Home
release-team@conference.openafs.org
Wednesday, August 24, 2016< ^ >
Room Configuration
Room Occupants

GMT+0
[05:58:44] Jeffrey Altman leaves the room
[07:11:10] Jeffrey Altman joins the room
[12:40:18] meffie joins the room
[13:32:22] jhg joins the room
[13:58:32] wiesand joins the room
[14:01:58] <meffie> hello
[14:01:58] kadukoafs@gmail.com/barnowl2200C8CA leaves the room
[14:01:58] kadukoafs@gmail.com/barnowl0226D601 leaves the room
[14:01:58] kadukoafs@gmail.com/barnowl2200C8CA leaves the room
[14:01:58] kadukoafs@gmail.com/barnowl2200C8CA leaves the room
[14:01:58] kadukoafs@gmail.com/barnowl0226D601 leaves the room
[14:02:08] <mvita> hello
[14:02:17] <wiesand> Hello
[14:02:32] <mvita> seems to be an echo
[14:02:48] <wiesand> o...o...o...
[14:03:06] <mvita> heh
[14:03:35] <mvita> I needed a laugh just then, tx
[14:03:47] <wiesand> Any news? Good or bad?
[14:03:59] <wiesand> In particular, Linux 4.8?
[14:04:47] <mvita> nothing new to report
[14:05:06] <mvita> how about getcwd?  have you tested the patches?
[14:05:11] <wiesand> No.
[14:05:31] <wiesand> I’m under a deadline and barely looked at openafs for the last week.
[14:05:41] <mvita> oh, sorry
[14:06:04] <wiesand> But Joe just commented in gerrit that he did :)
[14:06:36] <wiesand> Should be good enough, after all we’re planning to have prerelease next.
[14:06:59] <wiesand> So our userbase will test very thoroughly ;-)
[14:07:35] <mvita> good
[14:08:35] <wiesand> There’s not much to discuss from my point of view, since very little has happened since last week. Or did I miss anything?
[14:08:47] <wiesand> In particular, we still seem to have buildbot troubles.
[14:09:30] <wiesand> But thanks for looking at what I did to RECFOUNDDB.
[14:09:50] <meffie> buildbot troubles?
[14:10:02] <mvita> yes, this is the first I've heard
[14:10:05] <mvita> looking now
[14:10:10] <wiesand> It’s ubik though, and I’d like to have more than one +1 before merging it.
[14:10:17] <meffie> there's a number of slaves offline.
[14:11:21] kadukoafs@gmail.com/barnowlD10ED525 joins the room
[14:11:31] <meffie> btw, i have a set of patches to help "make check" on different platforms.
[14:11:53] <meffie> i'm going to add some nightly builders to do make check against master.
[14:13:39] <meffie> > In particular, we still seem to have buildbot troubles.
which trouble to you mean?
[14:13:42] <jhg> do we have a roadmap?
[14:15:25] <wiesand> Sure. We’re waiting for sufficient review of the changes discussed so far for inclusion in 1.6.19. In a few cases, this means on master. We’re also waiting for changes that make the client work on Linux 4.8.
[14:15:46] <wiesand> Working buildbot verification would help...
[14:15:55] <wiesand> And then we can issue 1.6.19pre1.
[14:16:26] <kadukoafs@gmail.com/barnowlD10ED525> Someone(tm) should pester the buildslave owners
[14:16:59] <meffie> ok, i'll do that.
[14:17:11] <wiesand> Thanks.
[14:17:55] <kadukoafs@gmail.com/barnowlD10ED525> I didn't get much done last week, either, though it does seem like my
two weeks of viral-induced sadness are almost over.
[14:18:17] <wiesand> Get well soon
[14:18:25] <meffie> oh, sorry. get well soon.
[14:19:16] <jhg> yes, be well
[14:19:22] <kadukoafs@gmail.com/barnowlD10ED525> Thanks.
[14:19:39] <jhg> have we documented a roadmap?
[14:20:07] <jhg> even a loose outline for the general goals and direction of release team
[14:21:05] <wiesand> I thought I’d posted that above....
[14:21:46] <jhg> I must have missed the link
[14:22:03] <wiesand> what link?
