[00:20:39] --- Simon Wilkinson has left [01:01:29] --- lars.malinowsky has become available [01:21:07] --- lars.malinowsky has left [01:21:17] --- lama has become available [02:43:43] i'll bet the generating toolchain isn't updated. sadly as i discovered the last time i looked into this, i don't know *where* it's generated [02:58:51] it means that =include doesn't work on windows [03:07:41] --- jaltman has left: Disconnected [03:43:36] Eek! Gerrit mis-cherry-picked http://gerrit.openafs.org/3999 onto openafs-stable-1_4_x and broke it. I submitted the diff as http://gerrit.openafs.org/4057 . [05:58:16] --- haba has become available [06:45:41] --- reuteras has left [07:11:56] --- deason has become available [07:16:50] are the man pages generated on the windows buildbot slaves? I don't see it, but I dunno when in the build it would be / what it would look like [09:37:43] --- haba has left [09:41:52] --- sxw has become available [09:56:04] I suspect we should move documentation generation to openafs.stanford.edu [09:56:55] The bad cherry pick is probably due to gerrit's new merge logic. That's still marked as experimental, and we could disable it. [10:06:51] --- sxw has left [11:22:18] This is --use-content-merge? [11:23:29] Oh dear god. gerrit-server/src/main/java/com/google/gerrit/server/git/MergeOp.java: m = MergeStrategy.SIMPLE_TWO_WAY_IN_CORE.newMerger(db); [11:24:05] No wonder it doesn’t work. Two-way merge is a fundamentally, irreparably broken algorithm. [11:25:15] Wait, maybe that doesn’t mean what I think it means. That’s the other case. [11:39:10] Complained to http://code.google.com/p/gerrit/issues/detail?id=859 [15:42:02] --- pod has left [15:46:46] --- Simon Wilkinson has become available [15:50:57] --- Simon Wilkinson has left [15:53:32] --- meffie has left [15:58:19] --- deason has left [16:03:06] --- jaltman has become available [16:22:48] --- jaltman/FrogsLeap has become available [16:23:37] --- jaltman/FrogsLeap has left: Disconnected [16:48:43] windows buildbot slaves build all documentation, verify all docbook, and generates all installers [16:50:06] the problem is now fixed. the NTMakefile was missing clean rules [17:12:04] --- rra has left: Disconnected [17:28:06] --- Russ has become available [20:30:01] It would be nice if running an old afsd against a new kernel module failed more gracefully. [21:25:22] Russ, do you want to tell me how ugly amd64_kfbsd81_deb60 is? [22:16:25] * Russ hehs. It's not horrible. [22:17:07] The _deb60 probably shouldn't be part of the in-tree sysname, I suspect; we've mostly not been putting the libc version into the in-tree one. [23:37:59] --- Russ has left: Disconnected