<@tdawson:fedora.im>
18:00:31
!startmeeting EPEL (2024-11-06)
<@meetbot:fedora.im>
18:00:32
Meeting started at 2024-11-06 18:00:31 UTC
<@meetbot:fedora.im>
18:00:32
The Meeting name is 'EPEL (2024-11-06)'
<@tdawson:fedora.im>
18:00:36
!meetingname epel
<@tdawson:fedora.im>
18:00:36
!topic aloha
<@meetbot:fedora.im>
18:00:37
The Meeting Name is now epel
<@salimma:fedora.im>
18:00:38
!hi
<@zodbot:fedora.im>
18:00:39
Michel Lind (salimma) - he / him / his
<@conan_kudo:matrix.org>
18:00:57
!hi
<@zodbot:fedora.im>
18:00:58
Neal Gompa (ngompa) - he / him / his
<@tdawson:fedora.im>
18:01:40
Hi Michel Lind 🎩 UTC-5 and Conan Kudo
<@dherrera:fedora.im>
18:02:04
!hi
<@zodbot:fedora.im>
18:02:05
Diego Herrera (dherrera) - he / him / his
<@xavierb:bachelot.org>
18:02:07
!hi
<@zodbot:fedora.im>
18:02:08
None (xavierb)
<@nirik:matrix.scrye.com>
18:02:12
morning
<@tdawson:fedora.im>
18:02:51
Hi Diego Herrera and Xavier bachelot
<@tdawson:fedora.im>
18:02:55
Morning nirik
<@salimma:fedora.im>
18:04:27
morning all
<@salimma:fedora.im>
18:04:35
well, lunchtime for me but we know it's an illusion
<@salimma:fedora.im>
18:04:38
42
<@nhanlon:beeper.com>
18:04:48
!hi
<@zodbot:fedora.im>
18:04:50
Neil Hanlon (neil) - he / him / his
<@tdawson:fedora.im>
18:04:59
Hi Neil Hanlon
<@carlwgeorge:matrix.org>
18:05:12
!hi
<@zodbot:fedora.im>
18:05:14
Carl George (carlwgeorge) - he / him / his
<@tdawson:fedora.im>
18:05:20
Hi Carl George
<@tdawson:fedora.im>
18:05:29
<@tdawson:fedora.im>
18:05:29
!topic EPEL Issues https://pagure.io/epel/issues
<@tdawson:fedora.im>
18:06:10
We have two issues today ... I'll do the one that doesn't lead us into the next topic.
<@tdawson:fedora.im>
18:06:33
!epel 303
<@zodbot:fedora.im>
18:06:33
● **Assignee:** Not Assigned
<@zodbot:fedora.im>
18:06:33
● **Last Updated:** a day ago
<@zodbot:fedora.im>
18:06:33
● **Opened:** 6 days ago by fmaurer
<@zodbot:fedora.im>
18:06:33
<@zodbot:fedora.im>
18:06:33
**epel #303** (https://pagure.io/epel/issue/303):**Can stgit be updated in epel9**
<@tdawson:fedora.im>
18:07:10
Just when we thought we'd seen them all ... :)
<@salimma:fedora.im>
18:07:14
Can't they do a `git revert` and commit that?
<@carlwgeorge:matrix.org>
18:07:33
yes, and that's probably what they should do
<@salimma:fedora.im>
18:08:11
I just asked that on the ticket because... uh no
<@carlwgeorge:matrix.org>
18:08:19
i'm really not inclined to set a precedent of "you can go ahead with an incompat update due to an accidental fastforward merge, with no other pressing reason"
<@salimma:fedora.im>
18:08:26
yep yep
<@salimma:fedora.im>
18:08:48
I've done that before, it happens, but it's not a reason as you said
<@carlwgeorge:matrix.org>
18:08:53
the only reason not to is to keep a linear git history, which everyone agrees is nice to have but is absolutely not necessary
<@salimma:fedora.im>
18:09:18
right. and in this case since the fedora branch is so far ahead, who cares if you're diverging
<@carlwgeorge:matrix.org>
18:09:28
fwiw, i did peek into the upstream release notes, and it's not as compatible as they described in the releng ticket because there are several removals
<@salimma:fedora.im>
18:09:30
if you're still keeping the branches up to date then it's a more annoying issue
<@carlwgeorge:matrix.org>
18:10:39
https://github.com/stacked-git/stgit/releases/tag/v2.2.0
<@carlwgeorge:matrix.org>
18:10:55
oh wait, wrong one
<@carlwgeorge:matrix.org>
18:10:59
https://github.com/stacked-git/stgit/releases/tag/v2.0.0
<@salimma:fedora.im>
18:11:52
yeah... idk. it's tricky
<@salimma:fedora.im>
18:12:05
can we wait for the person requesting this to answer or do we want to vote now?
