- Thursday, February 17, 2022

release-team@conference.openafs.org

Thursday, February 17, 2022< ^ >

kaduk@jabber.openafs.org/barnowl has set the subject to: openafs release team

Room Configuration

Room Occupants

GMT+0
[03:01:33] mvitale joins the room
[03:43:49] mvitale leaves the room: Replaced by new connection
[03:44:23] mvitale joins the room
[03:56:41] kaduk@jabber.openafs.org/barnowl joins the room
[16:58:39] meffie joins the room
[17:00:34] wiesand joins the room
[17:00:53] <meffie> greatings.
[17:00:53] <meffie> greetings
[17:01:07] <kaduk@jabber.openafs.org/barnowl> hello
[17:01:10] <wiesand> hi
[17:01:53] <mvitale> hello Ben!
[17:02:06] <mvitale> and Stephan!
[17:02:46] <mvitale> I have a new Mac M1 - my old chat client has got some quirks under emulation
[17:04:17] <wiesand> Ben (&Mark): Since changes are required to cope with Linux 5.17, we discussed last week wether to go for  a swift 1.8.8.2 (maybe sprinkling in some FBSD and Solaris specific changes) - would you agree?
[17:04:46] <meffie> i failed to take notes last week :(
[17:05:05] <wiesand> Well, there wasn't all that much to jot down…
[17:05:12] Cheyenne joins the room
[17:05:20] <Cheyenne> Hello all
[17:05:31] <mvitale> fine with me
[17:05:39] <kaduk@jabber.openafs.org/barnowl> A 1.8.8.2 would be reasonable, sure
[17:06:38] <wiesand> ok, that's what we'll go for then
[17:06:40] <mvitale> oh, there is a solaris change I wanted in 1.8.x, I wonder if it is already on your list?
[17:07:16] <mvitale> https://gerrit.openafs.org/#/c/13344/
[17:07:19] <wiesand> which one?
[17:07:47] <Cheyenne> Oh.. there is a master commit that deals with parallel makes, gerrit 14137 that I would like to see pulled in
[17:08:23] <Cheyenne> (simple makefile change)
[17:08:24] <mvitale> allows solaris builds to be more parallel - could allow speeding up our slow solaris builders
[17:09:15] <mvitale> oh yes 14137 as well
[17:09:47] <kaduk@jabber.openafs.org/barnowl> having opened them to try to refresh my memory of what they are, that
seems unproblematic for 1.8.8.2 to me (on first look, at least)
[17:10:12] <mvitale> ( this client is not autoscrolling with new messages, I have to manually drag the scroller each time)
[17:10:53] <wiesand> adium?
[17:11:41] <wiesand> 13344 and 14137 were not on my list yet, but seem good for 1.8.8.1
[17:11:46] <mvitale> yup
[17:12:40] <wiesand> I have the same problem since the update to Catalina, even w/o emulation.
[17:13:54] <wiesand> And I have a shiny new M1 Mac too, and I like it a lot, but am still using the old one for many tasks.
[17:14:01] <kaduk@jabber.openafs.org/barnowl> We could revisit the "what chat technology" question again if people
have other preferences that "work better"
[17:15:27] <mvitale> no, Ben, jabber is fine, it's Adium that is the issue
[17:15:41] <wiesand> Jabber is still a good choice IMO
[17:15:41] <kaduk@jabber.openafs.org/barnowl> ok
[17:15:59] <mvitale> the only reason I use it is because it supports both IRC and XMPP in one client
[17:16:11] <mvitale> but it does neither one well
[17:16:47] <mvitale> sorry to hijack the discussion
[17:17:06] <wiesand> np
[17:17:07] <Cheyenne> I've been using a jabber/irc gateway package (bitlbee) -- sets up a local "irc" server that connects to a jabber server
[17:17:27] <kaduk@jabber.openafs.org/barnowl> that's like why I use barnowl -- it does xmpp and irc and zephyr in
the same client, but only zephyr well
[17:19:48] <Cheyenne> Anyway -- I ran a linux-5.17-rc4+ build yesterday against master with no problems.
[17:20:46] <mvitale> thank you Cheyenne
[17:21:16] <kaduk@jabber.openafs.org/barnowl> yes, thanks
[17:21:28] <wiesand> thanks Cheyenne!
