Home
ejabberd
ejabberd@conference.jabber.ru
Wednesday, 15 August 2007< ^ >
badlop has set the subject to: Здесь говорят только по-английски! | Only english here! | website: http://ejabberd.jabber.ru/ | Erlang/OTP R11B-5 released; any interesting feature?
Room Configuration

GMT+4
[00:15:21] fox_a35 leaves the room
[00:30:00] kostix joins the room
[00:57:01] fox_a35 joins the room
[01:03:03] Christian leaves the room
[01:15:24] ermine leaves the room
[01:15:42] ermine joins the room
[01:25:35] <badlop> deryni: hi
[01:27:58] <badlop> are you Etan Reisner, who sent an email to ejabberd mailing list 2 years ago expressing interest in XEP-133: Service Administration?
[01:29:20] <badlop> http://article.gmane.org/gmane.network.jabber.ejabberd/634
[01:31:09] <deryni> Hah, I am.
[01:31:33] <badlop> did you submit a patch about xep133?
[01:32:20] <deryni> I don't recall exactly, but I don't believe I ever got around to it.
[01:33:05] <badlop> ah, ok; i ask because i my gsoc project includes implementing it in ejabberd
[01:33:52] <badlop> and i'm checking the 'state of the art' :)
[01:34:58] <deryni> Yeah, I believe that was a future goal of the project I was working on that never went anywhere, but I can look later and see if I ever started work on it.
[01:35:26] <badlop> if you don't happen to find anything, don't worry
[01:36:36] <deryni> Unfortunately I left the job that was having me do xmpp/ejabberd work and so I haven't touched it all in a while. Though I'm glad my muc default options patch went in, that's nice.
[01:40:58] ermine leaves the room
[01:41:58] <badlop> yes, it took him 14 months, but your patch finally found his path to svn :)
[01:47:14] fox_a35 leaves the room
[02:11:29] thomas joins the room
[02:19:25] <RogerD> badlop: ehm, any chance of fix for the problem I mentioned? :)
[02:23:50] <badlop> check the thread:
in-order delivery
http://lists.jabber.ru/pipermail/ejabberd/2006-March/thread.html
[02:30:22] badlop leaves the room
[02:34:13] teo leaves the room: Replaced by new connection
[02:34:13] teo joins the room
[02:34:26] legoscia leaves the room
[02:37:15] <RogerD> badlop: thanks, but I'm not sure it answers my question... unless it means "we don't mind" :)
[02:38:52] <RogerD> badlop: if I send someone several messages, I expect them to be delivered in that order - and they aren't
[02:39:32] <RogerD> badlop: and in this case, it IMHO doesn't make sense to wait for.. for what? messages don't imply response from server as in linked example with setting privacy lists
[03:00:14] <teo> RogerD: how do you send messages and how do you know that ejabberd reorders them?
[03:01:00] TobiasFar leaves the room
[03:18:36] RogerD leaves the room
[03:18:49] RogerD joins the room
[03:18:55] <RogerD> grm
[03:19:08] <RogerD> teo: sent with my program and checked with wireshark
[03:19:30] <RogerD> teo: messages are sent to server in correct order, but received permuted
[03:20:51] <teo> RogerD: are they all messages from the same JID1 to JID2?
[03:21:11] <RogerD> teo: they are sent to room
[03:21:47] <RogerD> from same JID to same server, room is on that server too
[03:21:59] <teo> the same room?
[03:22:06] <RogerD> that is - jid@server, room@conf.server
[03:22:08] <RogerD> yes
[03:22:24] <RogerD> all from jid@server to room@conf.server
[03:22:41] <RogerD> i'll look for captures, if I still have them...
[03:22:51] <teo> then it's strange. the order must be preserver
[03:25:25] <RogerD> argh
[03:25:29] <RogerD> server reset - brb
[03:25:37] <RogerD> I have those captures
[03:26:56] <teo> RogerD: could you run
tcpflow -c port 5222
and send all the messages, and receive them, and show me the dump?
