<@sgallagh:fedora.im>
19:00:05
!startmeeting FESCO (2024-06-17)
<@meetbot:fedora.im>
19:00:06
Meeting started at 2024-06-17 19:00:05 UTC
<@meetbot:fedora.im>
19:00:06
The Meeting name is 'FESCO (2024-06-17)'
<@sgallagh:fedora.im>
19:00:17
!meetingname fesco
<@meetbot:fedora.im>
19:00:18
The Meeting Name is now fesco
<@sgallagh:fedora.im>
19:00:26
Chairs: @conan_kudo:matrix.org, @ngompa:fedora.im, @nirik:matrix.scrye.com, @humaton:fedora.im, @zbyszek:fedora.im, @sgallagh:fedora.im, @jistone:fedora.im, @dcantrell:fedora.im, @mhayden:fedora.im, @tstellar:fedora.im
<@sgallagh:fedora.im>
19:00:31
!topic Init Process
<@sgallagh:fedora.im>
19:00:37
!hi
<@zodbot:fedora.im>
19:00:38
Stephen Gallagher (sgallagh) - he / him / his
<@nirik:matrix.scrye.com>
19:00:53
morning
<@humaton:fedora.im>
19:00:57
!hi
<@zodbot:fedora.im>
19:00:57
Tomáš Hrčka (humaton) - he / him / his
<@sgallagh:fedora.im>
19:01:51
Josh wrote me last night to say he was sick and wouldn't be able to run the meeting, so I took over for him
<@zbyszek:fedora.im>
19:01:52
!hi
<@zodbot:fedora.im>
19:01:53
Zbigniew Jędrzejewski-Szmek (zbyszek)
<@conan_kudo:matrix.org>
19:02:39
!hi
<@zodbot:fedora.im>
19:02:52
Neal Gompa (ngompa) - he / him / his
<@humaton:fedora.im>
19:04:01
soo nirik got devconf flu over tcp?
<@conan_kudo:matrix.org>
19:04:05
I don't think I'm feverish, but my throat hurts like hell and it feels like hay fever
<@conan_kudo:matrix.org>
19:04:17
I blame the train from Brno to Berlin
<@conan_kudo:matrix.org>
19:04:23
I was actually fine before that train ride
<@humaton:fedora.im>
19:04:37
I was fine until today morning
<@conan_kudo:matrix.org>
19:04:48
and Berlin is full of smokers, even more so than Brno
<@nirik:matrix.scrye.com>
19:04:52
heh. remote devconf flu! (but no, was on a family vacation and brought back our old unwelcome friend covid19)
<@sgallagh:fedora.im>
19:05:42
I count 5, so we have quorum. I'll give folks another minute or two to filter in since we really only have one topic.
<@zbyszek:fedora.im>
19:05:48
Hmm, looking at the template above, I think we didn't update it after elections.
<@sgallagh:fedora.im>
19:06:01
oof, whoops
<@sgallagh:fedora.im>
19:06:12
In my defense, it's not on the checklist.
<@conan_kudo:matrix.org>
19:06:14
we really also need to add constraints to fesco scheduling
<@conan_kudo:matrix.org>
19:06:26
because it has to be before releng and go/no-go
<@sgallagh:fedora.im>
19:06:40
https://docs.fedoraproject.org/en-US/fesco/Updating_Fedora_Engineering_Steering_Committee_Members/
<@conan_kudo:matrix.org>
19:07:03
everything should be updated except the badge
<@conan_kudo:matrix.org>
19:07:11
and the new timeslot
<@sgallagh:fedora.im>
19:07:37
Conan Kudo: We didn't update the meeting template to include the new members and remove the emeritus ones
<@sgallagh:fedora.im>
19:07:48
I'll do that after the meeting
<@sgallagh:fedora.im>
19:08:07
Who is missing? We should probably ping them directly.
<@zbyszek:fedora.im>
19:08:44
BTW, I made a ping in https://pagure.io/fedora-badges/issue/628.
<@humaton:fedora.im>
19:08:59
btw how do I claim fesco badge?
<@zbyszek:fedora.im>
19:09:09
OK, and I'll update the docs page to include updating the wiki.