[17:21:31] <kaduk@jabber.openafs.org/barnowl> are we still seeking other items for a 1.8.8.2 or do we think the
contents are pretty well settled?
[17:21:58] <meffie> i have nothing
[17:22:12] <wiesand> I think they are
[17:22:33] <mvitale> nothing for 1.8.8.2
[17:23:03] <wiesand> the "hot" thing this week is probably 14898
[17:23:09] <Cheyenne> wiesand: do you want me to pull in the 5.17 patches for 1.8? or do you want to take care of that?
[17:23:13] <mvitale> but after it is out, perhaps we could update our Solaris builders to run builds with more parallelization
[17:24:20] <wiesand> Cheyenne: please let me try
[17:24:44] <Cheyenne> Okay.. once you have the commits posted, I can run the build test
[17:25:19] <meffie> yeah, we have a deadline of tomorrow for 14898
[17:25:42] <kaduk@jabber.openafs.org/barnowl> Ah, yeah, Dave pinged me about gsoc topics, but this week was a busy
IESG telechat (ended just before this meeting) so I didn't get to
think about it at all
[17:25:52] <kaduk@jabber.openafs.org/barnowl> (only five weeks left in my term as SEC AD!)
[17:26:03] <meffie> yay!!
[17:26:44] <meffie> i for one am stoked to update all my old gerrits and get them merged! :) :)
[17:26:46] <kaduk@jabber.openafs.org/barnowl> It will be good to have more time for AFS, yes
[17:27:17] <wiesand> yay!
[17:27:52] <wiesand> ok, I'll merge and bring live 14898 this evening
[17:27:59] <meffie> for 14898, i believe we can add more project ideas (in fact google says we should be doing it year round), but we need to have an initial set asap for the foundation to submit the application, which is due on 2/21
[17:28:13] <kaduk@jabber.openafs.org/barnowl> (but there is a lot to try to pack in to those last five weeks)
[17:28:20] <meffie> thank you wiesand!
[17:31:03] <wiesand> that's about what I have for today
[17:31:50] <Cheyenne> Oh.. I will be on vacation next Thursday, but I'll give Mike a 5.17 status before I head out (Wednesday)
[17:31:54] <meffie> kaduk@jabber.openafs.org/barnowl, i added 'fileserver autotune' as a gsoc idea, since i was not able to make progress on that.
[17:32:26] <kaduk@jabber.openafs.org/barnowl> +1
[17:32:51] <meffie> ok, i wasnt sure if you agreed that would make a good project idea.
[17:32:52] <wiesand> makes sense
[17:33:46] <meffie> i decided to add it, then big forgiveness if it didnt make sense.
[17:34:05] <meffie> s/big/beg/
[17:34:26] <kaduk@jabber.openafs.org/barnowl> the foundation would probably need to shell out for some resources to
be able to have a decent benchmarking rig, but I think that there are
relatively straightforward things to do in that space in terms of
running the experiment and collecting data on what effect the various
parameters have.
[17:35:12] <kaduk@jabber.openafs.org/barnowl> scraping through old gerrits may be a good source of inspiration
[17:35:23] <meffie> we have a box we could loan.
[17:35:35] <meffie> (we use for perf testing)
[17:36:04] <meffie> that was my thinking too.
[17:36:35] <meffie> i did start going thru old gerrits, which triggered that idea.
[17:37:37] <kaduk@jabber.openafs.org/barnowl> the first thing that came to mind was how to get a bunch of different
clients to exercise that part of the server, but having a beefy server
to exercise is also useful :)
[17:37:56] <kaduk@jabber.openafs.org/barnowl> any hot topics for master this week?
[17:38:24] <kaduk@jabber.openafs.org/barnowl> I think I owe Andrew a response on 14041 (ubik tests)
[17:39:23] <kaduk@jabber.openafs.org/barnowl> and it looks like I had 14822 and 14861 open that I should probably
actually do something with
[17:39:32] <meffie> i dont think anything hot, just the regular backlog
