[00:46:35] --- Russ has left: Disconnected [01:37:35] --- dev-zero@jabber.org has become available [01:37:38] --- dev-zero@jabber.org has left: offline [01:39:45] --- Simon Wilkinson has become available [01:40:04] --- Simon Wilkinson has left [05:17:16] --- Simon Wilkinson has become available [05:25:48] --- SecureEndpoints has left: Replaced by new connection [05:26:03] --- SecureEndpoints has become available [06:31:20] "OpenAFS: We don't make new releases because we're bored" [06:46:52] huh? Because we are bored, we don't make new releases? [06:47:20] but when we're not bored, we do lots of releases? [06:47:48] No, as in, we're not making new releases just because we haven't got anything better to do. If you're running 1.4.x, for all values of x<10, and you see problems, there might be a reason for that. [06:52:36] sorry. bad humor. [06:52:59] --- mmeffie has become available [06:53:50] On a more serious note, anyone remember why we don't just turn -fPIC on globally on 64 bit platform builds? [06:54:45] I think doing so might be forbidden by Debian policy [06:56:02] Personally, I couldn't care about Debian... [06:56:36] We can always do it for some, and not for others. But it does mean that all of our .a libraries are unusable to build relocatable executables on 64 bit platforms. [06:59:26] I see we discussed this last year on openafs-info, without any real conclusion... [07:03:20] hmm [07:03:35] would it be possible to provide BOTH libraries? One with -fPIC and one without? [07:07:21] The issue Russ pointed out is that it isn't easy to provide one set of options for libraries, and another for code. So we can't build our libraries -fPIC, and others statically, especially for the case (perl-AFS) that most care about. [07:08:48] looks like src/WINNT/afsd/cm_dcache.c cm_BufWrite() has a performance problem. It constructs the BIOD list before the store synchronization logic. So we only permit one thread at a time to store data but mulitple threads can store the same data over and over again. [07:18:05] projects.html, roadmap.html, credits.html updated [07:20:32] SecureEndpoints: Disconnected does support conflict resolution strategies, just only one strategy at a time. [07:21:10] (ie the user won't ever get prompted, but can pick from 'server wins' or 'client wins' at reconnection time) [07:28:00] feel free to send me more detailed text about status and what needs to be done. I will post it [07:29:02] Will do. [07:31:30] I need a volunteer to put together a historical timeline from the beginning of OpenAFS to present detailing major events [07:32:03] the purpose is to point folks at it that claim that OpenAFS is dead [07:34:51] Of course, I could post the cocomo analysis http://www.secure-endpoints.com/openafs/cocomo.htm [07:36:15] The thing about cocomo is that it only shows changes which result in new code. I think something which shows about of _changed_ code since IBM_1_0 would be very interesting. [07:42:46] simon - agreed. [07:44:18] For instance, there are some files where cvs annotate finds virtually nothing but blank lines that date from version 1.0. Some of that is obviously due to the grand reformatting, but I think we've changed quite a lot over the years! [07:44:20] actually, it only reports changes in numbers of lines of code. it would be more expensive but we could perform an analysis of the commits to see number of lines of code changed, added, removed (ignoring whitespace). however, that would not be equivalent to the cocomo counts because that would still be counting physical lines instead of logical lines based upon the code expressions. [07:45:08] Somewhere or other we have a directory with all of the deltas in it since day 1. diffstat and a bit of perl could tell us how many physical lines we have changed ... [07:52:00] --- haba has become available [07:53:40] --- mmeffie has left [07:58:46] --- reuteras has left [08:00:18] i have that directory, but it's only deltas through, like, october [08:00:47] Was that directory generated by a script? [08:00:55] sort of [08:01:07] Script, then manually munged? [08:01:16] it was generated by my modified cvsps and then manually "helped" [08:01:40] So, an incremental run wouldn't be out of the question, then? [08:01:53] > credits.html updated [08:01:59] we have a script for this somewhere [08:02:29] it wouldn't, but i don't promise it's simple. last time i tried it misnumbered the new deltas and i had to hand-renumber [08:03:47] Can I get the pieces from you? [08:05:40] they should all be in an RT ticket. hang on [08:12:20] 