<@sgallagh:fedora.im>
19:09:29
Fabio Valentini: Michel Lind 🎩 Sorry for not pinging you sooner
<@sgallagh:fedora.im>
19:09:39
jednorozec: Aoife will award it to you
<@sgallagh:fedora.im>
19:09:49
I asked her to do so on Friday, but she might have forgotten
<@sgallagh:fedora.im>
19:11:44
Alright, let's get going, I guess.
<@sgallagh:fedora.im>
19:12:35
!topic #3222 Change: Make Tuned the Default Power Profile Management Daemon
<@sgallagh:fedora.im>
19:12:40
!fesco 3222
<@zodbot:fedora.im>
19:12:41
**fesco #3222** (https://pagure.io/fesco/issue/3222):**Change: Make Tuned the Default Power Profile Management Daemon**
<@zodbot:fedora.im>
19:12:41
● **Assignee:** smallorange
<@zodbot:fedora.im>
19:12:41
● **Last Updated:** 6 hours ago
<@zodbot:fedora.im>
19:12:41
● **Opened:** 6 days ago by amoloney
<@zodbot:fedora.im>
19:12:41
<@salimma:fedora.im>
19:12:46
!hi
<@zodbot:fedora.im>
19:12:46
Michel Lind (salimma) - he / him / his
<@conan_kudo:matrix.org>
19:12:53
welcome to the party :)
<@salimma:fedora.im>
19:13:11
I thought Neal was trying to skip this meeting 😅
<@conan_kudo:matrix.org>
19:13:14
unfortunately we ran out of drinks
<@sgallagh:fedora.im>
19:13:24
So, the person who voted -1 isn't here currently, possibly because I failed to ping them in time
<@conan_kudo:matrix.org>
19:13:25
Michel Lind 🎩: I was... but then the thing I was doing ended _just_ before this
<@salimma:fedora.im>
19:13:29
I'm still waiting too
<@salimma:fedora.im>
19:13:58
I'm at dinner and the phone is about to die, so apologies if I disappear
<@nirik:matrix.scrye.com>
19:14:07
I'd like to hear the reasoning behind the -1...
<@jsteffan:fedora.im>
19:14:11
!hi
<@zodbot:fedora.im>
19:14:12
Jonathan Steffan (jsteffan)
<@conan_kudo:matrix.org>
19:14:31
yeah I'm curious about it, since it seems outside of hadess, the feedback was either neutral or positive
<@zbyszek:fedora.im>
19:14:57
-1 in votes should be like pings, i.e. discouraged without additional context.
<@sgallagh:fedora.im>
19:15:15
Strictly speaking, we could opt to vote anyway.
<@sgallagh:fedora.im>
19:15:35
But I really don't want to do that prematurely. Fabio might have some very compelling argument we haven't heard yet
<@salimma:fedora.im>
19:15:59
We can vote to postpone?
<@nirik:matrix.scrye.com>
19:16:03
right. can we just punt this to next week?
<@sgallagh:fedora.im>
19:16:38
Does anyone know if this is time-sensitive? I don't think there's any specific deadline they'd need to meet.
<@decathorpe:fedora.im>
19:17:05
sorry, I missed the memo that meetings were moved from :30 to :00
<@sgallagh:fedora.im>
19:17:51
Fabio Valentini: I'm going by the fedocal; did we move it and forget to update the time?
<@decathorpe:fedora.im>
19:18:13
I'm not sure, I might just not have gotten the memo
<@salimma:fedora.im>
19:18:15
We've not voted a new time yet
<@sgallagh:fedora.im>
19:18:15
(Well, we've got a WhenIsGood in the pipeline now anyway).
<@zbyszek:fedora.im>
19:18:32
My calendar says that they've been at :00 since April 15th. I remember we vacillated a bit on the time, but no more details.
<@sgallagh:fedora.im>
19:18:32
Fabio Valentini: So... why the -1 on tuned?