[14:22:15] <jhg> oh, the comment
[14:22:19] <Jeffrey Altman> I believe that jhg is looking for long term horizon goals such as http://www.openafs.org/roadmap.html
[14:22:42] jhg nods
[14:22:49] <jhg> yes, thanks JeffreyAltman
[14:23:15] <wiesand> I think we’re nowhere near a position to make plans of this kind.
[14:23:34] <jhg> wiesand: on the contrary, we cannot afford not to
[14:23:36] <kadukoafs@gmail.com/barnowlD10ED525> "1.8 release around Christmas"
[14:23:49] <wiesand> Ben: right ;-)
[14:24:02] <jhg> with limited resources, it becomes even more critical to focus on what is important
[14:24:08] <jhg> i.e. planning
[14:25:23] <wiesand> 1) keep the 1.6 linux client alive, 2) fix the most grave 1.6 bugs as they are found, 3) work on getting 1.8 released
[14:25:30] <kadukoafs@gmail.com/barnowlD10ED525> I am not sure if you have been attending long enough to get a copy of
my list of 1.8 blockers
[14:25:37] <wiesand> And that’s already beyond our resources.
[14:26:31] <meffie> we should repost the list of 1.8 blockers to -devel when you are feeling better ben.
[14:27:37] <jhg> this content should probably get moved to the wiki (or something easy to keep alive and up to date)
[14:27:40] <jhg> http://www.openafs.org/roadmap.html
[14:28:22] jhg wishes he could write down wiesand's comments to do this
[14:28:30] <jhg> somewhere publicly accessible
[14:28:53] <wiesand> there are public transcripts of everything we say here
[14:29:00] <kadukoafs@gmail.com/barnowlD10ED525> (https://conference.openafs.org/release-team@conference.openafs.org/2016-08-24.html)
[14:29:39] <wiesand> right, thanks Ben
[14:30:09] <meffie> changes to www.openafs.org <http://www.openafs.org> can be submitted via gerrit, if i recall correctly.
[14:30:20] <wiesand> correct
[14:31:00] <mvita> so, about linux 4.8
[14:31:22] <jhg> okay, I will refresh it with wiesand's comments. and when I get ben's 1.8 blockers they can get linked there as well.
[14:31:39] <mvita> joe, would you have time to build master on 4.8 and do the time patches?
[14:32:37] <jhg> mvita: it builds and runs robotest just fine right now
[14:32:49] <jhg> if time patches are on master, then it is already done
[14:33:07] <mvita> no, there are no time patches for 4.8 on master
[14:33:19] <kadukoafs@gmail.com/barnowlD10ED525> It may also be worth putting some effort into making robotest exercise
more functionality
[14:33:26] <mvita> so perhaps the planned changes haven't been merged to linux yet
[14:33:49] <mvita> what distro did you use for 4.8?
[14:33:53] <jhg> ubuntu
[14:34:40] <jhg> new kernel debs are up. testing now.
[14:36:21] <wiesand> there are 4.8rc kernel debs for ubuntu?
[14:36:40] <jhg> wiesand: nightly, yes
[14:37:29] <meffie> what about pre rc changes?
[14:37:37] <jhg> meffie: limited resources
[14:38:19] <meffie> well, less resources are needed to put out a match than a forest fire.
[14:38:52] <jhg> true, however it is easier to test nightlies than track pre-release linux kernel development trees
[14:39:09] <jhg> and I'd rather invest time in getting AFS in-tree than playing that game
[14:39:45] <meffie> it will be too late then.
[14:39:58] <jhg> what? to get it in-tree?
[14:41:31] <meffie> to avoid breaking changes getting in tree
[14:42:24] <jhg> when we get some serious build resources, I'll put the kernel build bot scripts on them and we can pump out builds & tests for whatever git repo/hash we throw at it
[14:42:51] <jhg> until then, the ubuntu nightly debs run on my home box =)
[14:43:40] <meffie> yep. cool.
[14:48:45] <mvita> <researching 4.8rc2 time stuff>
[14:48:47] <meffie> any more news on the linux front?
[14:50:56] jhg shakes head
[14:51:46] <wiesand> On that "roadmap" topic:
[14:52:07] <wiesand> Maybe it’s not as obvious as I see it...