[17:40:20] <mvitale> thank you
[17:40:24] <meffie> i guess if we are doing a 1.8.8.2, should we include gerrit 14842?
[17:41:30] <kaduk@jabber.openafs.org/barnowl> sigh, I should pester jhutz again.  He did agree (IIRC) in principle
that a new GCO CellServDB release was in order.
[17:41:46] <meffie> ah, yes that makes sense.
[17:42:37] <Cheyenne> there's a small update to the curses autoconf (I need to clean up some trailing blanks).. -- this one is a nod to the openSUSE folks.
[17:42:52] <Cheyenne> (14889)
[17:43:06] <mvitale> yes, there are a number of issues with the current CSDB
[17:44:00] <kaduk@jabber.openafs.org/barnowl> ok, I will try to take a look at 14889
[17:44:06] <kaduk@jabber.openafs.org/barnowl> (it is small :)
[17:45:57] <wiesand> wasn't there an open build fix for gtx too?
[17:46:53] <meffie> yes, looking for the number
[17:47:01] <Cheyenne> already ported -- 12760
[17:47:11] <Cheyenne> 12763
[17:48:15] <wiesand> all the better
[17:48:17] <Cheyenne> I stumbled onto 12760 thinking it addressed the same problem as 14889
[17:48:45] <kaduk@jabber.openafs.org/barnowl> ah
[17:49:32] <Cheyenne> On a different distro related topic.. the gentoo folks have some patches dealing with a kernel built using clang
[17:49:51] <kaduk@jabber.openafs.org/barnowl> ooh, exciting
[17:50:01] <Cheyenne> currently our build forces the kernel build to use gcc
[17:50:20] <kaduk@jabber.openafs.org/barnowl> on a different different distro related topic, the debian kernel
package maintainer noticed that the current kernel breaks the 1.8.8
package, so I sould pull in 1.8.8.1 "soon"
[17:50:46] <meffie> Cheyenne: see the message i just added to 12760 ? i think this sort of backref is helpful.
[17:51:15] <Cheyenne> yes -- thanks
[17:51:38] <Cheyenne> meffie: I thought someone was looking at doing that automagically..
[17:51:58] <meffie> no, it is manual. feel free to add the backrefs on ones you find.
[17:52:47] <meffie> (if you can automate it, go crazy ;)
[17:53:17] <kaduk@jabber.openafs.org/barnowl> I think that modern gerrit might be able to cope with using the same
Change-ID for different branches, which would be a way to
automatically get this sort of visibility.  That might be a thing to
investigate.
[17:53:38] <meffie> that would be ideal.
[17:53:39] <wiesand> detecting those is simple: git log --grep=/12760 master
[17:55:14] <mvitale> wiesand:  some Adium possible workarounds discussed here: https://forums.cocoaforge.com/viewtopic.php?f=13&t=27155&start=80
[17:55:22] <meffie> hmm, maybe i could update cron job on the GerritsForX pages on wiki.openafs.org to just make a cross ref
[17:55:27] <wiesand> er, git log --grep=/12760 origin/openafs-stable-1_8_x
[17:55:43] <kaduk@jabber.openafs.org/barnowl> Hmm, worth a thought, I suppose
[17:55:50] <kaduk@jabber.openafs.org/barnowl> I have another meeting at the top of the hour -- anything else to
discuss today?
[17:56:00] <meffie> i have nothing.
[17:56:27] <wiesand> nothing here
[17:56:34] <wiesand> Adjourn?
[17:56:45] <kaduk@jabber.openafs.org/barnowl> sounds good.
Thanks everyone!
[17:56:49] <meffie> have a good day, thanks all.
[17:57:09] <wiesand> Thanks a lot everybody! (and Mark: thanks for the pointer)
[17:57:17] <Cheyenne> have a good week everyone
[18:52:22] meffie leaves the room
[19:23:16] wiesand leaves the room
[19:35:23] mvitale leaves the room
[19:42:46] mvitale joins the room
[19:45:42] mvitale leaves the room
[19:47:56] mvitale joins the room
[19:48:32] <mvitale> Adium scrolling workaround is working for me.  Crisis averted.
[20:44:46] meffie joins the room
[20:45:04] meffie leaves the room
[20:57:50] meffie joins the room
[21:32:24] meffie leaves the room