119102 [08:13:29] ta [08:14:05] if you need the existing directory, i can dredge that out too [08:14:39] If you've got the existing directory, and its got hand edits in it, that would be useful too. [08:14:55] Apparently there's some new fangled global filesystem you could put it in? [08:15:52] there is. and i did, but it's large and i didn't want to leave it around wherever i put it [08:21:20] I can give you a directory here to dump it in, if you like. [08:37:08] nah, i can stick it here. [08:37:16] sorry for disappearing, a bottle exploded in my kitchen [08:37:26] No problem. Exploded badly? [08:38:24] a jug of apple cider from the farm share was in fridge, fermented while i was gone, split side out of bottle; half gallon (of the gallon) ended up all over the floor, spilled through the door [08:38:48] Ouch. [08:38:59] I can remember having similar fun with homebrew. [08:39:16] i'm about to try homebrew; don't tell me that [08:40:21] I had too much of something == too much gas == shattered bottles [08:45:49] /afs/andrew.cmu.edu/usr/shadow/ps.tar.gz will be 52279633 in about 4 minutes [08:47:29] Cool. I'll grab it when I get in ... [08:47:35] Thanks! [09:01:41] --- haba has left [09:03:03] --- SecureEndpoints has left [09:22:20] --- SecureEndpoints has become available [09:40:51] --- bpoliakoff has become available [09:45:59] > I think doing so might be forbidden by Debian policy Last I checked, Debian policy applied to Debian, not OpenAFS [09:47:08] > all of our .a libraries are unusable to build relocatable executables Yes, that normally is a feature of archive libraries. People who have been spoiled because it happens to work on i386 should just get over it. In other news, we should ship a more meaningful collection of shared libraries. [09:48:35] Yes. A more meaningful collection of shared libraries would be very good. Because people who are gradually moving to x86_64 are going to be sad when none of their home grown tools work any more ... [09:49:09] --- mmeffie has become available [09:51:11] People who claim OpenAFS is dead usually were born yesterday and assume we're dead because we don't have/use whatever fad their favorite project is into. We do active development, we do releases, we have tons of mailing list and jabber and IRC (and zephyr) traffic => we're not dead, go away [09:53:11] > Somewhere or other we have a directory with all of the deltas in it /afs/.grand.central.org/project/openafs/deltas Note use of RW path. But it looks like the cron job has been broken since January :-( [09:53:38] jhutz - it's that last part of what you said that makes it hard to recruit new people: "...go away" [09:58:17] If you're interested in becoming involved, we want you. If you're interested in going around telling people we're dead, not. [09:58:21] And I don't think that the people who say OpenAFS is dead were "born yesterday". They may be lacking in technical clues, but unfortunately they tend to be the ones who make the policies. [09:59:51] I mean wrt open-source software. If they assume that not having a particular tool, or having a tool but not using it much, means there's no activity, then they haven't seen many projects [10:01:14] --- Simon Wilkinson has left [10:45:18] --- Simon Wilkinson has become available [10:45:57] Is there somewhere I can grab an up to date copy of the raw OpenAFS CVS from? [10:47:24] --- Russ has become available [10:48:35] /afs/.grand.central.org/project/openafs/cvs out-of-date like the deltas, but that'll be fixed shortly [11:28:44] --- bpoliakoff has left [11:32:53] --- bpoliakoff has become available [12:02:32] --- mmeffie has left [12:16:45] --- dev-zero@jabber.org has become available [12:17:42] --- dev-zero@jabber.org has left: offline [15:23:33] --- dev-zero@jabber.org has become available [15:23:37] --- dev-zero@jabber.org has left: offline [15:41:46] --- edgester has become available [15:58:19] --- dev-zero@jabber.org has become available [15:58:23] --- dev-zero@jabber.org has left: offline [17:44:06] --- RedBear has left [17:44:52] --- dwbotsch has become available [18:16:52] --- Russ has left: Disconnected [18:17:02] --- Russ has become available [18:17:07] --- tkeiser@sinenomine.net/owl has left [18:39:35] --- shadow@gmail.com/owl3F374034 has left [18:44:37] --- edgester has left [18:57:41] --- shadow@gmail.com/owlD3CF3980 has become available [21:19:22] --- RedBear has become available [22:39:54] --- reuteras has become available [23:11:19] --- dev-zero@jabber.org has become available [23:11:24] --- dev-zero@jabber.org has left: offline