<@salimma:fedora.im>
18:12:23
if they give added justification, and go through the whole process (announcing, waiting for discussion etc.) this might be ok
<@carlwgeorge:matrix.org>
18:12:43
true, technically they haven't formally started the process by sending an epel-devel email
<@salimma:fedora.im>
18:12:53
also I guess... is this normally used by humans or do we expect automation built around it
<@salimma:fedora.im>
18:12:59
if the latter then breaking changes are more painful
<@carlwgeorge:matrix.org>
18:13:23
this issue feels like a "tell me this will be approved before i take the time to seek approval"
<@jrichardson:matrix.org>
18:13:42
!hi
<@zodbot:fedora.im>
18:13:43
James Richardson (jrichardson)
<@tdawson:fedora.im>
18:13:46
Ya, and I have to say, the odds of it being approved are slim.
<@jrichardson:matrix.org>
18:13:50
sorry for being late
<@tdawson:fedora.im>
18:13:53
Hi James Richardson
<@carlwgeorge:matrix.org>
18:15:17
no way to know for sure, but it is a leaf package in epel9
<@salimma:fedora.im>
18:15:43
preapproval :)
<@tdawson:fedora.im>
18:15:55
I'm ok if we say something along the lines of "This would have to go through the proper process, and with the information we currently see, the odds are that it will not pass."
<@salimma:fedora.im>
18:16:09
I guess we can be nice and tell them how to request this with a higher chance of approval (and what our concerns are) and just close this?
<@salimma:fedora.im>
18:16:10
right
<@salimma:fedora.im>
18:16:32
these are the concerns, please make a case for why it's worth upgrading, and follow the process
<@carlwgeorge:matrix.org>
18:16:57
that's what i was aiming for in my first reply
<@carlwgeorge:matrix.org>
18:18:36
i don't think we should vote on this without the week discussion on epel-devel, and i think the issue already has the guidance needed to move it forward, so i think we can move on. anyone else can feel free to add more guidance on the issue, and the eventual mailing list thread when it happens.
<@nirik:matrix.scrye.com>
18:18:50
+1
<@tdawson:fedora.im>
18:18:52
Sounds good. Moving on.
<@tdawson:fedora.im>
18:19:15
Since the next open issue deals with EPEL10, I'm going to move to that topic before opening it.
<@tdawson:fedora.im>
18:19:27
!topic EPEL 10
<@tdawson:fedora.im>
18:19:32
!epel 300
<@zodbot:fedora.im>
18:19:33
<@zodbot:fedora.im>
18:19:33
**epel #300** (https://pagure.io/epel/issue/300):**EPEL 10 launch**
<@zodbot:fedora.im>
18:19:33
● **Opened:** 2 weeks ago by carlwgeorge
<@zodbot:fedora.im>
18:19:33
● **Last Updated:** 22 hours ago
<@zodbot:fedora.im>
18:19:33
● **Assignee:** carlwgeorge
<@salimma:fedora.im>
18:19:41
nice round number
<@tdawson:fedora.im>
18:19:57
Yep, easy for me to remember.
<@nirik:matrix.scrye.com>
18:20:00
sparta!
<@davide:cavalca.name>
18:20:58
!hi
<@carlwgeorge:matrix.org>
18:20:58
so we're just under two weeks from our target launch date. i'm ready to turn off automatic updates and use bodhi composes with a week testing period. i'd like to see that pipeline working for a little while before we announce.
<@zodbot:fedora.im>
18:20:59
Davide Cavalca (dcavalca) - he / him / his
<@davide:cavalca.name>
18:21:09
sorry my calendar is still screwed up from dst
<@salimma:fedora.im>
18:21:14
when do you want to do this Carl George
<@jrichardson:matrix.org>
18:21:26
me too Davide Cavalca
<@carlwgeorge:matrix.org>
18:21:34
i was hoping this week, and then see it work normally all next week
<@tdawson:fedora.im>
18:21:38
Carl George: Just remember to announce it on the mailling list so we know when to switch over to the new way of doing things.
