- Friday, July 12, 2019

release-team@conference.openafs.org

Friday, July 12, 2019< ^ >

Room Configuration

Room Occupants

GMT+0
[12:01:58] mbarbosa joins the room
[13:50:56] meffie joins the room
[14:41:08] kaduk@jabber.openafs.org/barnowl joins the room
[14:55:31] wiesand joins the room
[14:58:21] <meffie> good day wiesand
[14:58:33] <meffie> good morning kaduk
[14:59:12] <wiesand> hi
[15:00:23] <kaduk@jabber.openafs.org/barnowl> greetings
[15:05:08] <wiesand> this is about my worst day i've had for years
[15:05:17] <meffie> oh, sorry.
[15:05:39] <wiesand> i'm royally upset with the flakes
[15:05:43] <kaduk@jabber.openafs.org/barnowl> sorry to hear that
[15:07:13] <wiesand> so don't take me serious today - and the flakes also consumed the hours i had planned to spend on openafs this week
[15:08:37] <kaduk@jabber.openafs.org/barnowl> Should we just skip 1.8 today, then?
[15:08:49] <wiesand> noticed quite a bit of churn even on 1.8.x lately, but had no chance to follow up yet - do you have an "executive summary for me"?
[15:09:04] <wiesand> otherwise, yes, let's skip 1.8 today
[15:09:11] <kaduk@jabber.openafs.org/barnowl> Mike noticed some things that ought to be on 1.8 when working on the
1.9 NEWS
[15:09:36] <wiesand> I vow I'll find time during the weekend - the flakes are unavailable then anyway
[15:09:43] <kaduk@jabber.openafs.org/barnowl> mostly just the tidying up of ptserver -restricted mode, IIRC, but
maybe I'm forgetting something
[15:09:47] <meffie> yes, stuff already on the master branch that look like bug fixes.
[15:10:08] <wiesand> thanks for having those on your radar!
[15:10:37] <meffie> i just posted my complete list to the release team mail list.
[15:10:51] <meffie> i assume some of these could wait for 1.8.5 if you like.
[15:11:20] <wiesand> i'll have a look, but indeed P1 is to get out 1.8.4pre1
[15:11:41] <meffie> understood.
[15:12:26] <wiesand> noticed you also worked on the wish list - thanks again!
[15:12:39] <meffie> I also made some updates to the 1.8.x NEWS file for the things Stephan pointed out in gerrit.
[15:13:26] <kaduk@jabber.openafs.org/barnowl> I hadn't noticed that the 1.8.4 news got updated, too (just grabbed
the latest) -- thanks!
[15:13:38] <meffie> no problem
[15:14:20] <meffie> yes, i think the "three letter cell name" fixes are ready.
[15:15:25] <wiesand> That's good, thanks.
[15:15:39] <meffie> sorry for the long summary.
[15:15:48] <wiesand> I still wonder whether the "retsigtype" removal belongs into news.
[15:15:49] <kaduk@jabber.openafs.org/barnowl> (edit: I forgot the viced option parsing fixes from the list of new 1.8
pullups)
[15:17:00] <wiesand> And I still think 13584 is just fixing fallout of other changes new to 1.8.4pre1 and thus should either be mentioned along with those or not at all.
[15:17:02] <kaduk@jabber.openafs.org/barnowl> retsigtype removal seems relevant only to developers and only to those
who happen to have ancient platforms around
[15:17:37] <meffie> i will update the NEWS to avoid retsigtype text
[15:17:54] <wiesand> (these nit aside I *am* grateful for the help!)
[15:18:52] <wiesand> nits
[15:19:20] <meffie> remote: Updated Changes:
remote:   https://gerrit.openafs.org/13643 Update NEWS for 1.8.4pre1
[15:20:11] <wiesand> thanks
[15:21:01] <meffie> you are welcome.
[15:21:29] <meffie> too bad we dont have a tool to magically create an NEWS file from git log :)
[15:22:52] <kaduk@jabber.openafs.org/barnowl> :)
[15:23:18] <meffie> wiesand: anything else you need for 1.8.4pre1?
[15:23:47] <wiesand> We have Changelog. IMO the whole point of NEWS is to translate that into something suitable for users/admins w/o intimate knowledge of the code base