[03:29:34] <teo> RogerD: (just to make sure) 'server' is running ejabberd and 'conf.server' is running mod_muc?
[03:38:05] <RogerD> teo: it's jabber.cz - AFAIK it's ejabberd, but you can check it yourself
[03:38:20] <RogerD> I think I've even tested it on locally running ejabberd
[03:38:49] <RogerD> teo: capture from previous test is on http://roger.jikos.cz/dump1.pcap
[03:40:30] <RogerD> teo: it gets interesting at timestamp 29.859339, where messages come in wrong order - 48, 52, 49, 51, 50, 53
[03:42:37] <kostix> hm. this packet (#144) looks like complete crap in my wireshark
[03:42:57] <kostix> while #145 is ok and constitutes readable XML
[03:51:45] <teo> there shouln't be this reordering. seems like a bug
[03:54:11] <RogerD> as I thought :)
[03:54:33] elmex leaves the room
[03:54:56] <RogerD> I hope it can be fixed - ejabberd is really popular server
[03:54:58] <teo> the problem is how to reproduce it at a local server
[03:55:11] <RogerD> teo: just spam it, it should help
[03:55:18] thomas leaves the room
[03:55:34] <RogerD> teo: it doesn't happen always, took me several tries to get this capture
[04:01:06] <RogerD> well, it's really late here, I'll be back layter
[04:02:06] kostix leaves the room
[04:16:14] <teo> can't reproduce...
[08:09:21] zapw joins the room
[08:14:07] <zapw> /usr/bin/erlc -W -I .. -pz .. -o .. ELDAPv3.erl
/usr/bin/erlc -W -I .. -pz .. -o .. eldap_filter.erl
make[2]: Leaving directory `/root/ejabberd-1.1.2/src/eldap'
making all in pam
make[2]: Entering directory `/root/ejabberd-1.1.2/src/pam'
/usr/bin/erlc -W -I .. -pz .. -o .. epam.erl
./epam.erl:68: undefined macro ''WARNING_MSG''
./epam.erl:110: undefined macro ''WARNING_MSG''
. /epam.erl:21: function handle_info/2 undefined
./epam.erl:21: function init/1 undefined
make[2]: *** [../epam.beam] Error 1
make[2]: Leaving directory `/root/ejabberd-1.1.2/src/pam'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/root/ejabberd-1.1.2/src'
[08:32:12] zapw leaves the room
[11:02:58] Kev leaves the room
[11:10:24] teo leaves the room: Replaced by new connection
[11:10:24] teo joins the room
[11:14:08] Christian joins the room
[11:14:56] Christian leaves the room
[11:31:47] zapw joins the room
[11:49:31] ermine joins the room
[12:02:26] badlop joins the room
[12:03:27] notKev joins the room
[13:06:46] legoscia joins the room
[13:29:31] niekie joins the room
[13:36:33] Christian joins the room
[13:37:36] niekie leaves the room
[13:38:24] niekie joins the room
[13:57:38] elmex joins the room
[14:05:52] TobiasFar joins the room
[14:31:16] kostix joins the room
[14:39:10] kostix leaves the room
[14:43:52] kostix joins the room
[15:22:30] klaus joins the room
[15:45:06] teo leaves the room: Replaced by new connection
[15:45:08] teo joins the room
[15:47:59] <zapw> where do i d/l svn ver of ejabberd
[15:49:34] <klaus> svn co http://svn.process-one.net/ejabberd/trunk ejabberd
[16:24:02] <zapw> i got some error while trying to compile the epam module for 1.1.3/1.1.2
[16:24:24] <zapw> i just need to get this working , then i wont need the whole svn
[16:32:17] <zapw> how do i use subversion
[16:34:46] <zapw> never mind
[17:00:13] zapw leaves the room: offline
[17:18:17] klaus leaves the room: Logged out
[17:28:14] klaus joins the room
[18:04:57] klaus leaves the room: Logged out
[18:09:43] klaus joins the room
[18:15:25] klaus leaves the room: Logged out
[18:39:49] sss joins the room
[18:41:01] sss leaves the room: offline
[19:42:55] niekie leaves the room
[19:43:33] niekie joins the room
[19:54:36] niekie leaves the room
[19:58:22] niekie joins the room
[20:02:08] notKev leaves the room
[20:29:57] thomas joins the room
[21:03:54] bomer joins the room
[21:06:20] <bomer> i'm running ejabberd 1.0.0 on ubuntu 6.10. I've added: {mod_ctlextra, []}, to my ejabberd.cfg and restarted. when i run ejabberdctl, i do not see any new options. does anybody have any ideas?