<@nirik:matrix.scrye.com>
19:18:55
we decided it at the end of some random meeting in april... ;)
<@decathorpe:fedora.im>
19:19:56
basically, I'm not sure I like moving from a small, dedicated, efficient (space+time) point solution to a "everything and the kitchen sink" toolkit that looks bigger in both hard disk / memory / and boot time requirements
<@zbyszek:fedora.im>
19:19:57
Stephen Gallagher: please also make results visible.
<@conan_kudo:matrix.org>
19:20:33
Fabio Valentini: the main piece of feedback I got from our OEM contributors was that tuned gives them an opportunity to customize the profiles for their systems in a way that ppd can't
<@conan_kudo:matrix.org>
19:21:00
and from our perspective, there are so many systems that ppd does _nothing_ on because we don't have the required firmware data to make it do anything
<@conan_kudo:matrix.org>
19:21:17
because the tooling to grab that information is proprietary and unavailable to us
<@salimma:fedora.im>
19:21:25
Tuned is deprecated upstream right?
<@salimma:fedora.im>
19:21:27
Sorry ppd
<@salimma:fedora.im>
19:21:34
Ppd is deprecated upstream right?
<@salimma:fedora.im>
19:21:57
I have used tuned before and it's fine, I like that it now has GUI integration
<@zbyszek:fedora.im>
19:21:59
Do we have any measurements or concrete examples?
<@conan_kudo:matrix.org>
19:21:59
it was, and now it's not
<@blackwell:fedora.im>
19:22:17
!hi
<@zodbot:fedora.im>
19:22:19
Jason Blackwell (blackwell)
<@decathorpe:fedora.im>
19:22:20
and we have that data for tuned somehow?
<@conan_kudo:matrix.org>
19:22:21
the only concrete examples I can give you are my own computers, most of which the power-profiles-daemon settings do nothing
<@salimma:fedora.im>
19:22:33
Oh it's maintained again?
<@zbyszek:fedora.im>
19:23:10
How does one check if it's "doing something"?
<@conan_kudo:matrix.org>
19:23:15
we don't, but it is actually possible to override the profiles and set up custom aliases to them for tools to use... I had talked to a couple of the OEM contributors (Mark from Lenovo being one of them) who are interested in doing so
<@conan_kudo:matrix.org>
19:23:29
you should see energy levels going down if you go into power-saver
<@conan_kudo:matrix.org>
19:23:36
or fans spinning up for performance
<@decathorpe:fedora.im>
19:24:06
so I'm not sure how this is improving the status quo other than "OEMs might support this in the future, but they don't right now" ?
<@nirik:matrix.scrye.com>
19:24:17
I'm a bit surprised ppd would do 'nothing'... it uses some pretty standard interfaces I thought...
<@conan_kudo:matrix.org>
19:24:17
right now, they can't really do anything
<@conan_kudo:matrix.org>
19:24:35
they are only supported on some platforms, mostly Intel... not really ARM or AMD
<@conan_kudo:matrix.org>
19:24:48
though I think some really new AMD stuff works now that ppd is maintained by an AMD guy
<@conan_kudo:matrix.org>
19:25:21
none of my AMD machines seem to do anything with ppd except my Framework 16
<@salimma:fedora.im>
19:25:56
Pretty sure it works on my ryzen 7 but I can't check for a week
<@decathorpe:fedora.im>
19:26:20
ok, let me summarize my position: I don't like approving moving from A to B where A is "proven" just because B *might* do more things in the future. I'm going to stick with -1 vote but I am happy to be out-voted
<@sgallagh:fedora.im>
19:26:21
One major concern I have is this: they can't coexist.
<@conan_kudo:matrix.org>
19:27:00
only tuned-ppd cannot coexist with ppd
<@sgallagh:fedora.im>
19:27:01
You can have one or the other running on your system and tuned provides (from what I can tell) a superset of the functionality of PPD
<@conan_kudo:matrix.org>
19:27:03
tuned itself can
<@conan_kudo:matrix.org>
19:27:16
yes
<@sgallagh:fedora.im>
19:27:18
Is that new? I know we had a lot of issues with that on RHEL 9
<@conan_kudo:matrix.org>
19:27:25
yes it's very new
<@sgallagh:fedora.im>
19:27:37
The coexistence, I mean
<@conan_kudo:matrix.org>
19:27:43
no that isn't
<@conan_kudo:matrix.org>
19:28:02
tuned's functionality is mostly a manager of kernel tunables
<@decathorpe:fedora.im>
19:28:03
huh? I thought tuned.service and ppd.service conflict with each other ...
