[00:11:36] --- dev-zero@jabber.org has left [00:14:37] --- reuteras has become available [00:20:22] --- sxw mobile has become available [00:22:45] --- haba has left [01:33:09] --- haba has become available [01:35:46] --- kaj has left [01:38:37] --- abo has left [03:16:03] --- sxw mobile has left [03:26:43] --- sxw mobile has become available [03:31:00] --- abo has become available [03:34:06] --- sxw mobile has left [03:34:31] --- kaj has become available [04:47:35] --- sxw mobile has become available [04:49:07] --- sxw mobile has left [04:52:29] --- sxw mobile has become available [05:01:13] --- meffie has left [05:04:33] --- haba has left [05:56:47] --- sxw mobile has left [06:00:49] --- abo has left [06:01:03] --- abo has become available [06:10:32] --- haba has become available [06:23:40] --- sxw mobile has become available [06:52:18] --- deason has become available [07:37:10] --- sxw mobile has left [07:40:37] --- sxw mobile has become available [07:57:34] --- haba has left [08:07:14] --- sxw mobile has left [08:25:05] --- Kevin Sumner has become available [08:50:13] --- jaltman has left: Disconnected [08:58:58] --- haba has become available [09:00:12] --- reuteras has left [09:09:45] --- sxw mobile has become available [09:20:54] --- sxw mobile has left [09:26:39] --- sxw mobile has become available [09:39:12] --- Simon Wilkinson has become available [09:41:49] robots.txt? [09:46:25] --- Simon Wilkinson has left [09:48:15] --- Simon Wilkinson has become available [09:52:34] --- sxw mobile has left [09:53:12] There was some discussion over lunch that we might be switching from Heimdal to MIT (This again? sigh). One of my colleagues is under the impression that the newer version(s) MIT Kerberos don't support krb4, only 5. Is this true? And if so, is there something special we need to do to MIT Kerb while building to make sure that AFS continues to work? [09:54:15] * Kevin Sumner crosses fingers and hopes colleague is wrong. [09:57:39] Yes, MIT have dropped v4 support. [09:58:01] No, that doesn't break OpenAFS. All modern versions of OpenAFS can use Kerberos v5 tickets, as long as those tickets use DES encryption. [09:58:45] What does break OpenAFS is the fact that the forthcoming MIT Kerberos 1.8 release disables DES by default at the client side. [09:59:12] and the key point is only the afs service principal needs des encryption (the "afs/" or plain "afs" key) [09:59:15] We asked MIT for an interface to selectively enable DES again for us, and they've provided this. aklog from OpenAFS 1.4.12 onwards will work with MIT Kerberos 1.8 "out of the box" [10:00:38] Alright, good. Thanks. [10:14:19] Simon Wilkinson: I would assume we'd want afslore to appear in search engine results; can't just block everything from robots.txt [10:14:57] oh, crap. I used the wrong account again. for the mods on openafs-devel, would you mind approving my message? or you can delete it, and I'll resend with the proper account [10:15:26] deason: We should be able to block the CGIs, and point the robots at static pages. [10:15:49] Better yet, we should be able to block the robot that's swamping the server. [10:15:57] summatusmentis: I'll take a look. [10:16:08] it's a reply to your mail anyway [10:16:14] I thought the wiki pages (the useful information) was CGI, but I wouldn't know [10:16:15] Any idea which robot is clogging things up? [10:16:34] summatusmentis: "that would be my fault"? looks like it's already there to me [10:16:41] Derrick knows - afslore runs on one of his systems, when we couldn't get the version on grand upgraded. [10:16:54] deason: I jsut got a mail about needing mod approval [10:17:01] but yes, that's the right one [10:17:20] oh, you know what? ignore me [10:17:33] when I hit reply all, it copied openafs-doc on the mail [10:17:37] Oh - is that for the openafs-doc list? [10:17:40] I've got one of those too. [10:17:52] I wonder how many people on -doc aren't on -devel. [10:18:06] --- abo has left [10:18:14] yeah, the system load on the web server is back up to 60 [10:18:43] --- abo has become available [10:19:29] shadow: Which robot? Can we just drop the IP? [10:19:45] i did drop the ip. something else showed up now [10:19:57] the real answer is perl (twiki) is too damn heavy [10:20:18] I do wonder about Russ's git-wiki thing. [10:20:25] which? [10:20:31] We could probably get that up and running on o.s.e without much effort. [10:21:08] how do you gateway a cgi into it for people who don't have git? if we could do that, well, hell, we could do it today with the crappy web site we have [10:21:37] Yes. [10:21:37] ikiwiki? That has both a Git interface and a CGI interface. [10:21:39] That's the whole point. [10:21:42] and as far as robots.txt, the yahoo cralwer seems to, um, not so much obey it [10:21:45] That's the one. [10:22:14] how do we import content? is it packaged for debian? [10:22:34] Import content: