[00:59:27] --- dev-zero@jabber.org has become available [01:11:44] --- dev-zero@jabber.org has left: offline [01:53:30] --- tkeiser@sinenomine.net/owl has left [02:09:02] --- dragos.tatulea has become available [02:32:54] hi [02:40:05] --- dragos.tatulea has left [02:45:06] --- dragos.tatulea has become available [02:45:11] Simon Wilkinson: Around? [02:56:07] --- haba has become available [03:00:16] --- dragos.tatulea has left [05:22:37] --- dragos.tatulea has become available [05:40:34] --- sxw has become available [05:41:03] dragos.tatulea: Am now! [05:41:10] hi [05:41:26] hi [05:41:52] So, I have a few hours available for hacking openafs :). I was looking over my dissapeared cache pinning page :) [05:42:41] question one: Should there be a separate pinning thread? Or should I use the afs daemon thread? [05:42:43] Your disappeared page? [05:43:02] I'd use a separate thread. [05:43:21] Otherwise, you could easily starve the daemon thread. [05:43:27] good [05:43:39] --- SecureEndpoints has become available [05:44:37] --- dragos.tatulea has left [05:46:44] --- dragos.tatulea has become available [05:46:55] ping [05:47:04] --- dragos.tatulea has left [05:48:17] --- dragos.tatulea has become available [05:48:26] ping [05:48:29] oh [05:48:55] Still here.... [05:49:06] sxw: So, I was saying: When a file that has CPinned & ~CStatd set, the vcache should get pushed upfront in the pin queue and then the daemon should awakened. [05:49:24] (I'm thinking vcaches for now) [05:49:56] Yes. [05:50:26] So, you probably want the callback break handler (which unsets CStatd) to handle putting the pinned vcache in the refresh queue. [05:51:59] great [06:09:40] any ideas on this: http://pastebin.com/d3cc0c1c3 ? [06:09:57] What is @RXDEBUG@? [06:10:10] Just did a cvs up [06:10:28] Did you run regen.sh, and reconfigure? [06:10:39] I suspect its a new symbol that's appeared in configure and not been caught. [06:11:07] Our dependencies aren't very good at all, so you often need to manually refresh your tree after a cvs up. [06:11:19] Forgot about regen.sh [06:11:32] (I often do.) [06:11:53] We should probably make the makefiles rebuild configure if configure.in is newer. [06:17:35] Our dependencies are fine. Makefiles should not try to rebuild files that are in the source distribution. That's what regen.sh or autoreconf is for. [06:17:58] Our Makefiles aren't fine. There are many dependency issues in them. [06:18:45] I think the jury is out on what make should do if configure isn't up to date. [06:18:53] configure isn't part of the CVS distribution. [06:18:57] NO, we absolutely should NOT do that. I have enough trouble already with packages that attempt to be "helpful" that way and require me to jump through hoops to subvert this behavior so that I can build them from a source directory that isn't writable and also doesn't have timestamps in the "expected" order. [06:19:32] There is no "CVS distribution". I said "source distribution" and meant it. If it's in the source tarball, the rules shouldn't build it. [06:19:39] Or not automatically, anyway. [06:19:53] If it's in the source tarball, then the Makefile that comes with that tarball shouldn't try and build it. [06:20:18] The right way to deal with this is to use AM_MAINTAINER_MODE, so people making changes to configure.in et al can have automatic rebuilds without the rest of us being screwed. But automake transition never happened. [06:21:52] Nor should the Makefile that results in running configure from that tarball. If you want to somehow magically have different Makefiles in source tarballs than you get by checking out a CVS sandbox, I have no problem with that, as long as it doesn't result in multiple authoritative copies of things that have to be maintained in parallel (for example, generating one set from the other would be OK) [06:33:03] --- sxw has left [08:07:50] --- reuteras has left [08:59:16] --- dragos.tatulea has left [08:59:47] have a nice day [08:59:49] --- haba has left [09:04:10] --- dragos.tatulea has become available [09:44:48] --- dev-zero@jabber.org has become available [09:51:21] --- dev-zero@jabber.org has left: offline [10:14:22] shadow@gmail.com/owl936E8D6C: Around? [10:23:30] --- sxw has become available [10:38:34] --- dragos.tatulea has left [10:40:38] --- dragos.tatulea has become available [10:47:19] --- sxw has left [10:53:55] --- Russ has become available [11:09:46] hi [11:10:04] you don't need to send a complete client name with the jabber instance specifier: it won't make a difference [11:15:39] Yeah, I don't know where this practice came from of using complete JID's to address someone when a userid will do. [11:15:52] i prefer to be addressed by name [11:16:05] Or one of those [11:16:07] the practice is probably that stupid "alert on userid" support in clients [11:16:35] If that ties to a particular resource, then its users deserve to lose. [11:16:49] ... for example, if you use a different jid every time you start owl [11:17:37] i get a different jid every time i log into to gmail [11:32:14] shadow: Hi. [11:32:33] hi [12:01:47] I'm having trouble logging into TWiki. [12:02:58] My user still exists in the list. But it can't find it when trying to login (or using forgot my password). [12:03:25] i wonder if i screwed something up when i did the server migration. lemme look [12:03:34] (www.dementia.org moved to a new machine) [12:24:31] fixed [12:28:26] let me check [12:28:43] okay [12:28:48] I managed to log in [12:29:03] great [12:29:06] thanks! [12:30:32] --- manfred furuholmen has become available [12:32:09] --- asedeno@mit.edu/darkmatter has become available [12:32:59] --- asedeno@mit.edu/darkmatter is now known as asedeno [13:50:07] --- dragos.tatulea has left [13:55:28] --- tkeiser@sinenomine.net/owl has become available [14:03:31] --- shadow@gmail.com/owl936E8D6C has left [14:39:20] --- shadow@gmail.com/owl42DA2778 has become available [15:13:38] --- manfred furuholmen has left [15:15:36] --- manfred furuholmen has become available [15:25:38] --- manfred furuholmen has left [15:27:13] --- manfred furuholmen has become available [16:00:41] --- manfred furuholmen has left [16:02:19] --- manfred furuholmen has become available [16:21:22] --- manfred furuholmen has left [18:55:29] --- matt has become available [18:55:33] --- matt has left [21:01:31] --- Rrrrred has left [21:01:56] --- dwbotsch has become available [21:18:14] --- shadow@gmail.com/owl42DA2778 has left [21:21:16] --- kula has left [21:25:56] --- kula has become available [21:33:04] --- shadow@gmail.com/owl3F374034 has become available [21:34:39] --- dwbotsch has left [21:35:43] --- dwbotsch has become available [22:14:11] --- RedBear has left [22:31:49] --- brantgurga has left [22:32:10] --- brantgurga has become available [22:40:40] --- sxw has become available [23:04:31] --- reuteras has become available [23:19:34] --- dlc has left [23:26:14] --- sxw has left [23:45:15] --- dev-zero@jabber.org has become available [23:46:16] --- dev-zero@jabber.org has left: offline