<@conan_kudo:matrix.org>
19:28:11
they shouldn't?
<@sgallagh:fedora.im>
19:28:11
They do on RHEL 9
<@sgallagh:fedora.im>
19:28:19
Which is the source of my query above
<@zbyszek:fedora.im>
19:28:21
I think it's better if only one or the other is active at once. It'd be a nightmare to understand what is going on if both were active at once.
<@conan_kudo:matrix.org>
19:28:23
I guess this is like tlp and ppd too
<@conan_kudo:matrix.org>
19:28:50
there's a lot of guides to use tlp to replace ppd because of the issues around ppd's effectiveness
<@nirik:matrix.scrye.com>
19:29:07
I thought tuned-ppd conflicts with ppd?
<@sgallagh:fedora.im>
19:29:27
nirik: The packages conflict yes
<@sgallagh:fedora.im>
19:29:45
But also the tuned and ppd systemd units disable the other when either is started
<@conan_kudo:matrix.org>
19:30:07
this is an example of the kind of feedback we got about ppd: https://www.reddit.com/r/Fedora/comments/qpaa4g/tlp_vs_powerprofilesdaemon/
<@conan_kudo:matrix.org>
19:30:42
personally, I want our defaults to be better
<@conan_kudo:matrix.org>
19:30:51
and I've been frustrated with how hard it is to do that
<@nirik:matrix.scrye.com>
19:30:52
I don't think the fedora units do that... because you can't install them both at the same time...
<@nirik:matrix.scrye.com>
19:31:50
oh it does. fun
<@nirik:matrix.scrye.com>
19:31:54
Conflicts=tuned.service tlp.service auto-cpufreq.service system76-power.service
<@zbyszek:fedora.im>
19:33:42
Oh, oh. The packages don't have a Conflicts, but they conflict at the file level. So a packaging bug.
<@conan_kudo:matrix.org>
19:33:56
yes I suggested a fix for that
<@zbyszek:fedora.im>
19:33:57
- file /usr/share/dbus-1/system.d/net.hadess.PowerProfiles.conf from install of power-profiles-daemon-0.21-2.fc40.x86_64 conflicts with file from package tuned-ppd-2.23.0-1.fc40.noarch
<@zbyszek:fedora.im>
19:33:57
- file /usr/share/dbus-1/system-services/net.hadess.PowerProfiles.service from install of power-profiles-daemon-0.21-2.fc40.x86_64 conflicts with file from package tuned-ppd-2.23.0-1.fc40.noarch
<@zbyszek:fedora.im>
19:33:57
Transaction failed: Rpm transaction failed.
<@conan_kudo:matrix.org>
19:34:16
in the change discussion I described how to fix it
<@nirik:matrix.scrye.com>
19:34:19
yep.
<@zbyszek:fedora.im>
19:34:42
Anyway, I think we're going into the weeds now.
<@zbyszek:fedora.im>
19:34:54
I'm sure the packaging details will get resolved.
<@sgallagh:fedora.im>
19:35:30
Fabio Valentini: So your main concern is that we're replacing something lightweight with something heavier?
<@conan_kudo:matrix.org>
19:35:51
honestly, at the end of the day, I don't care which one we use, but I do consider it embarrassing that we can't match Windows on the same machine for power management
<@conan_kudo:matrix.org>
19:36:07
and I'm all for trying something that could potentially let us get better at this
<@sgallagh:fedora.im>
19:36:13
Conan Kudo: And the counter-argument is that PPD itself is too lightweight to be effective on a non-trivial amount of hardware?
<@decathorpe:fedora.im>
19:36:16
that's one concern, yes
<@conan_kudo:matrix.org>
19:36:21
yes
<@sgallagh:fedora.im>
19:36:47
Fabio Valentini: Do you have others?
<@decathorpe:fedora.im>
19:37:16
that is the most substantial one.