<@salimma:fedora.im>
18:21:53
if you do it later this week that works for me I guess. say end of Friday?
<@tdawson:fedora.im>
18:21:54
Hi Davide Cavalca
<@carlwgeorge:matrix.org>
18:22:16
it really depends on when i can steal nirik's attention 😀
<@tdawson:fedora.im>
18:22:28
I'm planning on getting "add epel-release to CentOS's extras-common repo" done later this week. Thursday or Friday.
<@salimma:fedora.im>
18:22:44
but yeah on a more serious note, as long as it's properly announced I'm fine, I was speaking in jest
<@nirik:matrix.scrye.com>
18:22:46
happy to help with it whnever...
<@xavierb:bachelot.org>
18:22:49
it would be nice to get ffmpeg in before launch
<@carlwgeorge:matrix.org>
18:22:51
side tags will still work, so if you have a big stack just use that, no need to rush it through
<@nirik:matrix.scrye.com>
18:22:59
happy to help with it whenever...
<@salimma:fedora.im>
18:23:05
yeah. I was hoping to overtake limb in the chart of most active packagers ;)
<@tdawson:fedora.im>
18:23:26
Xavier bachelot: We're very close, I'm hoping by the end of this week.
<@yselkowitz:fedora.im>
18:23:40
oh and here I was thinking I was next!
<@salimma:fedora.im>
18:23:41
but I also need to check if my branching tool can work with side tags so that's nice opportunity to test anyway
<@carlwgeorge:matrix.org>
18:23:49
and remember having a testing period isn't the end of the world, things can still move quickly with karma
<@carlwgeorge:matrix.org>
18:24:00
and remember having a testing period isn't the end of the world, things can still move quicker with karma
<@salimma:fedora.im>
18:24:09
you're probably next! I can only see the top 5 but I know I will be building that many packages
<@salimma:fedora.im>
18:24:32
yeah... not for hundreds of packages where the only reasonable thing to say is "well it worked for building X"
<@carlwgeorge:matrix.org>
18:25:04
nirik: any ideas on that zuul failure on the pr? from what i can tell it doesn't seem related to any of the files the pr modifies.
<@carlwgeorge:matrix.org>
18:25:14
https://pagure.io/fedora-infra/ansible/pull-request/2343
<@nirik:matrix.scrye.com>
18:25:37
the CI check was broken until somewhat recently, so likely thats just all old stuff that wasn't failing before.
<@nirik:matrix.scrye.com>
18:25:43
good to fix, but shouldn't block things.
<@xavierb:bachelot.org>
18:26:32
the FTIs are not moving much lately
<@tdawson:fedora.im>
18:26:45
Carl George: I know you want to have it working for a full week, but would it be possible to hold off on the switch until Monday? The problem I'm seeing is nobody being around to give Karma if someone is building on Friday or the Weekend.
<@tdawson:fedora.im>
18:27:32
Will that really affect things? Hmm ... maybe not. We can do overrides and such if it affects us. Just a thought.
<@carlwgeorge:matrix.org>
18:27:37
my concern is if something doesn't work quite right, and then we spend multiple days getting it working, cutting into the time to observe the pipeline working correctly
<@tdawson:fedora.im>
18:27:53
Ya, you are correct.
<@carlwgeorge:matrix.org>
18:28:02
it doesn't have to be a full week, but i would prefer it
<@carlwgeorge:matrix.org>
18:28:23
between side tags and overrides, i think we're good, and i can point that out in the email i send
<@salimma:fedora.im>
18:29:09
or side tags
<@carlwgeorge:matrix.org>
18:29:12
plus i want to spend next week writing the announcement and docs, not troubleshooting the build pipeline
<@salimma:fedora.im>
18:29:30
I guess it's more a concern, if switching Friday, that it's extra load for nirik to fix things if they're broken
<@salimma:fedora.im>
18:29:47
so actually maybe Thursday is better, 'not-working-on-weekend' wise
<@carlwgeorge:matrix.org>
18:29:55
so more reason to do it today 😀
<@nirik:matrix.scrye.com>
18:29:57
we had it working before, so I hope it will not have too many problems.
