- Wednesday, August 26, 2015

release-team@conference.openafs.org

Wednesday, August 26, 2015< ^ >

Room Configuration

Room Occupants

GMT+0
[07:47:07] stephan.wiesand joins the room
[07:47:28] <stephan.wiesand> test test
[07:57:47] stephan.wiesand leaves the room
[13:20:32] am i me?
[13:20:38] nope :(
[13:21:09] Daria Brashear joins the room
[13:45:29] meffie joins the room
[13:54:14] stephan.wiesand joins the room
[13:55:16] <stephan.wiesand> test test test
[13:56:05] <Daria Brashear> hi
[13:56:29] <Daria Brashear> jeff is beside me but says he has no jabber server. whch i guess is true since the disk containing it is on the other side of me
[13:56:48] <stephan.wiesand> Hi
[13:56:50] kaduk joins the room
[13:58:03] <stephan.wiesand> Hi Ben
[13:58:24] <stephan.wiesand> Those 800ms sure are noticable
[13:59:02] <kaduk> Hi.
[13:59:27] <stephan.wiesand> So what did I miss last week?
[13:59:36] <stephan.wiesand> (except the fun)
[14:00:23] <Daria Brashear> it was fun. the boat was nice. the band was cool. pittsburgh is still awesome
[14:01:30] <kaduk> Hmm.  Simon and jhutz and I had some bilateral discussions about compound identities for rxgk and probably have a way forward.  Docker can help a lot for spinning up test cells.  kAFS is not dead.
[14:01:30] <meffie> it was a good conference. lots of good talks and smart people.
[14:01:43] <kaduk> It was both good and fun, yes.
[14:01:59] <stephan.wiesand> Pity I missed out.
[14:02:05] <Daria Brashear> yes, we missed you
[14:02:11] <meffie> yes
[14:02:11] <kaduk> jhutz wants to not be an afs3-stds co-chair any more; someone should throw more hats in the ring.
[14:02:19] <kaduk> We did miss you, yes.
[14:02:21] Daria Brashear backs slowly away
[14:02:40] <meffie> maybe someone from mit? :)
[14:02:59] <stephan.wiesand> How does that "backs slowly away" thing work?
[14:03:07] <kaduk> There aren't a whole lot of likely candidates at MIT, but I guess I can ask around.
[14:04:27] <kaduk> IPv6 is still hard.  Some people may try to spin up test cells with my rxgk work-in-progress and see what breaks.  1.8 might happen.
[14:04:46] <stephan.wiesand> kAFS would be a major relief
[14:05:33] <stephan.wiesand> "1.8 might happen" seems actual news to me.
[14:05:42] <stephan.wiesand> I think I had given up on it.
[14:06:14] <kaduk> "But there's only two pages of changes in gerrit left"
[14:06:23] <Daria Brashear> lol
[14:06:44] <stephan.wiesand> We had more than that in the 1.6.x queue at times.
[14:07:13] <stephan.wiesand> But I doubt that's the actual problem.
[14:07:41] <stephan.wiesand> Anyway I'm delighted about the recent activity level.
[14:07:58] <stephan.wiesand> Can we have a BPW every fortnight to keep that up?
[14:08:00] <meffie> bens post was good on -devel. thanks ben!
[14:08:13] <kaduk> I try.
[14:08:38] <Daria Brashear> if someone else wants to organize workshops, i'd be in. like. we were supposed to go to cambridge...
[14:09:15] <stephan.wiesand> Which cambridge?
[14:09:24] <Daria Brashear> uk
[14:09:27] <stephan.wiesand> (MA or UK??
[14:09:37] <Daria Brashear> "the one i can't just get in the car and drive to"
[14:09:52] <stephan.wiesand> "the one I can"
[14:09:56] <Daria Brashear> yup
[14:10:18] <Daria Brashear> well, mostly. i assume they won't let you drive through the chunnel
[14:10:26] <Daria Brashear> and i hope you don't try
[14:10:27] <meffie> is owen still there?
[14:10:49] <Daria Brashear> i assume so. we just got email from him asking for his email addresses on the openafs lists to be normalized to
[14:10:53] <Daria Brashear> (something new that i forget)
[14:10:57] <stephan.wiesand> I can drive onto a railway wagon or a ferry.
[14:11:56] <stephan.wiesand> Before we continue discussing 1.8: any Linux news?
[14:12:18] <Daria Brashear> asked marc
[14:12:26] <kaduk> Apparently someone at Ubuntu is paying some attention to the openafs package