<@decathorpe:fedora.im>
19:37:25
I would be happer if it weren't Python ...
<@zbyszek:fedora.im>
19:37:31
FWIW, the approach to make tuned-ppd a separate service is a bit strange. It effectively doubles the memory usage.
<@zbyszek:fedora.im>
19:38:04
It probably makes the whole thing slower and more fragile too.
<@conan_kudo:matrix.org>
19:39:47
that was done specifically to make it swappable
<@conan_kudo:matrix.org>
19:40:03
and so that tuned itself wasn't constrained by ppd's limited interface
<@conan_kudo:matrix.org>
19:40:17
at least that's what Kate mentioned when I talked to her about it before
<@sgallagh:fedora.im>
19:40:31
I kinda wish that PPD didn't resume maintenance; if it remained dead upstream, that would be so much simpler...
<@conan_kudo:matrix.org>
19:40:41
yes
<@conan_kudo:matrix.org>
19:41:00
tbh, I really don't care that it's Python
<@conan_kudo:matrix.org>
19:41:13
that's no worse than many of the Windows power management things being written in C#
<@conan_kudo:matrix.org>
19:41:25
they still beat us even with all that
<@zbyszek:fedora.im>
19:41:29
That seems totally bogus. The requirement is that the dbus endpoint is provided. If we pipe it through a second daemon, this changes nothing to how it works and what can be pushed through it.
<@salimma:fedora.im>
19:41:47
C# should be more performant though
<@conan_kudo:matrix.org>
19:42:04
we could also run Python through Nuitka if it really mattered
<@conan_kudo:matrix.org>
19:42:12
then it would be compiled
<@zbyszek:fedora.im>
19:42:15
I'm pretty sure we don't want C# ;)
<@salimma:fedora.im>
19:42:31
Oh indee
<@salimma:fedora.im>
19:42:37
Oh indeed
<@conan_kudo:matrix.org>
19:42:57
again, I don't really care, I just know what we're doing now isn't working
<@salimma:fedora.im>
19:43:01
On windows you already have it anyway
<@nirik:matrix.scrye.com>
19:43:04
so, should we vote here? or is there any further info we need?
<@sgallagh:fedora.im>
19:43:28
To be clear, we're voting on what the default power management software on Fedora will be. We're NOT voting to drop PPD forcibly from the distro.
<@sgallagh:fedora.im>
19:43:34
Correct?
<@conan_kudo:matrix.org>
19:43:37
yes
<@nirik:matrix.scrye.com>
19:43:47
right.
<@conan_kudo:matrix.org>
19:44:06
I think tuned-ppd is probably going to have more chances to work on ARM than ppd will
<@nirik:matrix.scrye.com>
19:44:08
well, default power management software on desktop media...
<@sgallagh:fedora.im>
19:44:27
Right, tuned is already the default for Server, last I checked
<@conan_kudo:matrix.org>
19:44:33
yup
<@conan_kudo:matrix.org>
19:44:46
and iirc, tuned-ppd is replacing ppd in RHEL 10
<@conan_kudo:matrix.org>
19:44:59
so I would hope that means _someone_ cares about perf for it
<@sgallagh:fedora.im>
19:45:03
That is what I see in CentOS Stream 10 right now, yes
<@sgallagh:fedora.im>
19:46:04
OK, at this point, I'm going to call the vote for this Change.
<@sgallagh:fedora.im>
19:46:18
Unless anyone feels there's an argument that has not been made yet?
<@zbyszek:fedora.im>
19:47:15
IIUC, the votes from the ticket stand, unless explicitly changed?
<@sgallagh:fedora.im>
19:47:21
Proposed: FESCo approves the Change to use tuned as the default power-management tool for desktop installs. P-P-D remains in the distribution as an alternative that can be manually installed.