[14:52:14] <wiesand> visit https://gerrit.openafs.org/#/q/status:open+project:openafs+branch:openafs-stable-1_6_x
[14:52:31] <wiesand> forget everything marked -2 or not touched for more than two months
[14:52:49] <wiesand> 9 changes left
[14:53:08] <wiesand> review on most of them is a 1.6.19pre1 blocker
[14:53:19] jhg nods
[14:53:32] <wiesand> then look at the same page for the master branch
[14:54:12] <wiesand> besides your own ones, those prefixed ubik: were discussed in recent meetings as candidates for inclusion in 1.6.19
[14:54:16] <kadukoafs@gmail.com/barnowlD10ED525> (I had gone through all gerrit changes for master and -2'd "not for
1.8" most of them, several months ago.)
[14:54:27] <wiesand> Again, typically we’re blocked on code review.
[14:55:11] <wiesand> And that really is the "road map" regarding 1.6.x as far as it makes sense to define (well, plus linux 4.8 changes of course)
[14:55:38] <kadukoafs@gmail.com/barnowlD10ED525> Yeah, the 1.6 roadmap is ideally "1.8 exists and people move to that"
[14:56:23] <wiesand> And for 1.8 it’s similar. As Ben says. Forget what’s marked -2, help making Ben confident to merge the rest.
[14:56:40] <wiesand> Ben: yes.
[14:57:19] <wiesand> But at the pace we’re moving... maybe next Christmas then...
[14:57:50] <jhg> ok, sounds reasonable. perhaps we can flesh out some of the robotest suites and run them against these.
[14:58:06] <kadukoafs@gmail.com/barnowlD10ED525> Well, for master we can further prioritize than just that
[14:58:29] <kadukoafs@gmail.com/barnowlD10ED525> There are a bunch of things that I didn't list as blockers for 1.8
that are not marked -2, since they could go in if they got review and
such
[14:59:22] jhg nods
[15:00:13] <kadukoafs@gmail.com/barnowlD10ED525> Did Joe get on the release-team mailing list? (I was kind of
wondering whether most of what we've been sending there could just go
to -devel)
[15:00:50] <wiesand> It probably could
[15:01:26] <jhg> I registered for it. pending admin approval. I don't see any reason to separate it from -devel at this point.
[15:01:26] <meffie> i think -devel is better for devel
[15:02:29] <meffie> the -devel archives are searchable
[15:03:03] <meffie> https://lists.openafs.org/pipermail/openafs-devel/  but the release-team archives require a password
[15:03:10] <wiesand> Maybe it would make sense to (ab)use the "topic" for the blocker lists(s)?
[15:03:33] <jhg> wiesand: which one? for the IRC room?
[15:03:41] <wiesand> no, gerrit
[15:03:56] <wiesand> Ben can edit the topic of any change
[15:04:29] <meffie> oh, that is an interesting idea.
[15:04:30] <jhg> ah
[15:04:32] <wiesand> setting it to "1.8-blocker" for the changes in question would make the list available with a single click
[15:04:45] <mvita> nice
[15:05:07] <jhg> hmm... I suppose that is one way to tag it. I wish we could just add a tag.
[15:05:34] <wiesand> other useful ones would be "1.6-wanted" and "1.6-blocker" (yes on master)
[15:06:57] <meffie> i'd hate to lose are topics tho. e.g. dvhints
[15:07:19] <wiesand> Just an idea...
[15:07:48] <jhg> what is the status of the other branches? 1.4 deprecated. 1.6 ... support until?, 1.7 ... 1.8 ...
[15:08:22] <wiesand> Mike: yes, pity, but then those lists are rather short
[15:08:34] <wiesand> 1.4: dead
[15:08:54] <jhg> meffie: I haven't seen any -devel traffic since I signed up for it...
[15:08:59] <wiesand> 1.6: support until 1.8 is out and considered better/more stable than 1.6
[15:09:03] <wiesand> 1.7: dead
[15:09:33] <wiesand> 1.8: To be cut. By Christmas. For a couple of years now.
[15:10:07] <meffie> jhg, me neither.
[15:10:15] <jhg> Windows IFS implementation ?
[15:10:27] <meffie> ?
[15:10:29] <jhg> meffie: so why not merge with -release?
[15:10:38] <jhg> The 1.8 series will become the first stable release of OpenAFS to include the
[15:10:38] <jhg> Windows IFS implementation.&nbsp; No other new features will be added to 1.8.