[15:24:10] <meffie> yes, i was just kidding. sorry.
[15:24:17] <wiesand> And that's where I have an obvious advantage :)
[15:25:16] <meffie> yes, i agree we need a NEWS file with human readable text. i'm just punchy from working on them for 1.8.x and 1.9.x.
[15:25:50] <kaduk@jabber.openafs.org/barnowl> It's friday noon in ... 35 minutes ;)
[15:26:16] <wiesand> Mike: I think there's already more available for 1.8.4pre1 than I would have imagined a few weeks ago. Thanks.
[15:26:52] <meffie> ok, thanks.
[15:27:03] <wiesand> Ben: looks like you're eager to get on to master/1.9.0 - please go ahead
[15:27:29] <kaduk@jabber.openafs.org/barnowl> I dunno about "eager", just wanted to keep the meeting from stalling
[15:28:48] <kaduk@jabber.openafs.org/barnowl> A bunch more new stuff showed up in gerrit, that i only partially
caught up with
[15:30:31] <kaduk@jabber.openafs.org/barnowl> But I got to merge a bunch of things at least
[15:30:38] <meffie> yay
[15:31:13] <kaduk@jabber.openafs.org/barnowl> (including Andrew's afs_linux_read_cache page-unlocking fix and the
dcache length verification stuff)
[15:32:39] <meffie> yay, thank you.
[15:33:23] <wiesand> NB looking forward to Yadav joining us
[15:33:32] <kaduk@jabber.openafs.org/barnowl> Now that debian 'buster' released, the krb5 package in unstable has
dropped support for single-DES entirely.  In the process of doing that
the maintainer noticed that openafs still calls
krb5_enable_weak_crypto() (and friends) and wondered if he was going
to totally break us.
[15:33:50] <wiesand> ouch
[15:34:12] <kaduk@jabber.openafs.org/barnowl> So in response to that I submitted 13689 to at least make those calls
conditional, so it's more clear what's going on
[15:35:43] <wiesand> makes sense to me at first glance
[15:35:44] <kaduk@jabber.openafs.org/barnowl> Mike has already reviewed, so thanks :)
[15:35:45] <meffie> ok, thanks.
[15:36:11] <meffie> yes, not sure if my vote is any good, but it looks like it makes sense to me.
[15:36:24] <kaduk@jabber.openafs.org/barnowl> I kind of want to follow up in master/1.9-only that removew the calls
(and option) entirely, but maybe that's too fast
[15:36:49] <kaduk@jabber.openafs.org/barnowl> I mostly  am worried that I missed a call site or have bogus
documentation or something
[15:38:02] <wiesand> Then chances are there'll be both missed call sites and bogus  docs ;-) Fine for 2.0 IMHO.
[15:38:28] <kaduk@jabber.openafs.org/barnowl> Anyway, among the other new stuff in gerrit that I didn't look at
closely we have a bunch of cleanup/build-robustness stuff from Andrew,
which is nice but probably not urgent.  Do people know of anything
that is particularly urgent?
[15:38:58] <wiesand> not me, but htta's hardly surprising
[15:40:34] <kaduk@jabber.openafs.org/barnowl> Other than trying to keep working on stuff as it comes in, I think the
plan is to push 1.9.0 once the 1.8.4 NEWS solidifies (since we are
effectively rebased ontop of it) -- Mike's NEWS draft has done all the
hard work, and there's just the mechanical version number bumps
and release work to do.
[15:40:55] <kaduk@jabber.openafs.org/barnowl> Though I guess the announcement should probably have some sketch of
instructions for how to test the things we want testing of...
[15:42:20] <wiesand> right
[15:42:22] <meffie> nothing urgent.
[15:43:37] <kaduk@jabber.openafs.org/barnowl> I don't think I have anything else for this week.
[15:43:47] <wiesand> Re Yadav: his TZ is GMT+5:30 all year round?
[15:44:14] <meffie> i think so. we will need to ask him
[15:46:06] <wiesand> Having parcticipants from -9:00 to +5:30 is "interesting".
[15:46:29] <kaduk@jabber.openafs.org/barnowl> Almost makes me wonder if we should rotate through meeting times so