<@carlwgeorge:matrix.org>
18:30:28
the one thing that makes me nervous is the repomd override bit, but we'll see how that goes
<@carlwgeorge:matrix.org>
18:30:56
anyways, i'll send the email today, and folks should expect the cutover sometime later this week
<@salimma:fedora.im>
18:31:48
thanks for the heads up :)
<@carlwgeorge:matrix.org>
18:31:50
other than that cutover, for the launch we just need some docs and the announcement
<@carlwgeorge:matrix.org>
18:33:12
unless anyone else has epel10 things to bring up, i think we can move on to the next topic
<@xavierb:bachelot.org>
18:34:15
Carl George: will you file more FTIs or untag stuff on the announcement ?
<@tdawson:fedora.im>
18:34:34
Who is doing the "add epel-release-latest symlink" ? I was just thinking that needs to be done before the documentation.
<@jrichardson:matrix.org>
18:34:45
I am
<@carlwgeorge:matrix.org>
18:34:49
the main thing i'm tracking to untag is nextcloud. it's down to just one missing dep, but that one is stall because it might be included in rhel/centos
<@jrichardson:matrix.org>
18:35:01
well, with Carl George
<@carlwgeorge:matrix.org>
18:35:20
the work for the symlink is done i think, we just need to verify it works once we're on bodhi composes
<@carlwgeorge:matrix.org>
18:35:29
the work for the symlink is done i think, we just need to verify it worked once we're on bodhi composes
<@tdawson:fedora.im>
18:35:30
OK, thank you.
<@carlwgeorge:matrix.org>
18:35:48
it's part of the new-updates-sync script that bodhi runs
<@carlwgeorge:matrix.org>
18:36:09
i theory once epel10 is being pushed by that, we'll have the symlink
<@carlwgeorge:matrix.org>
18:36:16
in theory once epel10 is being pushed by that, we'll have the symlink
<@tdawson:fedora.im>
18:36:40
OK, that was the last question I had. I'm going to move on.
<@tdawson:fedora.im>
18:36:50
!topic Old Business
<@tdawson:fedora.im>
18:36:57
Did we have any old business this week?
<@salimma:fedora.im>
18:37:48
I have something for open floor later
<@tdawson:fedora.im>
18:37:54
I haven't seen Smooge in the past couple meetings. I hope he's doing well.
<@tdawson:fedora.im>
18:38:32
I'll give it one more minute, then move to open floor if nothing comes up.
<@tdawson:fedora.im>
18:39:13
!topic General Issues / Open Floor
<@tdawson:fedora.im>
18:39:26
Michel Lind 🎩 UTC-5: you said you had something, go for it.
<@salimma:fedora.im>
18:39:58
yes
<@salimma:fedora.im>
18:40:21
so Carl and I have encountered issues branching packages that already have epel-packagers-sig in the collaborator list ... if we were not granted wildcard access to epel*
<@salimma:fedora.im>
18:41:13
since right now releng assumes it's up to us to ask the maintainer to fix the access ... is this something we can just fix in the policy and say "if you've granted access to epel-packagers-sig before but it was limited to some branches, we can ask releng to fix it directly"
<@tdawson:fedora.im>
18:41:37
I honestly didn't know that was a thing you could do (set it to certain epel levels)
<@salimma:fedora.im>
18:41:55
related: can we ask releng to override the bugzilla assignee for EPEL, for packages where the maintainer expressed they're not interested in EPEL but never fix the setting?
<@carlwgeorge:matrix.org>
18:42:00
yeah collab access is a wildcard pattern, but sometimes it gets setup as just for a specific branch
<@salimma:fedora.im>
18:42:02
yeah turns out 'epel7,epel8,epel9' works
<@salimma:fedora.im>
18:42:05
I did not realize either
<@tdawson:fedora.im>
18:42:16
If they are set to certain levels, that is going to be a mess when we start getting epel10.3 and things like that.