[14:13:09] <kaduk> They picked the four linux-4.2 changes from gerrit but not the allocator changes before it, which I guess caused things to be a bit broken, but then they went back to auto-importing from debian once I pulled the linux-4.2 support in there.
[14:13:12] <stephan.wiesand> Thanks. Daria: You also approved 11945. This was the last blocker I believe, but what about Andrew's comments?
[14:14:01] <stephan.wiesand> Ben: Good news re Ubuntu.
[14:14:24] <stephan.wiesand> Adapting 11951 w/o 11933 shouldn't be hard. But it still needs to be done.
[14:14:46] <Daria Brashear> andrew may be right, but 1) marc would need to comment and honestly 2) it could be done over this if needed
[14:14:47] <stephan.wiesand> That's why I went with pulling up the allocator change.
[14:14:48] <kaduk> For 11945, I currently suspect that we will need to try it and see if it breaks.
[14:15:33] <stephan.wiesand> Fine.
[14:15:50] <stephan.wiesand> Scripts seems to be a candidate...
[14:16:15] <kaduk> It does...
[14:17:42] <stephan.wiesand> So, once 11945 is merged we can pull it up and finalize 1.6.14.1.
[14:18:55] <stephan.wiesand> I expect to have reasonable network connectivity most of the time the next few weeks, and leisure to work on openafs releases.
[14:19:22] <meffie> woot
[14:19:24] <stephan.wiesand> But if I'm unresponsive, please feel free to go ahead w/o me.
[14:20:24] <stephan.wiesand> (I have yet to succeed recharging that prepaid UMTS card or getting a new one)
[14:20:40] <kaduk> (Ah.)
[14:20:56] <stephan.wiesand> 1.6.14.1 seems straightforward at this point, unless there are any bad surprises in Linux 4.2 final.
[14:21:36] <stephan.wiesand> You don't need me for that. But I'll gladly do it, and chances are it will work.
[14:23:33] marc says rc8 happened; release likely this weekend; still on target
[14:24:03] <stephan.wiesand> [I'm on the way to Iceland. Right now the vessel is approaching the Faroese Islands for a short stop-over, and we'll arrive in Seydifjordur tomorrow morning. I'll then stay on the Island for 20 days and have a 3 days jourbey back home]
[14:24:06] <Daria Brashear> marc says rc8 happened; release likely this weekend; still on target
[14:24:22] <stephan.wiesand> Thanks.
[14:24:22] <Daria Brashear> gesundheit
[14:24:34] <stephan.wiesand> ?
[14:24:54] <Daria Brashear> i assume Seydifjordur is sneezing
[14:25:04] <kaduk> The icelandic place names are always a mouthful.
[14:25:15] <kaduk> But I hear it's beautiful there; have fun and relax!
[14:25:29] <stephan.wiesand> It's a nice little town in eastern iceland.
[14:25:47] <stephan.wiesand> Ben: That's exactly the plan.
[14:26:01] <meffie> sounds wonderful!
[14:26:14] <stephan.wiesand> In fact, it has already started :)
[14:26:22] <kaduk> :)
[14:26:41] <stephan.wiesand> And I'm escaping the next heat wave at home.
[14:27:01] <stephan.wiesand> And the one after.
[14:28:13] <stephan.wiesand> Ok, so much for 1.6.14.1. We need 11945 merged, the rest is straightforward.
[14:28:28] <stephan.wiesand> Ben: Want to discuss 1.8 in some more detail?
[14:28:54] <kaduk> Hmm, where to start.
[14:29:15] <kaduk> I guess we had largely left off at the logging changes that Mike was working on.
[14:29:45] <kaduk> I think I made some review comments on those, but don't really remember what since I was going through a lot of changes in a row.
[14:29:48] <meffie> yes, and the softsig conversion.
[14:30:13] <kaduk> Do you remember more about what was needed?
[14:30:18] <meffie> i'll respond to the comments on the logging patches.
[14:30:33] <meffie> the blocking issue for me was not breaking the windows builds.
[14:30:45] <meffie> wrt to softsig
[14:30:50] <kaduk> (11945 needs a rebase; working on it)
[14:31:37] <meffie> i have a windows vm here, i'll work there