<@nirik:matrix.scrye.com>
19:48:10
I'm +1 (hadn't voted in ticket yet as I was out last week)
<@conan_kudo:matrix.org>
19:48:12
Stephen Gallagher: +1
<@decathorpe:fedora.im>
19:48:28
still -1
<@zbyszek:fedora.im>
19:48:40
Still +1
<@salimma:fedora.im>
19:48:50
+1
<@sgallagh:fedora.im>
19:48:50
zbyszek: "Any FESCo member is assumed to retain the last vote that they made in the ticket prior to the meeting. Note that this does not carry over to new proposals, with the exception of clarification edits." So yes, ticket votes remain since we aren't changing the proposal
<@sgallagh:fedora.im>
19:48:59
I'm +1
<@salimma:fedora.im>
19:49:26
Still +1
<@sgallagh:fedora.im>
19:50:15
#agreed FESCo approves the Change to use tuned as the default power-management tool for desktop installs. P-P-D remains in the distribution as an alternative that can be manually installed. (+7, 0, -1)
<@sgallagh:fedora.im>
19:50:30
I count only Josh as not voting. Confirmed?
<@sgallagh:fedora.im>
19:50:59
!topic Next week's chair
<@sgallagh:fedora.im>
19:51:35
Who wants it?
<@salimma:fedora.im>
19:52:11
I'll be jet lagged but I can do the week after
<@sgallagh:fedora.im>
19:52:42
Actually, maybe we should hold off on selecting the chair until we know the results of the WhenIsGood.
<@zbyszek:fedora.im>
19:53:08
But we need somebody to make a decision based on the poll…
<@zbyszek:fedora.im>
19:53:12
I can take it.
<@sgallagh:fedora.im>
19:53:16
OK, thanks
<@sgallagh:fedora.im>
19:53:23
!action zbyszek to chair the next meeting
<@sgallagh:fedora.im>
19:53:30
!topic Open Floor
<@zbyszek:fedora.im>
19:53:34
I have something.
<@sgallagh:fedora.im>
19:53:41
Go ahead
<@zbyszek:fedora.im>
19:54:15
https://pagure.io/fork/zbyszek/fesco/fesco-docs/c/37ec880b3c3608529cbb81761ce01de60429fa46 is in a pull request that implements a policy we approved a while back. Please comment on the pull request. If there are no comments, I'll merge it.
<@zbyszek:fedora.im>
19:54:28
(I.e. not now, but let's say tomorrow.)
<@sgallagh:fedora.im>
19:54:46
That's not the MR
<@sgallagh:fedora.im>
19:55:03
https://pagure.io/fesco/fesco-docs/pull-request/90
<@sgallagh:fedora.im>
19:55:16
And I thought I approved that last week, but I must not have hit submit
<@zbyszek:fedora.im>
19:55:30
Yeah, that's the non-whitespace part of that MR.
<@nirik:matrix.scrye.com>
19:56:02
LGTM. 🛳️it. :)
<@conan_kudo:matrix.org>
19:56:13
yup ship
<@salimma:fedora.im>
19:56:19
Iacan check in an hour
<@salimma:fedora.im>
19:56:25
But yeah don't wait for me
<@sgallagh:fedora.im>
19:56:28
I hit merge
<@salimma:fedora.im>
19:56:32
I can check in an hour
<@zbyszek:fedora.im>
19:56:42
OK, great.
<@sgallagh:fedora.im>
19:56:43
Sorry about that; I really thought I did that already.
<@zbyszek:fedora.im>
19:57:02
NP. I forgot too.
<@zbyszek:fedora.im>
19:57:03
Next:
<@zbyszek:fedora.im>
19:57:12
https://fedorapeople.org/groups/schedule/f-41/f-41-key-tasks.html says that we have a few deadlines coming up.
<@zbyszek:fedora.im>
19:57:19
4 Change Checkpoint: Proposal submission deadline (System Wide Changes) Tue 2024-06-25 Tue 2024-06-25 0
<@zbyszek:fedora.im>
19:57:19
3 Change Checkpoint: Proposal submission deadline (Changes requiring mass rebuild) Tue 2024-06-25 Tue 2024-06-25 0
<@zbyszek:fedora.im>
19:57:19
5 Change Checkpoint: Proposal submission deadline (Self Contained Changes) Tue 2024-07-16 Tue 2024-07-16 0
<@zbyszek:fedora.im>
19:57:19
2 Change Checkpoint: Proposal submission deadline (Changes requiring infrastructure changes) Wed 2024-06-19 Wed 2024-06-19 0
<@zbyszek:fedora.im>
19:57:43
Just a reminder to everyone.