everyone gets inconvenienced some of the time.
[15:46:51] <wiesand> So is having 3 parties with slightly different (or no) DST rules
[15:47:42] <meffie> i think he is used to having meetings with people in different time zones
[15:48:10] <meffie> but it is good to ask for preferences.
[15:48:23] <wiesand> sure
[15:48:55] <wiesand> Nb this is one fo those cases of "problems I'd rather have than not" :)
[15:49:16] <meffie> in other news, cheyenne is testing (locally) a change to the buildbot config to create a "force gerrit number" build.
[15:49:48] <wiesand> [wants that badly]
[15:49:51] <meffie> that should let us replay gerrit verification
[15:50:05] <meffie> me too.
[15:50:13] <kaduk@jabber.openafs.org/barnowl> That is pretty exciting, yes
[15:51:28] <meffie> he's going to setup a local gerrit to test it better.
[15:51:40] <meffie> before we make any changes to the buildbot.
[15:51:51] <wiesand> Nice plan :)
[15:51:58] <meffie> thanks :)
[15:52:04] <wiesand> Anything else to discuss today?
[15:52:14] <kaduk@jabber.openafs.org/barnowl> Oh
[15:52:21] <wiesand> oh?
[15:52:25] <meffie> oh?
[15:52:32] <kaduk@jabber.openafs.org/barnowl> Mike's note about local gerrit for the buildbot testing reminded me
that I'm still supposed to take more gerrit updates
[15:52:37] <wiesand> oh? ?
[15:52:49] <kaduk@jabber.openafs.org/barnowl> Apparently there's a 3.0.0 now and it sounds like a very different
appearance/UI
[15:52:59] <meffie> oh great ! :(
[15:53:08] <wiesand> .0.0 sounds scary ;-)
[15:53:31] <kaduk@jabber.openafs.org/barnowl> But I was thinking of stopping at a 2.15 or 2.16 release for now
[15:53:36] <kaduk@jabber.openafs.org/barnowl> (https://www.gerritcodereview.com/releases-readme.html)
[15:55:08] <kaduk@jabber.openafs.org/barnowl> Anyway, no solid ETA for scheduling the upgrade, but it probably will
not be during business hours for anyone here.
[15:55:43] <wiesand> Chances are we'd cope, so if 3.x is the way forward, let's face it
[15:56:05] <meffie> kaduk, if you are spending some time on gerrit, we should consider making a new/modern ssh keys for the buildbot and an account that does not use Jason's personal domain in the name.
[15:56:36] <kaduk@jabber.openafs.org/barnowl> I think when I skimmed repo-discuss they said they'd even keep a
version of 2.15 around for a while since some sites couldn't upgrade
easily.  (We may be one of them, actually, since I don't  think I have
an easy path to java 9)
[15:57:05] <kaduk@jabber.openafs.org/barnowl> That's a good point; we had talked about a new email account at least
a while ago and  I forgot to do anything  about it
[15:57:10] <meffie> currently we are using ssh-dss key type
[15:58:47] <meffie> and return mail address is <buildbot@rampaginggeek.com>
[15:59:13] <kaduk@jabber.openafs.org/barnowl> indeed
[15:59:14] <meffie> i'd think that should be buildbot@openafs.org or some-such.
[15:59:51] <kaduk@jabber.openafs.org/barnowl> Something to keep in mind.  Anything else for today?
[16:00:01] <wiesand> yes, and DSA keys should probably phased outr too…
[16:00:02] <meffie> that's all.
[16:00:33] <wiesand> Friday noon was … six hours ago… adjourn?
[16:00:36] <kaduk@jabber.openafs.org/barnowl> Okay. Thanks everyone!
[16:00:48] <meffie> (i'll gen new keys and send a public key)
[16:00:53] <wiesand> Thanks a lot Ben & Mike!
[16:01:15] wiesand leaves the room
[16:01:17] <meffie> thanks!
[16:01:48] <meffie> have a good weekend.
[16:07:05] meffie leaves the room
[19:16:46] kaduk@jabber.openafs.org/barnowl leaves the room
[19:59:39] mbarbosa leaves the room
[19:59:42] mbarbosa joins the room
[20:43:49] mbarbosa leaves the room
[20:43:49] mbarbosa joins the room
[21:40:14] mbarbosa leaves the room