<@xavierb:bachelot.org>
18:42:39
that is a regexp, so it can be fixed
<@carlwgeorge:matrix.org>
18:43:01
i don't think it's regex, because `epel*` works
<@xavierb:bachelot.org>
18:43:12
or glob, rather
<@salimma:fedora.im>
18:43:22
if it's a glob CSV won't work :)
<@salimma:fedora.im>
18:43:27
there seems to be some parsing going on
<@carlwgeorge:matrix.org>
18:43:59
if we did this, i think the policy we should aim for would be along the lines of "if you're granting collab access to epel-packagers-sig, it must be on `epel*`, not on specific epel branches"
<@carlwgeorge:matrix.org>
18:44:16
then asking releng to fix is asking releng to help a package comply with policy
<@conan_kudo:matrix.org>
18:44:59
yup
<@carlwgeorge:matrix.org>
18:45:01
i suggest figuring out the wording in a docs pr, then we can vote on the pr
<@salimma:fedora.im>
18:45:26
that sounds good
<@salimma:fedora.im>
18:45:55
is it reasonable to propose a similar PR for the bugzilla assignee field?
<@salimma:fedora.im>
18:46:08
that's also something that sounds like if we make it explicit in policy we can ask releng to fix
<@salimma:fedora.im>
18:46:36
maybe the condition will be "the main admin responds in a comment that they're not interested in epel" or "the main admin never did any epel builds"
<@tdawson:fedora.im>
18:46:45
I don't think the packagers-sig can be put in the bugzilla assignee field, can it?
<@salimma:fedora.im>
18:46:50
I have one package where the epel assignee is ... 'orphan' :(
<@salimma:fedora.im>
18:46:58
you can
<@salimma:fedora.im>
18:47:02
it happens to a lot of Rust packages
<@salimma:fedora.im>
18:47:07
@epel-packagers-sig would work
<@tdawson:fedora.im>
18:47:14
Oh ... well that would be nice.
<@carlwgeorge:matrix.org>
18:47:15
oh is this the weird thing where only the main admin can change the epel bz assignee?
<@salimma:fedora.im>
18:47:19
yep
<@salimma:fedora.im>
18:47:28
and the main admin often won't esp for packages that go through the stalled flow
<@carlwgeorge:matrix.org>
18:47:45
should be making a policy to work around a bug, or just work on fixing the bug?
<@tdawson:fedora.im>
18:47:48
The orphan one is also this wierd thing - https://pagure.io/epel/issue/199
<@salimma:fedora.im>
18:48:37
well... if we have a policy for bugzilla assignee we might as well handle the case where the assignee is invalid too
<@tdawson:fedora.im>
18:48:50
TLDR; If a package get's orphaned in Fedora, and then picked up by someone, it orphans both Fedora and EPEL, but only Fedora get's picked up by the person.
<@salimma:fedora.im>
18:49:09
it just seems silly where if the maintainer is not interested in epel but does not care enough to set the field properly, we end up missing bugs
<@salimma:fedora.im>
18:49:18
that definitely should be fixed
<@salimma:fedora.im>
18:49:35
also: if the package is orphaned but the epel assignee is already different, does it still reset it?
<@tdawson:fedora.im>
18:49:50
I have no idea.
<@carlwgeorge:matrix.org>
18:50:48
i honestly don't know what a policy to resolve this would look like
<@salimma:fedora.im>
18:51:26
let's deal with the branching first then ... I'm swamped this week so if someone wants to take a stab, be my guest
<@salimma:fedora.im>
18:51:31
otherwise I can try writing this next week
<@salimma:fedora.im>
18:51:52
we won't vote without discussing it anyway, so next wed seems like a reasonable timeline :)
<@carlwgeorge:matrix.org>
18:52:05
i think we should touch on the most recent epel-devel thread before we close out the open floor
<@carlwgeorge:matrix.org>
18:52:31
<@carlwgeorge:matrix.org>
18:53:24
i share in miro's concerns
<@tdawson:fedora.im>
18:53:33
I'm not saying I asked Miro to ask this, but Miro has changed the EPEL maintainer on two packages to me during this EPEL10 round.
<@tdawson:fedora.im>
18:53:54
Err ... possibly three.
<@salimma:fedora.im>
18:54:43
this ownership thing is tricky
<@carlwgeorge:matrix.org>
18:54:46
we're all human, and i know i've missed bz updates before, this is just a reminder to everyone to be mindful of this problem
<@salimma:fedora.im>
18:55:45
but yeah idk, in terms of fixing bugs I think "getting a package branched" is lower priority than fixing actual bugs unless it's demonstrably blocking something else? (many people don't link their bugs together)
<@tdawson:fedora.im>
18:56:06
Here's one concern I have. One of those three packages should have been taken care of by the epel-packagers-sig. And I thought it odd that nobody worked on it. But now that I see the thing above, I'm wondering if Miro is setting things too tight, like EPEL9, and the epel-packagers-sig isn't even seeing the emails.
