[00:53:50] --- dev-zero@jabber.org has become available [01:46:40] --- dev-zero@jabber.org has left [02:05:10] --- dev-zero@jabber.org has become available [03:42:18] --- SecureEndpoints has left: Replaced by new connection [04:52:26] --- SecureEndpoints has become available [06:44:22] --- SecureEndpoints has left [06:58:36] --- dmontuori has become available [07:08:36] --- matt has become available [07:43:11] --- abo has left [07:44:20] --- reuteras has left [07:45:41] --- SecureEndpoints has become available [09:22:20] --- dev-zero@jabber.org has left [12:09:51] in Init_VLdbase, it looks like if we go to pass 2, we leak a transaction. at the end of pass 2 we EndTrans, but not at the end of pass 1. comments? [12:14:21] afs event 0,133 ? [12:14:37] er, what? [12:15:03] a useful afs windows error message [12:16:09] so it had nothing to do with my question. fine. see also "i hate unthreaded chat systems" [12:16:43] shadow: Do you not AbortTrans in any codepath that means you go to pass 2? [12:17:17] wasn't this thing supposed to be threaded? [12:17:38] we can't use the threading, apparently. [12:17:50] like, apparently the only threaded choice is zephyr [12:18:00] There's no client support. I think the protocol might let us do so, but none of the clients care. [12:18:11] But, anyway, we digress. [12:18:18] at least unlike irc there's not the nickserv stupidity. i am now unable to auth as dbrashear on freenode and they can't help me for now [12:18:48] that's a business decision [12:19:01] what is? [12:19:02] They can't help you or they won't? [12:19:14] nickserv is stupid regardless. it's not real authentication [12:20:00] My reading of the Init_VLdbase code is that the only way you make it into pass 2 is if CheckInit returns VL_EMPTY, in which case you've called AbortTrans() [12:20:01] So defeat it [12:20:06] simon, i guess you're correct. the one case we don't abort we return [12:20:27] Yup. The code could be ... clearer [12:38:17] --- mmeffie has become available [12:58:55] Rrrrred: where are you getting 133? [15:37:30] --- mcohan has left [15:44:03] --- matt has left [15:46:26] --- mcohan has become available [15:58:11] Is the uniquifier guaranteed to always start at 1, or is it possible to have a valid file with a 0 uniquifier ? [16:05:20] --- dmontuori has left [16:07:43] I believe it always starts at one because every root directory I've ever seen is 1.1 [16:08:42] Cool. I just found somewhere in disconnected that assumed that, and it occurred to meet that I'd never seen it explicitly stated [16:18:52] the thing I would be more concerned about is what happens when it wraps [16:20:13] Indeed. I think I need to improve afs_dir_Enumerate a little further ... [16:21:20] I'm going to have to rethink all of this stuff for hardlinks, anyway, so I'll probably just deal with it then ... [16:48:18] --- mcohan has left [16:48:36] --- mcohan has become available [16:54:50] --- mcohan has left [16:54:51] --- mcohan has become available [17:36:50] --- mmeffie has left [18:06:48] --- matt has become available [18:17:34] --- Russ has left: Disconnected [18:36:45] --- Russ has become available [19:38:45] --- matt has left