Don't care. We move the stuff we want. [10:22:35] --- deason has left [10:22:38] ah, it is [10:22:47] Packaged For Debian: If Russ is using it, I suspect it will be. [10:22:49] ideally i'd rather move it all. [10:22:52] and it is packaged [10:23:11] Well, at the moment we have 0% usable wiki content. [10:23:13] --- haba has left [10:23:32] wait another 30 seconds. then 100% will be useable until the crawler comes along again :) [10:23:37] I think there are scripts for converting various formats to ikiwiki. What is it in right now? [10:23:43] TWiki [10:23:53] http://ikiwiki.info/tips/convert_MoinMoin_and_TWiki_to_ikiwiki/ [10:24:09] If we know the UA of the crawler, we could get Apache to black hole it. [10:24:17] --- deason has become available [10:24:33] twiki, iirc [10:24:43] oh, wow, scrolling down is good [10:24:44] --- abo has left [10:25:13] --- abo has become available [10:26:15] --- haba has become available [10:27:06] wtf. iptables with drop is not so much dropping. [10:27:51] Is it below your established/related rule? [10:28:35] i restarted the web server. no connections are established. [10:28:40] hrrmmm. [10:28:43] (well, "were" established) [10:30:05] in any case, it's useable at the moment [10:31:14] --- abo has left [10:31:55] --- abo has become available [10:33:19] I really should just find some time and write pinning. [10:34:00] --- kaj has left [10:39:02] if you're bored, write persistence. pinning, eh :) [10:39:24] Not bored, sadly. [10:40:27] But yeh, persistence is more important. It's also harder, because you need to not just persist the vcache, but every other structure the cache manager uses. (cell information, volume information, ... ) [10:40:49] And then you need to do the right thing with CStatd... [10:41:59] --- jaltman has become available [10:47:21] --- sxw mobile has become available [11:13:21] --- Simon Wilkinson has left [11:13:21] --- Simon Wilkinson has become available [11:13:37] --- Simon Wilkinson has left [11:14:12] --- sxw mobile has left [11:17:30] --- jaltman has left: Disconnected [11:24:44] --- jaltman has become available [11:25:49] --- Simon Wilkinson has become available [11:37:03] --- sxw mobile has become available [11:40:22] --- kaj has become available [11:40:42] --- kaj has left [11:40:46] --- kaj has become available [11:44:08] --- Russ has become available [11:48:24] andersk: I'll fix your gerrit auth problem just as soon as I'm on a network that won't drop out when I'm in the middle of a database transaction. [11:49:19] Thanks. [12:13:46] --- Simon Wilkinson has left [12:15:19] --- sxw mobile has left [12:56:58] --- sxw mobile has become available [12:59:06] --- kaj has left [13:13:02] --- dev-zero@jabber.org has become available [13:13:34] --- dev-zero@jabber.org has left [13:17:00] --- Kevin Sumner has left [13:18:38] --- Russ has left: Disconnected [13:20:00] --- Simon Wilkinson has become available [13:24:29] --- dev-zero@jabber.org has become available [13:29:25] andersk: You around? [13:29:30] Hi. [13:29:36] Could you try gerrit now? [13:29:54] --- Russ has become available [13:30:00] “Contact site administrator” [13:31:11] Also, when I log in with https://login.launchpad.net/+id/TKJNYQ7 , I get the wrong account (1000038). [13:31:34] That's easy to fix, I think. [13:31:42] Need to see what's wrong with the first one. [13:33:28] Ah. [13:33:38] The problem is that you delegate, but both IDs are registered with gerrit. [13:33:43] (with different user IDs) [13:33:52] So the second problem is causing the first. [13:34:02] My real account is 1000027. I ended up with a bunch of bogus accounts while testing a while ago; I think they include 1000036, 1000038, 1000039, 1000040. Will it work if you delete those? [13:34:29] It should do. I wonder if you ever used them ... [13:35:11] I don’t think so. [13:36:21] --- sxw mobile has left [13:37:49] Try now ... [13:38:31] That worked, thanks. [13:38:42] Yay! Hopefully that's all fixed for you now ... [13:46:59] --- kaj has become available [13:53:47] --- kaj has left [13:53:57] --- kaj has become available [14:13:57] --- bpoliakoff has become available [14:15:11] --- bpoliakoff has left [14:43:54] As Jake has reminded me, GSoC organizational deadlines are in two weeks. If there are any developers that are interested in mentoring, please let me know. I will not be mentoring this year although I will be acting as the administrator. [14:46:05] --- kaj has left [15:12:40] --- dev-zero@jabber.org has left [15:28:25] --- deason has left [17:01:38] --- deason has become available [18:45:05] --- Russ has left: Disconnected [20:14:10] --- Russ has become available [21:16:00] --- abo has left [21:16:41] --- abo has become available [21:19:39] --- haba has left [21:22:34] --- haba has become available [21:42:06] --- reuteras has become available [21:50:28] --- deason has left [23:36:09] --- Russ has left: Disconnected