<@salimma:fedora.im>
18:56:13
I probably should stop using my python-packagers-sig access to branch packages in the future though and ping to get epel-packagers-sig added instead
<@salimma:fedora.im>
18:56:30
the bugzilla email can't be set to a specific branch I think
<@salimma:fedora.im>
18:56:34
it's just all of Fedora and all of EPEL
<@salimma:fedora.im>
18:56:54
but yeah ... should we bring back the triage thing we used to have for security issues?
<@salimma:fedora.im>
18:57:01
make it monthly or so so it's not too burdensome
<@carlwgeorge:matrix.org>
18:57:03
that's just shifting the problem i think
<@carlwgeorge:matrix.org>
18:57:23
the problem isn't how we get access to branch, it's about ongoing maintenance
<@jrichardson:matrix.org>
18:57:38
Sorry, for all the new people that might be watching (like me :)) can someone break this down a bit? People are building packages that they don't own or maintain...why?
<@carlwgeorge:matrix.org>
18:57:48
i know smooge ran into this problem in epel8, and regretted branching too many things
<@salimma:fedora.im>
18:58:05
well, people have access to these via group ACL
<@salimma:fedora.im>
18:58:13
but they might be building it only as a dependency
<@jrichardson:matrix.org>
18:58:27
Michel Lind 🎩 UTC-5: Ah! thanks
<@carlwgeorge:matrix.org>
18:58:39
yeah generally they have access somehow, but it's all too easy to build the thing you need to unblock yourself for something else, and then ignore the package
<@salimma:fedora.im>
18:58:49
even in packaging ecosystems that collectively maintain everything (Rust and Golang) the policies are inconsistent
<@salimma:fedora.im>
18:59:17
Rust folks normally also update in epel if the package is branched, golang folks normally only update in Fedora and I end up rebasing when the leaf package I need is bumped
<@tdawson:fedora.im>
18:59:40
James Richardson: In the past, EPEL9 timeframe, people have said they would maintain the EPEL branches of packages. But then after they did the initial epel9 (or epel8) builds, they didn't to any other maintenance. And when other people are asking for them to be branched and build for epel10, they still aren't even doing that.
<@carlwgeorge:matrix.org>
18:59:40
imo, epel-packagers-sig has grown to the point that it's really difficult to stay on top of all the emails
<@jrichardson:matrix.org>
19:00:10
Thanks guys
<@salimma:fedora.im>
19:00:14
yeah, though Miro set the rhbz to a person not to the group right?
<@salimma:fedora.im>
19:00:20
so emails can be filtered differently
<@salimma:fedora.im>
19:00:35
but yeah without the rhbz assignee it will be trickier
<@tdawson:fedora.im>
19:00:35
I agree. That's why I had to drop. I couldn't do the ones I know where mine alone, because they were getting mixed in.
<@jrichardson:matrix.org>
19:01:22
Looks like we're at time
<@tdawson:fedora.im>
19:01:36
James Richardson: Thanks, I totally lost track.
<@jrichardson:matrix.org>
19:01:46
No worries
<@salimma:fedora.im>
19:01:52
I think needinfo-ing the person who last built in EPEL and setting them as the EPEL assignee by default would make sense
<@salimma:fedora.im>
19:01:58
oh yeah we're out of time
<@tdawson:fedora.im>
19:02:06
Thank you all for coming and the good discussions we had.
<@carlwgeorge:matrix.org>
19:02:11
so there's no exact fix for this, just everyone doing their part
<@tdawson:fedora.im>
19:02:24
And thank you all for all you do for EPEL and it's community.
<@jrichardson:matrix.org>
19:02:50
Thanks Troy Dawson
<@tdawson:fedora.im>
19:02:52
I'll talk to you all next week, if not sooner .... hmm ... I always say that ... I'll have to think of something different for next week.
<@xavierb:bachelot.org>
19:02:54
thanks Troy Dawson, thanks all
<@tdawson:fedora.im>
19:03:09
!endmeeting