[01:27:08] --- haba has left [02:03:30] --- haba has become available [02:35:23] --- Russ has left: Disconnected [02:46:20] --- dev-zero@jabber.org has left [02:49:08] --- dev-zero@jabber.org has become available [05:16:59] --- pod has become available [05:56:17] --- jaltman has left: Disconnected [06:23:05] --- Simon Wilkinson has left [06:43:28] --- mmeffie has become available [06:44:07] --- deason has become available [06:50:08] --- abo has left [07:49:36] --- abo has become available [07:49:56] any of yinz remember if we're not maintaining the web site via only git because grand lacks git, or if there's a better reason? [07:52:18] well, anyway, thereifixedit. [shadow@GRAND htdocs-git]# hostname GRAND.CENTRAL.ORG [shadow@GRAND htdocs-git]# which git /usr/local/bin/git [07:52:42] so i'm also gonna make export_htdocs able to export the web site unless someone screams right quick [07:53:12] (and before jhutz complains, yes it's properly installed and depot copied it in) [07:59:18] --- matt has become available [08:03:05] rx afs3: jhutz: that's great [09:05:30] --- deason has left [09:21:43] --- haba has left [09:23:18] --- gendalia@gmail.com/owlCF44859C has become available [09:24:27] --- deason has become available [09:57:44] > because grand lacks git That's not why [09:59:13] The web site that's in CVS on grand is not going to stop using CVS. AFAIK Russ is working on transitioning to a copy maintained in AFS and managed by git. [10:20:51] --- sxw mobile has become available [10:22:06] --- dev-zero@jabber.org has left [10:23:41] rwe have a copy managed by git. we've been shadowing for weeks. [10:24:32] hence "how i already dealt with this" [10:34:13] --- sxw mobile has left [10:47:42] --- Russ has become available [10:51:10] I think the only piece was missing was for jhutz to have a chance to repoint the grand web server at AFS. [10:52:37] well, for the moment the old process still ~works, but now the repo is git [10:52:46] because, well, i'm lazy [10:52:53] The grand web server won't be repointed; the stuff in AFS is exported by a different server. But I'll be happy to change the DNS aliases around if you think that version is ready; I've just been waiting for the go-ahead [10:53:16] > now the repo is git ... except not the same as the repo Russ set up? [10:53:47] the exact same repo [10:54:24] all i did was: build git on grand. git clone. create a export_htdocs_git script. run it. [10:54:47] Fine. Though I argue that is false laziness. A truly lazy person would have waited for Russ and I to appear and made us deal. :-) [10:55:36] dude, i'd rather you got kula his bits [10:55:44] we have these drafts here and coming [10:55:53] and standardisation needs to move... [10:56:50] jhutz: So far as I know, the version exported into the AFS production volume you gave me should be fine. I just ran the prod htdocs release script to ensure that it's up-to-date. [10:56:52] I think kula should be set up. [11:00:02] The plan has been to migrate what numbers content there is to the new moinmoin, so it no longer lives in CVS on grand in a web site I want to make go away. [11:07:25] --- sxw mobile has become available [11:30:35] --- sxw mobile has left [11:31:57] --- dev-zero@jabber.org has become available [12:51:11] --- Simon Wilkinson has become available [12:51:46] --- Russ has left: Disconnected [12:55:07] --- jaltman has become available [13:48:14] --- Russ has become available [13:59:31] --- jaltman has left: Disconnected [14:03:29] --- jaltman has become available [15:16:14] I don't think whats on HEAD builds with --enable-checking on any platform I have. [15:16:32] That's entirely possible. [15:16:43] See README.WARNINGS [15:17:29] I've made the files I updated in my current changeset build with the various warning flags --enable-checking would entail. [15:17:42] Cool. Thanks. [15:17:59] --enable-warnings will give you the same flags as --enable-checking, just without the fatalness. [15:18:24] Ok. [15:18:43] The first fatal I hit here is in XDR generated stubs. [15:19:25] or, rather, rpcgen [15:19:28] > I've made the files I updated in my current changeset build... Did they build that way before you made your changes? [15:19:31] er, rxgen [15:20:08] I think they did, but, I split the changes in half. Simon may have been complaining about something in the other half. [15:20:53] We actually supress the fatalness in most of the rxgen generated files [15:21:10] ok, my Makefile hack didn't quite make it [15:21:54] rxgen has the whole function-taking-an-array-as-a-parameter problem. [15:22:09] I guess the first fatal I hit in a normal compile is something in ptserver/db_verify.c [15:22:51] what is the error? and location? [15:23:18] If it's a 64bit platform then it's not that exciting. [15:23:23] yah [15:23:34] We know we've still got 64bit errors. [15:23:58] That's why --enable-checking isn't the default. It's there to stop the appearance of new errors, not to magically make all of the existing ones go away. [15:24:11] (Although, if it could do that, that would be very useful) [15:29:00] In any event, I am now defeated by trying to figure out writeback congestion deadlocks in Linux, and am going to sleep ... [16:10:21] --- matt has left [17:08:57] --- jaltman has left: Replaced by new connection [17:08:58] --- jaltman has become available [19:07:42] We should point AFS contributors at the local Git commit hook that fails commits if there is trailing whitespace. [19:07:44] > I don't think whats on HEAD builds with --enable-checking on any > platform I have. anything outside platform-specific? [19:08:16] There's a commit hook in the Pro Git book that does that. [19:09:45] we should [19:10:44] --- Russ has left: Disconnected [19:21:49] --- dwbotsch has left [19:22:15] --- dwbotsch has become available [19:45:09] --- jaltman has left: Replaced by new connection [19:45:10] --- jaltman has become available [20:19:05] --- Russ has become available [21:07:11] --- deason has left [21:07:28] --- stevenjenkins has left [21:07:28] --- stevenjenkins has become available [21:07:41] --- deason has become available [21:58:41] --- deason has left [21:59:25] --- deason has become available [22:20:38] --- deason has left