[00:00:05] --- Simon Wilkinson has become available [00:27:23] --- Simon Wilkinson has left [00:27:46] --- Daria has left [00:29:58] --- mvita has left [00:56:18] --- Marc Dionne has become available [00:57:20] --- kula has left: Lost connection [00:59:59] --- Daria has become available [01:01:59] --- Simon Wilkinson has become available [01:03:14] --- mvita has become available [01:04:17] --- meffie has become available [01:20:57] --- kula has become available [01:21:39] --- kula has left: Lost connection [01:23:02] --- haba has become available [01:29:45] Quick and dirty Heimdal package (installs in /usr/heimdal/) for Centos: /afs/pdc.kth.se/public/ftp/pub/bambi/yum/centos/6/local/x86_64/RPMS/heimdal-1.5.2-5.x86_64.rpm [01:30:15] --- stephan.wiesand has become available [01:30:36] That's not really what Jeffrey and Ken were talking about, though. The plan is to get a Heimdal package into Fedora itself, so it is available as part of the OS. [01:30:57] Simon: yes of course. [01:31:17] Quick and dirty for someone who wants it kindof as a gap fill [01:34:30] Of course Ken is doing the right thing (TM) ;) [01:35:36] If you want the packages that will eventually be in Fedora you can ask Ken [01:37:00] Or build them yourself using the specfile that's in bugzilla [01:41:15] https://gitorious.org/ktdreyer/heimdal-packaging/source/97aecaed997b27c6ec71f776579e038965cb07be: [01:49:04] --- Daria has left [01:49:54] --- Derrick Brashear has become available [01:52:51] --- meffie has left [01:52:52] --- meffie has become available [02:45:31] should anyone feel like having a look at my first stab at the 1.6.8 release notes: gerrit 10958 [02:46:55] if you end up repushing, might as well rebase away the trailing whitespace [03:05:12] --- Marc Dionne has left [03:19:45] --- mvita has left [03:30:19] --- Marc Dionne has become available [03:33:57] --- mvita has become available [04:25:45] --- kula has become available [04:26:16] --- kula has left: Lost connection [04:27:15] "So, how about those C strict aliasing rules?" [04:30:10] --- stephan.wiesand has left [04:50:38] --- Derrick Brashear has left: Disconnected [04:53:24] --- meffie has left [04:53:29] --- mvita has left [04:53:31] --- mvita has become available [04:53:46] --- Marc Dionne has left [04:56:07] --- mvita has left [04:56:08] --- mvita has become available [04:57:59] --- haba has left [04:58:04] --- Simon Wilkinson has left [05:04:51] --- mvita has left [05:28:34] --- Derrick Brashear has become available [06:13:36] --- mvita has become available [06:18:23] --- meffie has become available [06:24:24] --- meffie has left [06:31:31] --- mvita has left [06:31:32] --- mvita has become available [06:39:50] --- mvita has left [06:41:19] --- haba has become available [06:50:12] --- cudave has left: Disconnected [06:54:54] --- Derrick Brashear has left: Disconnected [07:01:50] --- haba has left [08:10:27] --- ballbery has left [08:13:18] --- ballbery has become available [08:47:50] --- Simon Wilkinson has become available [08:51:40] --- Marc Dionne has become available [08:53:56] --- ktdreyer has become available [08:56:27] btw, USGS's Heimdal 1.6 packages are at http://butst.usgs.gov/pub/usgs/free/el/testing/6/x86_64/repoview/ . This is the version that will eventually go into Fedora [08:57:22] it's the "1.6" branch in that Git repo Jeff linked earlier [09:20:22] --- Derrick Brashear has become available [09:32:43] There is an RT ticket for the 'getcwd' issue. [09:33:32] I guess I can mention the points that I noted during Simon/Andrew's talk, here. [09:34:21] VolForward would need to change for IPv6, but also there would need to be a rethinking for how it would work with rxgk departmental fileservers. [09:34:26] --- mvita has become available [09:35:14] I thought we had "always" agreed that ubik is an implementation detail, and changes to ubik behavior explicitly do not need to go to AFS3-stds. [09:35:17] You can, subject to policy, forward from a shared-key fileserver to a departmental one. Forwarding in the opposite direction requires client operation [09:35:52] ubik: I think that is the general assumption, but I am not sure whether it has ever been explicitly stated. It probably should be. [09:35:52] Right. I think jhutz had some ideas for ways to go from, e.g., one departmental to a different one. [09:36:07] --- meffie has become available [09:36:59] I guess I sort of talked to Andrew in person, but it looks like he has some sense of the layout of the vldb format on disk. It would be nice to have some written notes that we could refer to. (I could probably flesh it out into a more detailed description if Andrew's time is limited.) [09:37:48] --- Simon Wilkinson has left [09:37:57] --- Simon Wilkinson has become available [09:38:36] so one constraint in all of this is backward compatibility especially vis-a-vis IBM [09:38:39] I'd be happy to review the document. The format is a little more complicated than it was shown in those slides - and there are things like hash tables that you need to think about [09:38:53] which means some things are kinda stone-graven [09:39:00] Not much [09:39:07] The last thing I wanted to note is that making the tie-breaking "lowest IP" always be an IPv4 address (if one is present) is the only thing that makes sense if there is going to be a dbserver that doesn't know about IPv6 (to get the transition right). [09:39:19] We have to continue to support IBM AFS clients. That's pretty much it, as I understand it. [09:39:34] kaduk: Yeah, that was the conclusion I came to. [09:39:45] *nods* [09:43:38] Should someone (meffie?) send some notes on our stack usage investigations to -devel? [09:43:56] That's probably a useful place for them to end up [09:44:37] I suppose that the ticket would not be wrong, either. [09:45:08] Depends if you want to provoke wider discussion, really. Not many folk read RT [09:46:49] I think we determined that yesterday, yes :) [10:01:21] --- Simon Wilkinson has left [10:03:32] --- Simon Wilkinson has become available [10:03:38] --- meffie has left [10:04:59] --- haba has become available [10:27:33] --- Simon Wilkinson has left [10:29:22] --- Simon Wilkinson has become available [10:34:06] --- stephan.wiesand has become available [10:53:58] --- Marc Dionne has left [10:58:59] --- Derrick Brashear has left: Disconnected [11:03:25] --- stephan.wiesand has left [11:06:31] --- Simon Wilkinson has left [11:07:54] --- ktdreyer has left [11:17:00] --- haba has left [11:17:00] --- haba has become available [11:24:02] --- haba has left [15:34:42] --- kula has become available [16:13:29] --- haba has become available [16:13:46] --- Marc Dionne has become available [16:30:27] --- Marc Dionne has left [23:12:23] --- Marc Dionne has become available [23:50:20] --- kula has left: Lost connection