campaign/openpower/
17:33:16 <dgilmore> smooge: they do help
17:34:05 <smooge> dgilmore, I quit looking at those on the last time IBM said it was going to do this in the early 2000's and then changed their minds and all of that alternative power went away silently
17:34:56 <smooge> my main thing was hoping they would be the ones championing it versus two very overworked work-aholics
17:35:07 <dgilmore> smooge: now that IBM got rid of their x86 business they actually need it to be a sucess
17:36:25 <smooge> possibly. They let go a portion of their PPC people last month so I really don't know.
17:36:34 <smooge> in any case.. that isn't really my point
17:37:14 <dgilmore> I think it is worth our while to support
17:38:20 <dgilmore> so it seems everyone but smooge is for supporting ppc64le
17:38:39 <smooge> I don't doubt that. And I expect it is. I want someone else shouldering it so when something doesn't compile on PPCle for some reason that x86_64 people don't understand.. it isn't on you to move a mountain
17:38:52 <smooge> I am +0
17:39:35 <smooge> nirik, you have a say?
17:39:53 <maxamillion> yeah, that's a fair concern I think .... all said and done, it comes down to supportability and if it's just going to be left to like 2 or 3 people who may or may not have a PPCle background
17:40:01 * nirik cannot pay attention right now, sorry
17:40:51 <maxamillion> I absolutely have no experience with PPCle nor access to any hardware so if I ran into an arch specific build issue I'm not sure how useful I would be in the effort.... I'm willing to learn though, just not sure that's the best strategy
17:41:30 <maxamillion> I'll +0 also because I'm kind of on the fence, no real preference, some pros/cons that I think kind of stalemate in my mind
17:41:58 <smooge> dgilmore, if we had secondary architecture EPEL I would be +1 in an instant.. if someone from the PPCle team comes in and says "We are happy to support this and make this work" I am also +1 in an instant
17:42:37 <smooge> if you say we can drop ppc and replace it with ppcle... I will be +1 so fast that the Earth will turn backwards and Lois Lane will be saved
17:42:50 <dgilmore> smooge: okay. the power community wants to support it
17:43:00 <dgilmore> and most if not all le issues hit fedora also
17:43:04 <dgilmore> and get fixed there
17:43:21 <dgilmore> smooge: at least for now we will have both be and le
17:43:57 <smooge> well I can't get everything I want :)
17:44:18 <dgilmore> we would need to build the arch specific rpms in mock and import to koji and change the arches
17:44:35 <dgilmore> implementation is not hard
17:44:44 <smooge> let us move this to an email list vote also so that nirik can focus on it when he isn't sick/dealing with broken shit. We will finalize next week
17:45:27 <dgilmore> sure
17:45:37 <smooge> Since I am not blocking with a +0 I am going to say that it is mostly deal with any side line items and we can finalize and figure out what needs to be done to get the builds going
17:46:14 <smooge> dgilmore, can you get some people from ppc on EPEL-devel as that is where questions for things 'broken' will end up.
17:46:40 <smooge> #info email vote on PPCle inclusion to EPEL to be finalized next week
17:47:30 <smooge> #topic Open Floor
17:47:56 <smooge> anything for the open floor? broken builds etc?
17:48:19 <smooge> if not we can close this out in 2 minutes.
17:48:30 <dgilmore> smooge: sure
17:48:54 <smooge> thanks dgilmore
17:49:21 <bstinson> nothing on my end
17:50:26 <smooge> #endmeeting