[21:06:53] <bomer> errr... 6.06 rather.
[21:13:03] <deryni> Silly question but did you install that module?
[21:19:06] Kev joins the room
[21:40:40] <bomer> yes. i copied the mod_ctlextra.beam file after running ./build.sh in the mod_ctlextra directory under ejabberd-modules. I copied the file to /usr/lib/erlang/lib/ejaberd-1.0.0/ebin/
[21:41:40] sss joins the room
[21:43:32] <deryni> Did you see that the README says "Module for ejabberd 1.1.2 or newer"?
[21:45:28] <bomer> doh!
[21:45:57] <bomer> time to upgrade. thanks for pointing that out lol
[21:45:57] <sulci> bomer: Recipient is not in the conference room
[22:00:16] sss leaves the room
[22:01:26] sss joins the room
[22:05:43] <badlop> bomer: 1.1.3 and older also have this bug: https://support.process-one.net/browse/EJAB-162
[22:06:04] <badlop> howeverm if you know the name of the command, you can use it without problem (i think)
[22:08:18] kostix leaves the room
[22:09:05] <bomer> i really wish shared rosters were able to be shared across vhosts
[22:11:59] <badlop> you can formalize your wish by submitting it to the bug tracker
[22:13:01] <Kev> well, I think psa's got a shared roster xep on his .plan
[22:13:15] <Kev> so it's probably not worth doing much work on them until then?
[22:14:44] <teo> i'd say that shared roster shouldn't require support in clients, so it worth fixing or extending it anyway
[22:14:55] <badlop> in that case, the bug issue will be marked as `won't fix' :)
[22:15:29] <badlop> `won't fix' <--- in case of xep
[22:16:49] <bomer> well, in a shared roster now, i can add members from other vhosts. their presence isn't sent until those users are on the rosters of the newly added members from the other vhost.
[22:17:00] <bomer> and having the ability to use @all@vhost would be nice, too
[22:19:09] <badlop> in any case, note that the future of your feature request is *uncertain*
[22:19:34] <badlop> from what i know, mod_shared_roster was written by aleksey for fun
[22:20:16] <badlop> and the vhost feature was implemented with money from process-one
[22:21:37] <badlop> some people are trying to kick process-one out and away from ejabberd development
[22:21:50] <Kev> badlop: what is the story with that?
[22:21:56] <badlop> so, the future of your feature request is... *uncertain*
[22:21:59] <Kev> that's the second time I've heard you say it in two days
[22:22:16] <badlop> and in two different places
[22:22:24] <bomer> is mod_shared_roster not maintained anymore?
[22:22:24] <Kev> indeed
[22:22:30] <Kev> (@badlop)
[22:25:22] <badlop> bomer: submit the feature request and let's see :)
[22:26:27] <bomer> indeed we shall :)
[22:29:25] <bomer> feature requests are submited via bugzilla?
[22:29:45] <badlop> Kev: i was also surprised to see that obstructionist behaviour in people that i considered was interested in improving ejabberd
[22:30:19] <badlop> bomer: russian people like teo and ermine want bugzilla, and also sander
[22:30:44] <badlop> workers of process-one only read Jira
[22:31:05] <badlop> russian people strongly reject to use Jira at all
[22:31:20] <badlop> for reasons they don't reveal, maybe because Jira uses Java
[22:33:00] <badlop> aleksey, the original author of ejabberd, didn't publicly reveal his opinion yet
[22:34:05] <deryni> I think you (and just about everyone else) has been overreacting to this whole situation.
