[01:09:21] --- shadow@gmail.com/owlE16C8B4F has left [03:54:43] --- manfred furuholmen has left [05:21:53] --- tkeiser@sinenomine.net/owl has left [05:34:31] --- Claudio Bisegni has become available [05:36:36] --- Claudio Bisegni has left [05:43:26] --- shadow@gmail.com/owl98511FE0 has become available [06:06:06] --- matt has become available [06:46:34] --- tkeiser@sinenomine.net/owl has become available [07:10:03] --- dmontuori has left [07:10:13] --- dmontuori has become available [07:22:32] --- manfred furuholmen has become available [07:29:14] --- Derrick Brashear has left [07:35:08] --- floh has left [07:42:31] --- summatusmentis has left [07:44:53] Hi folks. Simon Wilkinson around? [07:52:20] --- Russ has become available [07:56:13] --- Derrick Brashear has become available [07:58:34] simon's busy lighting a show [07:58:51] AH! (nice) [07:59:10] what do you need/ [08:00:43] I just wanted to ask if even the latest and greatest ssh+gssapi-kex does log logins on 2 different lines (principal here and host there) and if there is a wishlist if not ;) [08:01:38] But now I need a phone. [08:03:14] less than you think. the bridge is unhappy [08:04:42] Yeah, no workie. [08:05:23] Answer from protein based service: "Holder of conference call not active" [08:06:24] --- reuteras has left [08:08:16] Back on the openssh issue I thought Simon would know the answer right away, so the lazy person I am.... But I don't need the answer here and now, so you all can keep on walking (nothing to see here :) [08:10:21] russ and haba see tom's mail [08:10:49] TX [08:14:21] --- dev-zero@jabber.org has left [08:49:26] --- manfred furuholmen has left [08:53:17] --- manfred furuholmen has become available [09:12:49] --- Derrick Brashear has left [09:32:40] --- dev-zero@jabber.org has become available [09:53:54] --- haba has left [09:53:54] --- haba has become available [10:11:28] --- summatusmentis has become available [10:30:08] --- summatusmentis has left [10:57:01] --- Derrick Brashear has become available [10:58:47] --- dev-zero@jabber.org has left [11:01:56] --- haba has left [11:08:42] --- manfred furuholmen has left [12:03:56] --- summatusmentis has become available [12:43:17] --- Derrick Brashear has left [13:13:33] --- SecureEndpoints has left [13:14:42] --- Derrick Brashear has become available [13:45:36] --- dev-zero@jabber.org has become available [14:18:14] --- manfred furuholmen has become available [14:32:28] --- summatusmentis has left [14:39:15] --- summatusmentis has become available [14:44:57] derrick: can you check this git conversion http://www.beolink.org/downloads/git ? [14:46:21] some clean up still needed .. [14:47:35] in a half an hour, if i did not drink myself under the table, i will tell you more. [14:48:51] .-) [14:49:24] did you check out the draft plan i offered for releases? [14:49:46] I think a lot of things must be fix [14:49:59] in git or in my plan? [14:50:11] in git [14:50:37] tommorrow i will do some test on the code base [14:50:49] i didn't read you plan [14:50:53] ok. that's not shocking. did you check the scripts mike meffie posted the other day? [14:51:01] --- dev-zero@jabber.org has left [14:51:02] no [14:51:17] you should read my plan. i'm not doing NaNoWriMo, so if you want to read my great works it's that or my blog. [14:51:40] and meffie scripts ? [14:51:41] ok. now it tells me i am waiting 49 minutes [14:51:45] hang on [14:54:03] /afs/sinenomine.net/public/openafs/projects/git_work [14:54:35] I know too many people doing NaNoWriMo [14:57:04] --- SecureEndpoints has become available [15:00:27] --- manfred furuholmen has left [15:12:15] I need a non-sucking jabber client [15:14:22] depends what it sucks [15:19:06] --- Simon Wilkinson has become available [15:20:10] Im using ichat [15:20:12] :( [15:20:31] Well, that sucks for MUC. Not bad for person to person. [15:20:38] I like Adium and owl. [15:20:53] I like iChat over Adium, I don't know how to use owl [15:21:00] cli client? [15:21:11] barnowl, really [15:21:19] Yes, it's a multi-protocol cli client. [15:21:55] Do we think shadow is under the table yet? [15:22:01] I should get kula to install it [15:22:39] unless he dropped below in the last three minutes, or has his net connected device down there with him, I'd say no [15:23:02] owl rocks my wold, but i want twitter [15:23:19] shadow needs at least 1 more to be really drunk [15:23:21] you want cli twitter? [15:23:32] i want owl twitter. [15:23:36] yeh [15:23:38] i want *one* IM client [15:23:40] Really drunk, or Munich drunk? [15:23:53] twitter isn't quite IM, but fair enough [15:23:56] really drunk. munich drunk would preclude even typing [15:24:04] I tend to like GUI clients, except irc [15:24:11] 7 minutes of happy hr. where's lexi. [15:24:19] gui clients fail in screen [15:24:31] what?!? since when? :-P [15:24:44] you use owl for irc too? [15:24:45] where is bacon? [15:24:56] bacon at 6:30 [15:25:05] i used owl for aim irc jabber zephyr [15:25:34] no aim ATM [15:25:51] not enough of my friends use jabber [15:26:11] Get a better class of friends? :) [15:27:08] i agree with simon [15:29:02] ok, ups calls, they say they are delivering a package tomorrow and here is the tracking number. the web site says the number is not valid. what's up/ [15:29:15] nice [15:29:20] scam? [15:30:02] to make me do what? enter a tracking number into google? [15:30:02] shadow: Because I'm feeling generous, repo RPMs are now in the new usual place. [15:30:18] thanks! [15:31:14] your generosity in this time of drunk is appreciated [15:31:41] I wish I had drunk more. Now I'm just fall-over-tired ... [15:32:20] none of my friends are terribly geeky in any way [15:34:13] ups website takes a while to get tracking numbers, seemingly [15:35:17] tired is good. i slept 4.5hours last night [15:37:40] sleep is better [15:38:05] sleep is for pitt students [15:38:23] moose says that too. I disagreee [15:38:25] -3 [15:38:40] er... +3 -e [15:39:59] -3 what? [15:40:09] I hit 3 instead of e [15:40:21] stop that [15:40:35] I bet the wine will show up before the tracking number hits the web site [15:40:38] 3 is e to one significant digit. [15:40:52] 1'm jus+ s0 1337 [15:40:58] I feel so dirty [15:41:11] |>00|>! [15:41:41] You are bringing back my misspent youth writing IRC bots. [15:42:06] I considered writing an IRC bot once... [15:42:18] It's a fun way to learn TCP networking. [15:43:05] i've been to 13375d@13 [15:43:27] it's a good place to tak train pictures [15:43:33] take [15:44:14] manfred: i have it all. i will check when i can actually parse the result and not have it hurt my brain [15:46:44] --- summatusmentis has left [16:00:24] --- Simon Wilkinson has left [16:06:58] i can do 1.4.8 if i can just edit this perl script [16:07:05] lesse if i am sober enoygh [16:07:13] --- Marc Dionne has become available [16:10:15] If I can make a little suggestion for the 1.4.8 release page: s/Fedora Core/Fedora/ - the last "Fedora Core" was FC6 [16:20:07] the release page is made by a script. let me try to edit it and pray my emacs skills stil work [16:23:03] emacs skills: still work [16:23:50] always avoided getting any emacs skills [16:24:43] fixed in the web site in about 45 seconds [16:25:00] weel the issue is how fast i shed them when drinking [16:25:27] there "Fedora" [16:25:29] should 2.6.28 support be mentioned in the release notes? or is it goofy to support an unreleased kernel [16:25:57] rlrease notes written before we applied the patch. rpms too late to respin [16:26:18] assume "it was already supported" unless we say otherwise? [16:26:18] what? no cool code name for the release? [16:26:31] "septuple revenge" [16:26:39] octuple [16:26:48] pretend math still works [16:28:00] ah, I see "Fedora" now -thanks [16:28:38] yeah. i wonder if i can send the announcement without botching it [16:33:29] there. have some openafs 1.4.8 [16:36:04] announced, and tweeted. [16:37:42] there's an openafs twitter? [16:37:56] doesn't look too botched [16:38:02] twitter.com/openafs [16:38:18] won't follow you back as fast as barack obama, but... [16:39:00] kids and their twitter and their facebooks and whatnot. [16:39:15] will i be able to get an sms message the next release? [16:41:18] sure. [16:41:24] cuz we love you long time [16:41:48] can interrupt your date so you can punt that guy you hate and go home. "oops i have a priuor engagement" [16:43:55] ENOTMAKINGSENSE [16:46:10] --- summatusmentis has become available [16:48:29] sorry. i will drink more [16:48:30] sms [16:48:39] rhetorical [16:48:57] --- summatusmentis has left [17:02:18] --- dmontuori has left [17:07:56] --- edgester has become available [17:08:36] hi everyone [17:12:58] hi doctor nick [17:18:37] I'm reviewing the fileserver man page and noticed the "16 character" file name language [17:18:59] should that be changed to "16 bytes" since unicode is now in play? [17:19:31] --- matt has left [17:39:44] --- summatusmentis has become available [17:50:31] edgester - which version? [17:51:00] cvs [17:51:05] head [17:51:30] not sure which part you're talking about..openafs-devel-1_5_x is what I'm looking at, but I can switch. [17:51:34] can you give a larger context? [17:51:59] in doc/pod8/fileserver.pod [17:52:24] ah, you mean the '15' character..not '16'. [17:52:29] it talks about the 64000 file limit in a directory and that is with filenames <16 chars [17:52:36] ye [17:53:19] I don't believe secureendpoint's changes to add unicode support touch the fileserver [17:53:29] I could be wrong, though. [17:54:33] no, they don't [17:54:44] Yeah, but that sort of supports the point. [17:54:51] The client stores Unicode characters, but the server sees bytes. [17:55:02] And the underlying data format stores bytes, so the 15-character constraint is really about bytes. [17:55:04] but the <15char <> 15 bytes with unicode [17:55:27] they do not [17:55:45] so s/15 chars/15 bytes/ ? [17:55:48] no [17:55:57] chars are not bytes [17:56:06] it would be 63 bytes [17:56:19] (or 60, i need to look) [17:56:54] well, from a unicode standpoint, the 15 bytes could map to 4 chars or less [17:57:32] 15 characters is not 15 bytes [17:57:38] you areconfused [17:57:45] http://picasaweb.google.com/shadow/SouthSideNeighborhoodWalk#5267260087081190322 [17:57:48] er, oops [17:58:06] the 15 characters in the pre-unicode word is 60 bytes. not 15. [17:58:17] 15 bytes was *never relevant* [17:58:29] huh? [17:58:37] your choices are "less than 64" or "4 times 15" [17:58:47] nothing was ever limited to 15 bytes [17:58:51] * Russ is pretty sure that the documented restriction on the size of directories is based on a size of filename buckets which is expressed in bytes. [17:58:53] the max files in a dir is 64000 when filesnames < 16 bytes [17:59:14] --- G6x7CwKmDOwY2 has become available [17:59:31] er [17:59:33] crap [17:59:35] bits [17:59:51] * Russ changes the man page to express the limit in number of beers. [18:00:01] yes. russ is correct. [18:00:38] --- G6x7CwKmDOwY2 has left [18:00:44] man-page-fileserver-octet-20081111 [18:00:54] woot [18:01:05] ok, so should I replace "characters" with "bytes" int he context of filename length re ma files in a dir [18:01:07] ? [18:01:20] * Russ used octets since it's the IETF-blessed term. [18:01:33] ok [18:01:46] --- summatusmentis has left [18:01:48] --- summatusmentis has become available [18:02:02] should I updated my cvs copy to integrate those changes? [18:02:19] sorry, playing with new clients, is this working? [18:02:29] hi summ [18:02:34] Yup! It's in head already, pulling up now. [18:02:39] hi edgester :) [18:03:19] * Russ occasionally has to change things in OpenAFS to feel like a gatekeeper. ;) [18:03:23] russ, no [18:03:28] "Each additional block provides an additional length of 32 ! octets for the name of the entry." [18:03:30] try 16 [18:03:57] That bug was already in there. [18:04:09] slot 1: 0-15. slot 2: 16-31. slot 3: 32-47 and so on [18:04:10] ok [18:04:12] I didn't change that bit. But I can right now. You sure on 16? [18:04:32] blob is a blob. 16 is 16 [18:04:51] I was assuming that it was 32-byte blobs and the first blob stored additional metadata, but I'll certainly take your word for it. [18:04:59] i will check [18:06:23] I have a patch for fileserver.pod ready, how should I proceed? [18:06:48] looks like i am right [18:06:51] todd agrees [18:06:53] http://www.openafs.org/pipermail/openafs-info/2002-September/005812.html [18:07:03] edgester: I dunno -- what is it addressing? [18:07:07] wait [18:07:10] let me... [18:07:17] no, i am wrong? [18:07:22] stil reading code [18:07:46] s/vicepzz/vicepiv/ and made partition limits clear [18:08:01] i should just try the shit [18:08:37] i think i was right the first time [18:08:43] (and so was todd) [18:08:55] edgester: Go ahead and merge it with current unstable and send it to me, and I'll do something or other with the file bucket thing at the same time. [18:09:50] sent [18:14:41] i have an inkling that the subsequent blobs hold 32 octets. [18:15:39] they should but i don't think they do [18:15:50] i may be reading the code wrong [18:17:20] i'm gonna try it [18:19:28] i'm reading src/dir/dir.c, it looks like it gets some number of blobs, fills in a flag, the vnode and uniqifier, and then just splats the name after that. [18:20:09] edgester: committed. [18:20:20] thanks! [18:21:24] that's what i have open [18:23:16] i am doing the math inGetBlob. i wonder if i am loomgink at this wrong [18:27:23] the math in NameBlob seems to be easier, and that's what Create uses to determine how many blobs it needs [18:27:39] Yeah. [18:27:47] And it says 32 bytes for each additional block. [18:27:51] and it would seem to support ==russ [18:28:20] There's basically 15 bytes of overhead added on to the length of the filename plus one. [18:28:52] yeah, NameBlob does seem sto yeah [18:28:59] I wish it weren't >> 5 instead of a divide that would be more intuitive. [18:29:15] But >> 5 is / 32. [18:29:28] yeah, i had to wake up some dull bits of my brain to do that [18:29:51] Okay, I'm reverting my change and putting it back at 32. :) [18:30:05] i used bc. i can't do bitshift now [18:31:47] actually, this is starting to make more sense than the last time i looked at it. i should move my python package to munge volume dumps closer to the top of The Project List(TM) [18:32:11] yeah so i should shut up while i am not able to do math [18:40:07] 1.4.8 annoucement submitted to freshmeat [18:41:45] gmail now has a/v built into the web client :) [18:42:24] --- tkeiser@sinenomine.net/owl has left [18:43:49] woot thanks [18:48:23] --- Derrick Brashear has left [18:49:24] --- Marc Dionne has left [18:49:42] --- tkeiser@sinenomine.net/owl has become available [18:51:46] shadow_gmailcom: what's outstanding yet for the foundation to happen? [18:52:01] out of curiousity [18:53:05] shadow@gmail.com/owl98511FE0: ugh, 1.4.8 is out, now I gotta push out a new mameo package [18:53:10] shadow@gmail.com/owl98511FE0: :P [18:53:25] good night everybody [18:53:45] night edgester sleep well [18:54:15] --- edgester has left [18:56:10] --- summatusmentis25565 has become available [18:56:10] --- summatusmentis25565 has left [18:56:10] --- summatusmentis91681 has become available [18:56:31] --- summatusmentis91681 has left [19:18:31] --- Derrick Brashear has become available [19:25:51] alas, sobriety returns [19:25:56] jake still here? [19:28:33] yep [19:39:55] i fail. what'd you last ask? [19:40:57] just what is left blocking the foundation from happening [19:41:55] ah, yes. [19:42:28] we need to do some more work on filling in some missing details in the charter, and then there will be one more comment period [19:42:47] I see [19:42:59] and what's required for an individual membership? [19:56:06] contributions, which are not necessarilt monetary in nature [19:56:50] > i want owl twitter. So write it. [19:57:37] derrick, I see, what's the monetary amount? [20:00:35] lazy [20:00:41] derrick, jaltman, anyone-else-who-cares: http://www.dementia.org/twiki/bin/view/AFSLore/OpenAFSWebsiteProjectPage [20:03:07] > it would be 63 bytes No, it would not. It would be 15 bytes, because the rule is "you can have 64000 filenames if each one fits in a single blob", and the initial blob of a directory entry has room for 15 bytes. [20:03:47] yes. catch up [20:07:29] > "Each additional block provides an additional length of 32 > ! octets for the name of the entry." > try 16 Try 32. Have you actually _read_ the directory code? [20:07:55] > I was assuming that it was 32-byte blobs and the first blob stored > additional metadata, You were assuming correctly. [20:09:47] your ability to catch up sucks [20:49:30] summatusmentis: thank you. I would look at it now but I've got an early morning flight to catch [20:49:57] no worries, whenever. I sent out an email notifying everyone also. Where are you off to? [20:50:08] logging for this conference appears to be broken. [20:50:26] or maybe not. [20:56:24] Boulder [20:56:46] I'm glad I missed the misdiscussion of the directory entries. [20:57:14] I'm a big fan of the Rockies, beautiful [20:57:20] it is [20:57:26] but no wifi on the flight [20:57:52] is that even remotely normal? [20:58:06] I suppose I don't fly to populated areas very often [20:58:23] american now has wifi between new york and many major cities [20:59:04] ah, that's awesome :) [21:00:16] which will make it harder to figure out why the connection cloning patch doesn't appear to result in multiple connections being established between my client and the file server. [21:00:21] anyway, good night all. [21:00:54] sleep well [21:58:32] --- manfred furuholmen has become available [22:06:35] shadow: where can i find your blog ? [22:07:39] --- dev-zero@jabber.org has become available [22:12:17] --- manfred furuholmen has left [22:55:26] --- reuteras has become available [23:10:48] --- Russ has left: Disconnected [23:35:10] --- dev-zero@jabber.org has left