<@sgallagh:fedora.im>
19:58:25
I'll try to get that done tomorrow
<@zbyszek:fedora.im>
19:58:33
What's ODCS?
<@conan_kudo:matrix.org>
19:58:40
On Demand Compose Service
<@sgallagh:fedora.im>
19:58:45
On-Demand Compose Service, but it's dead
<@nirik:matrix.scrye.com>
19:58:46
yes please. I was going to ask you about doing that.
<@nirik:matrix.scrye.com>
19:58:53
I am not sure it needs a change tho?
<@conan_kudo:matrix.org>
19:59:08
it does because it will massively change how ELN works
<@salimma:fedora.im>
19:59:16
I'm going to try and finish my mock deprecation before putting up new proposals 's 😅
<@sgallagh:fedora.im>
19:59:25
nirik: It probably requires coordination with releng, so better safe than sorry
<@salimma:fedora.im>
19:59:25
I'm going to try and finish my mock deprecation before putting up new proposals 😅
<@nirik:matrix.scrye.com>
19:59:30
oh, how so?
<@conan_kudo:matrix.org>
19:59:34
everything currently goes through ODCS, but if that goes away, the whole releng path for ELN is completely different
<@nirik:matrix.scrye.com>
19:59:46
odcs just calls pungi.
<@nirik:matrix.scrye.com>
19:59:54
we are just changing it to call pungi directly.
<@conan_kudo:matrix.org>
20:00:13
IIRC our content filtering thingy is in ODCS
<@nirik:matrix.scrye.com>
20:00:35
I'd have to look, but I think it just passes that to pungi config.
<@sgallagh:fedora.im>
20:00:36
No, that's `prepopulate.json` and that's a pungi feature
<@sgallagh:fedora.im>
20:00:47
So I think we're okay there
<@nirik:matrix.scrye.com>
20:01:05
so really I don't think anyone externally will notice...
<@conan_kudo:matrix.org>
20:01:06
we don't produce ISOs, but we do produce VM images
<@conan_kudo:matrix.org>
20:01:13
do the VM images go through ODCS?
<@nirik:matrix.scrye.com>
20:01:21
yes, pungi makes them
<@sgallagh:fedora.im>
20:01:24
Again, they go through Pungi
<@conan_kudo:matrix.org>
20:01:31
I know that davdunc has it on his todo to redo the ELN Guest image
<@conan_kudo:matrix.org>
20:01:41
because it currently uses tooling we are trying to get rid of
<@nirik:matrix.scrye.com>
20:02:11
yeah, and taking odcs out of the picture means we can adjust the pungi config as we like instead of having to poke odcs to make it like we want.
<@conan_kudo:matrix.org>
20:02:11
yes but it uses imgfac :(
<@sgallagh:fedora.im>
20:02:12
And we do produce at least one ISO image: netinst
<@sgallagh:fedora.im>
20:02:27
We produce qcow2 images
<@sgallagh:fedora.im>
20:02:33
We don't have an AMI process
<@conan_kudo:matrix.org>
20:02:41
this cycle, Fedora Cloud is trying to migrate all users of imgfac away so we can shut it down
<@conan_kudo:matrix.org>
20:03:04
we're starting to contact everyone and provide their options and what assistance we can offer
<@sgallagh:fedora.im>
20:03:56
Anyway, I don't think we need to plan the entire effort here. I'll start a Change proposal for ELN
<@conan_kudo:matrix.org>
20:04:03
sounds good
<@sgallagh:fedora.im>
20:04:29
!action @sgallagh to submit a Change to migrate Fedora ELN away from ODCS
<@sgallagh:fedora.im>
20:04:37
Anything else for Open Floor?
<@sgallagh:fedora.im>
20:05:52
OK, I'll take that to be "no".
<@sgallagh:fedora.im>
20:06:14
Thank you for participating, folks. Don't forget to answer the WhenIsGood by Friday!
<@sgallagh:fedora.im>
20:06:17
!endmeeting