[00:09:44] --- dev-zero@jabber.org has left [00:10:05] --- reuteras has become available [00:42:27] --- dev-zero@jabber.org has become available [01:07:20] --- haba has become available [01:10:08] --- Simon Wilkinson has become available [01:48:08] --- Simon Wilkinson has left [03:30:49] --- Claudio Bisegni has become available [03:38:20] --- Claudio Bisegni has left [04:10:59] --- gendalia@gmail.com/owlD3D5926A has left [04:16:09] --- Simon Wilkinson has become available [04:16:43] --- Simon Wilkinson has left [04:41:09] --- dlc has become available [05:34:47] --- Jeffrey Altman has left: Replaced by new connection [05:49:09] --- matt has become available [06:52:21] --- deason has become available [08:04:35] --- reuteras has left [09:29:49] --- cclausen has become available [09:57:54] --- haba has left [10:18:13] --- Russ has become available [10:59:57] --- Jeffrey Altman has become available [11:08:43] --- Simon Wilkinson has become available [11:10:55] --- dwbotsch has left [11:11:21] --- Rrrrred has become available [11:33:41] --- dev-zero@jabber.org has left [11:48:22] --- haba has become available [12:17:56] --- dev-zero@jabber.org has become available [12:23:40] could that dynroot thing on -info possibly be dd7fd7140ef5f5852c28788eed8e821997f52be4 ? [12:24:09] sec [12:24:21] it's conceivable. i'd like to see what his fstrace says [12:32:49] I suspect the problem is that .htaccess looks like a short name read/write cell that is to be dynanically added based upon AFSDB records. [12:33:05] I would like to know if he is seeing DNS traffic or not [12:34:17] For the Windows client one solution to his problem would be to create an .htaccess entry in the Freelance database which is a symlink referring to a real file somewhere in AFS [12:35:12] i asked all the questions i think we need the answers to. has he reappeared? [12:35:18] nope [12:35:50] Why would an ordinary dotfile appear to be a cell mount? [12:36:16] /afs/.andrew.cmu.edu is the convention for readwrite mounts [12:36:18] The issue is what happens if you try to do stat /afs/.htaccess [12:36:36] Yuk. [12:36:38] --- Claudio Bisegni has become available [12:36:46] --- Claudio Bisegni has left [12:36:47] so, basically, in a dynroot, we have an expectation that it would be a short name for an rw mount [12:36:52] We probably have the same problem with /afs/.DS_Store [12:37:10] (and whatever other nonsense we deal with elsewhere for stupid GUIs) [12:37:12] fakestat might handle that before dynroot sees it [12:37:47] yeah, I had a problem with the Visual Studio installer stating random files in \\AFS\all [12:38:04] although I had that mapped to a drive letter at the time [12:38:11] it got fixed though. not sure how [12:38:22] I have quite a few lighting applications that won't run on my Mac until I turn off AFS, because they take a wander into /afs and get lost. [12:38:32] yeah [12:38:52] I've seen many Mac OS installers start searching through all files on a system for files to update [12:38:58] its bad when they get lost in AFS [12:40:46] On another tack - I've just had a problem reported to me with 1.4.10, which looks like directory callbacks getting lost when a file is being repeatedly deleted and recreated. [12:41:11] i'd bet not. i'd bet a race [12:41:28] Well, a race is my suspicion too. [12:41:55] But it's definitely callback related, because creating a file in the directory from another machine causes the first client to magically see the file again. [12:42:17] I apparently have fstrace output waiting for me. [12:42:46] i think it's callback related in terms of handling and a race, rather than the callback being lost [12:43:20] like out of order handling [12:43:34] That seems reasonable. I will have to dig when I get a moment. [13:23:50] --- shadow@gmail.com/owl1A054DC7 has left [13:24:13] --- shadow@gmail.com/owl642EA9BE has become available [15:14:17] --- deason has left [15:34:28] --- deason has become available [15:46:51] --- Simon Wilkinson has left [15:57:08] --- shadow@gmail.com/owl642EA9BE has left [16:22:34] --- shadow@gmail.com/owl5B75C82F has become available [17:03:00] --- matt has left [18:49:02] --- cclausen has left [20:00:53] --- dev-zero@jabber.org has left [21:34:38] --- kaduk@mit.edu/owl has left: Lost connection [21:34:38] --- asedeno has left: Lost connection [21:34:38] --- andersk@mit.edu/vinegar-pot has left: Lost connection [21:58:58] --- deason has left [22:58:56] --- reuteras has become available