- Thursday, April 21, 2022

release-team@conference.openafs.org

Thursday, April 21, 2022< ^ >

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

Room Configuration

Room Occupants

GMT+0
[13:36:16] Cheyenne leaves the room
[15:55:25] meffie joins the room
[16:00:35] <kaduk@jabber.openafs.org/barnowl> greetings
[16:00:54] wiesand joins the room
[16:01:04] <meffie> greetings
[16:01:53] <wiesand> Hello
[16:04:10] <wiesand> No news from my side again, sorry. I've been "under the weather" lately.
[16:04:33] <meffie> oh, sorry to hear. i hope you feel better soon.
[16:05:23] <kaduk@jabber.openafs.org/barnowl> yes, I hope you feel better soon as well
[16:07:17] <kaduk@jabber.openafs.org/barnowl> I spent time yesterday looking into the java crypto CVE rather than reviewing
change in gerrit, though I guess I did get a couple new reviews
posted.
[16:07:49] <wiesand> Just looking at Cheyenne's Whiteboard in the Wiki. Good start. I hope I'll find time to add to it soon.
[16:07:54] <meffie> which jkd version is running on gerrit?
[16:08:05] <meffie> jdk
[16:08:11] <kaduk@jabber.openafs.org/barnowl> We're currently on java 1.8 :(
[16:08:38] <meffie> well, that means we are not vuln for that particular bug?
[16:08:40] <kaduk@jabber.openafs.org/barnowl> And I think the version of gerrit is old enough that its libraries
don't even support ECC ssh keys
[16:08:58] <kaduk@jabber.openafs.org/barnowl> For that particular bug, yes ... I expect there are plenty of others
waiting, though
[16:10:54] <kaduk@jabber.openafs.org/barnowl> I confess I did not even pull up the whiteboard yet :(
[16:11:32] <meffie> i think the whiteboard is a great idea. looks good so far
[16:11:50] <meffie> https://wiki.openafs.org/devel/Whiteboard/
[16:12:00] <kaduk@jabber.openafs.org/barnowl> Yup, looking now :)
[16:12:17] <kaduk@jabber.openafs.org/barnowl> It does look like it will be a good tool to help coordinate work
[16:12:45] <kaduk@jabber.openafs.org/barnowl> Are there any sections that seem useful for coordinating work for
1.9.x/master at the moment?
[16:12:57] <kaduk@jabber.openafs.org/barnowl> (i.e., that we would add in order to do such coordinating)
[16:13:24] <wiesand> It's a lot shorter than what I'm carrying around though, and it lacks the path conflicts between changes. Those will need to be added to make it truly useful.
[16:14:49] <meffie> i think this initial draft was focused on changes for linux :)
[16:15:04] <wiesand> So far it seems strictly 1.8.x only, though it does mention a few changes being worked on on master.
[16:16:03] <meffie> fwiw, i prefer to use git to make changes to wiki.openafs.org  https://wiki.openafs.org/devel/GitWiki/
[16:16:05] <kaduk@jabber.openafs.org/barnowl> There is nothing wrong with being focused on linux or 1.8 (or both);
I'm just brainstorming about whether there are other potential use
cases to consider.  No need to actually do anything right now
[16:19:47] <meffie> in other news, GSoC submissions deadline passed. we had small number of submissions, but i thing we have one or more strong ones.
[16:19:57] <meffie> s/thing/think/ ;)
[16:20:11] <wiesand> For reference, my current version of the text file I'm working with is https://www.zeuthen.desy.de/~wiesand/1.8-next.txt
[16:20:37] <meffie> oh!
[16:20:38] <wiesand> (and I'm going to remove it in a few minutes again)
[16:20:57] <meffie> are you going to move it to wiki.openafs.org?
[16:22:01] <meffie> to the new Whiteboard page that is?
[16:23:19] <wiesand> well, as is it's not really usable for collaborative editing IMO, but yes, getting that info into the wiki and living there was my idea behind the "whiteboard".
[16:24:19] <meffie> yes, i see. it is quite detailed.
[16:25:38] <meffie> it seems the file tree could be generated.
[16:28:57] <meffie> i'll talk to cheyenne and see if he can write a tool in "icon"
[16:29:25] <wiesand> that would help a lot, but I haven't gotten around to trying that yet