[15:11:08] <meffie> i'm just saying the release-team archives require a password to see.
[15:11:09] <wiesand> just forget the roadmap you read
[15:11:20] <jhg> and btw, whose google analytics is on the webpage? UA-5995928-1
[15:11:28] <jhg> wiesand: I am updating it with this conversation
[15:12:15] <wiesand> Keep the "rxgk" topic, this may land on master after 1.8.x was cut. Throw everything else away.
[15:12:27] <meffie> > Just an idea...
i think the gerrit topics is a good idea. i dont have a better one :)
[15:14:46] <wiesand> Of course I could be wrong. The long term roadmap is not my turf.
[15:18:23] <meffie> that was my understanding as well.
[15:23:11] <meffie> are we dismissed wiesand? :)
[15:24:13] <wiesand> Let’s call it a meeting, yes.
[15:24:26] <jhg> how do I change the website? I have a commit ready to go, but perhaps no access to do so
[15:24:56] <wiesand> submit to gerrit, project openafs-web
[15:25:01] <jhg> k
[15:25:29] <wiesand> git push ssh://gerrit.openafs.org/openafs-web.git HEAD:refs/publish/master
[15:26:05] <wiesand> Ben and I have the powers to merge changes there and bring them live.
[15:26:12] <meffie> cool!
[15:26:30] <wiesand> I’m not going to merge a roadmap change though, unless it’s approved by the gatekeepers/gurardians.
[15:26:37] <meffie> i know jhutz is a busy guy.
[15:27:11] <wiesand> As I said, not my turf.
[15:27:50] jhg wonders what his gerrit password is
[15:28:24] <jhg> I thought I was using ssh keys
[15:28:33] <meffie> you mean your ssh key password?
[15:28:48] <jhg> meffie: not sure...
[15:28:53] <jhg> it's not the same
[15:30:46] <jhg> git push ssh://gerrit.openafs.org/openafs-web.git HEAD:refs/publish/master
[15:30:50] <meffie> unless it's asking for a password for the key, you probably need to fix your .ssh/config to set the identyfile used
[15:30:53] <jhg> ssh: connect to host gerrit.openafs.org port 22: Operation timed out
[15:30:53] <jhg> fatal: Could not read from remote repository.
[15:30:53] <jhg> Please make sure you have the correct access rights
[15:30:53] <jhg> and the repository exists.
[15:30:54] <jhg> Exit 128
[15:31:18] <meffie> that is not the correct port number
[15:31:20] <jhg> http instead of ssh asks for the bizarre login
[15:32:20] <meffie> Host gerrit.openafs.org User <Username> IdentityFile ~/.ssh/id_rsa Port 29418
[15:32:50] <wiesand> you have to set it up just the same as the main project
[15:35:25] <kadukoafs@gmail.com/barnowlD10ED525> Thinking of rxgk and 2.0, has there been any movement on IPv6?
[15:37:30] <jhg> https://gerrit.openafs.org/#/c/12372/
[15:38:37] <meffie> kadukoafs, not recently, the work i did to convert int32 to sockaddr is still setting in my github.
[15:38:59] <kadukoafs@gmail.com/barnowlD10ED525> Not a huge fan of using search URLs as a roadmap (vs. some static text
describing the current contents), but thanks for putting something
together.
[15:39:35] <jhg> it's not likely to change and I felt it was more expressive than "click here" for xxxxx
[15:40:18] <jhg> however, I am not tied to it. the amount of errors removed by simply having that as opposed to what was there before is a value in and of itself
[15:42:07] <wiesand> I guess we’re finished for today?
[15:42:14] <kadukoafs@gmail.com/barnowlD10ED525> Something like that.
[15:42:34] <jhg> sure
[15:42:44] <wiesand> Thanks everyone!
[15:42:48] wiesand leaves the room
[15:42:52] <mvita> bye
[15:44:25] <jhg> ciao
[16:52:43] meffie leaves the room
[17:38:22] kadukoafs@gmail.com/barnowlD10ED525 leaves the room
[17:45:32] kadukoafs@gmail.com/barnowlD10ED525 joins the room
[20:35:26] Jeffrey Altman leaves the room
[20:35:26] Jeffrey Altman joins the room
[22:20:59] Jeffrey Altman leaves the room
[22:20:59] Jeffrey Altman joins the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!