[14:32:03] <kaduk> Did you manage to get a windows build environment set up without too much gnashing of teeth?
[14:32:19] <meffie> someone did it for me! :)
[14:32:43] <kaduk> Huzzah
[14:33:51] <meffie> the point is to avoid ifdef hell, of course.
[14:34:58] <kaduk> Of course.
[14:35:08] <meffie> i'm really keen on having the external-log-rotation stuff in, so i'll make it a priority.
[14:35:16] <kaduk> Thanks!
[14:35:46] <kaduk> Now if I can just convince someone (else) to work on converting supergroups from a build-time option to a run-time option...
[14:36:11] <meffie> oh, i didnt realize that was even possible?
[14:36:11] stephan.wiesand leaves the room
[14:36:25] stephan.wiesand joins the room
[14:36:48] <kaduk> I believe it should be possible, though I haven't done enough investigation to be fully confident.
[14:37:16] <stephan.wiesand> Is that realy a "must have" for 1.8?
[14:37:45] <kaduk> I am not inclined to block the release on it, making it just a "nice to have" in my book.
[14:38:06] <stephan.wiesand> It certainly is a "nice to have". But not a blocker.
[14:39:00] <kaduk> Looks like Mark has some Makefile clean rule tweaks that are -1'd; maybe we should see if he can update those (though ISTR that some of the comments were not entirely clear on what they wanted to see happen)
[14:39:07] <stephan.wiesand> And if possible, that feature would even be neat in 1.6.
[14:39:41] <meffie> a working shake-loose-vcaches would be useful too.
[14:39:43] <kaduk> I think it would be a pretty invasive change that might be too big for 1.6.  But I guess we'll see.
[14:40:20] <kaduk> Yes, shake-loose-vcaches would be nice.
[14:40:53] <meffie> i've never heard for a request for a runtime change for supergroups.
[14:41:32] <meffie> i'd think we'd just want to sunset the old format.
[14:41:45] <kaduk> I don't think it's worth making it something you can enable without restarting the processes, but just a command-line argument.
[14:42:32] <kaduk> The motivations I remember were so that all ptservers would recognize the file format and to have the code be always compiled so that static analysis and such pick up on it.
[14:43:39] <kaduk> Anyway, these things should probably be enough to keep us busy this week.
[14:43:40] <meffie> ah, yes.
[14:44:11] <stephan.wiesand> [2h to Tórshavn - and my dinner buffet :) ]
[14:44:12] <kaduk> And I'll try to continue working on merging the backlog of approved changes (help is of course welcome)
[14:44:37] <meffie> excellent.
[14:46:33] <meffie> marcios fix for the softsig unit test execution would be helpful. gerrit 11977
[14:46:43] <stephan.wiesand> Ben, thanks for rebasing 11945.
[14:47:20] <kaduk> Sure thing; just waiting for buildbot to chime in...
[14:49:31] <stephan.wiesand> Are we done for today? Anything else to discuss?
[14:49:39] <kaduk> meffie: My perl is rusty^Wnonexistent; is $Bin some form of magic?
[14:49:50] <kaduk> stephan: I think we're pretty much done
[14:50:39] <meffie> it's a standard perl thing to find yourself
[14:50:51] <kaduk> Ah.
[14:52:00] <meffie> core module: http://perldoc.perl.org/FindBin.html
[14:53:27] <stephan.wiesand> Ok. Thanks a lot for being here today! Please feel free to go on w/o me, but I'll go on deck now and watch the Islands approach. It can be quite a view...
[14:53:49] <meffie> enjoy!
[14:53:50] <kaduk> Possibly even better than the view of Pittsburgh from the river (but don't tell Daria I said that)
[14:54:27] <stephan.wiesand> ok, won't.
[14:54:35] <stephan.wiesand> Thanks again, and bye.
[14:54:37] stephan.wiesand leaves the room
[14:56:12] kaduk leaves the room
[14:57:17] meffie leaves the room
[15:20:29] <jhutz@jis.mit.edu/owl> > asking for his email addresses on the openafs lists to be normalized
... which I should actually do
[15:36:12] Jeffrey Altman joins the room
[18:37:08] Daria Brashear leaves the room