[16:30:12] <meffie> it looks like a cross reference of the source files and which commits make changes to that file (regards if the changes conflict) ?
[16:30:20] <meffie> (regardless)
[16:30:25] <kaduk@jabber.openafs.org/barnowl> yes
[16:30:38] <wiesand> I'm afraid I don't have anything else for today
[16:30:38] <kaduk@jabber.openafs.org/barnowl> that's (my understanding of) what gerrit uses as the "path conflict"
detection mechanism
[16:31:21] <meffie> i think a interesting list would be the order of the commits as the exist on the "upstream" master branch.
[16:32:11] <meffie> i'm still not sure why we care about path conflicts. seems merge conflicts are the real problem.
[16:32:29] <kaduk@jabber.openafs.org/barnowl> I don't have so much, either; I seem to have confused myself with the
apple-silicon stack because we reordered patches, so my previous -1
was incorrect.  It looks like I'm blocked on merging any of it by
confirming that the commit in the middle of the stack that uses the
saved "sysnameinfo->name" information from the first commit does so in
a useful manner, but that was one of the commits where I confused
myself by the reordering.  So I guess that means I need to look again
when I'm actually awake.
[16:32:43] <meffie> maybe a chart to show merge conflicts would be helpful.
[16:33:00] <wiesand> right - if two changes touch the same file, and you merge them in the wrong order, you get to rebase one of them (which may be trivial or not)
[16:33:14] <kaduk@jabber.openafs.org/barnowl> path conflicts can be an indicator that two changes are not
commutative, even if there are not any merge conflicts detected by
git.  It really needs a human reviewer to look and find out, I think.
[16:33:27] <kaduk@jabber.openafs.org/barnowl> Also what Stephan said :)
[16:34:26] <wiesand> and things get the more "interesting" the more the backlog grows
[16:34:27] <meffie> oh, i see. path conflict means it "might" be a merge conflict.
[16:34:28] <kaduk@jabber.openafs.org/barnowl> Oh, and Mike, while I have you here: for 14950, do you think it's okay
to leave the cell number printed as a signed integer for now, until
the actual code handling those numbers gets updated?
[16:34:51] <meffie> yes, just pointing out that one too.
[16:35:26] <kaduk@jabber.openafs.org/barnowl> ok, thanks
[16:36:06] <kaduk@jabber.openafs.org/barnowl> And please convey my thanks to Marcio for working on removing LWP
usage (14940, 14941)
[16:36:14] <meffie> yes! +1
[16:36:26] <wiesand> that's also why I loathe those security updates - they invalidate all your work on disentangling things immediately (and I think that's what happened to 1.6.x)
[16:37:33] <meffie> well, no one likes security issues, except the bad guys.
[16:38:03] <meffie> but yeah, it's always a disruption.
[16:38:08] <kaduk@jabber.openafs.org/barnowl> agreed
[16:38:31] <kaduk@jabber.openafs.org/barnowl> any other topics for today?
[16:38:51] <meffie> wiesand: do you want to see the GSoC proposals?
[16:39:07] <wiesand> removing LWP usage - sigh, yes good thing, but more code skew since I guess that's not for stable again…
[16:39:24] <meffie> i can send an invite (or just the docs)
[16:40:18] <meffie> that's all i have.
[16:40:42] <wiesand> Mike: I'd love to, but frankly I'll not be able dedicate enough time to them to make that useful (and keep in mind I'm still not a developer)
[16:41:18] <wiesand> anyway, it won't do harm…
[16:41:29] <meffie> ok, i can just send just for you info in case you have time to look.
[16:42:01] <wiesand> thanks
[16:42:12] <wiesand> adjourn?
[16:42:28] <kaduk@jabber.openafs.org/barnowl> seconded
[16:42:28] <meffie> second. have a good day/evening.
[16:42:33] meffie leaves the room
[16:42:41] <kaduk@jabber.openafs.org/barnowl> thanks all, and have a good week
[16:42:53] <wiesand> bye then, cu next week
[16:43:00] wiesand leaves the room
[17:10:22] Cheyenne joins the room