[22:34:41] <badlop> deryni: ok, and now the solution: where should bomer submit it?
[22:34:56] <bomer> http://www.jabber.ru/bugzilla/show_bug.cgi?id=367 <http://www.jabber.ru/bugzilla/show_bug.cgi?id=367>
[22:35:27] <bomer> it was the first option i found from the web site
[22:36:11] <deryni> The fact that the two sides haven't been able to decide which bugtracker to use is an indication of a lack of desire to actually resolve that issue, likely a desire to want to fight, and unfortunate.
[22:36:35] <deryni> All that being said, given that the people likely to be making the changes have stated that they only really want to use Jira the answer is use Jira.
[22:38:18] <legoscia> they have stated that with their actions, but have they stated that with words too?
[22:39:41] <deryni> Have they stated what? That the process-one people want to use Jira?
[22:41:30] <legoscia> they obviously don't read bugzilla on a regular basis, but i can't remember them having said so publicly…
[22:42:59] <deryni> http://lists.jabber.ru/pipermail/ejabberd/2007-August/002890.html <- For one instance.
[22:44:52] <legoscia> ah ☺
[22:45:19] <deryni> Which isn't exactly 'official word' but is pretty close, from at least one person.
[22:45:27] <legoscia> indeed
[22:45:40] fox_a35 joins the room
[22:45:42] <deryni> Some people even took it upon themselves to migrate patches from bugzilla to jira (which is how one of mine just went in).
[22:46:45] <badlop> deryni: yes, and that patch was awaiting in bugzilla for more than a year, right?
[22:46:52] <fox_a35> Всем привет)
[22:47:11] <badlop> i went with the pragmatic solution: submit to Jira; and i got the patch applied ---> success
[22:47:40] <deryni> Yes, my patch had been sitting around for a while. Had I still been actively working on it and known about Jira I likely would have moved it myself.
[22:47:57] <deryni> I'm not trying to excuse process-one for ignoring the bugzilla.
[22:48:15] <deryni> They must have known about it and could have checked it.
[22:48:42] <deryni> I'm just saying that the 'anger' which this is causing is out of proportion to the problem, as far as I'm concerned.
[22:48:48] Федот joins the room
[22:49:16] <fox_a35> Федот: :-)
[22:50:10] <fox_a35> ping
[22:50:12] <sulci> fox_a35: Pong from you: 2.13 seconds
[22:50:46] Федот leaves the room
[22:51:06] fox_a35 leaves the room
[22:53:15] <badlop> <bomer> it was the first option i found from the web site <--- do you mean this page?
http://ejabberd.jabber.ru/bugzilla
[22:55:34] fox_a35 joins the room
[22:55:36] Федот joins the room
[22:56:42] <ermine> badlop: do we reject jira because of it is on java?
[22:56:52] <ermine> interesting opinion
[22:57:26] <badlop> if that isn't the reason, then what is the real reason, or reasons?
[22:57:31] Dimka joins the room
[22:57:47] <fox_a35> ermine: :-* :-*
[22:57:53] <Федот> fox_a35: ;)
[22:58:13] <ermine> real is only one: bugzilla appears before jira and nobody said that jira becomec primary bugtrack
[22:58:29] <fox_a35> Dimka: :-)
[22:58:44] <bomer> yes badlop
[22:58:45] <Dimka> fox_a35: :-)
[22:58:47] <badlop> who should say that so you accept it?
[22:59:08] <ermine> вы, если пришли сюда только лыбаться, валите отсюда, пока не забанили
[22:59:53] <Dimka> Do you speak russian?
[23:00:25] <ermine> Dimka: нет, и это конференция не клуб знакомств
[23:00:26] <badlop> or who should speak against Jira, so all we move to Bugzilla?
[23:00:50] <ermine> badlop: nobody said nothing
[23:00:52] Федот leaves the room
[23:01:44] <ermine> badlop: can you find in maillist something about it in past years?
[23:02:10] <fox_a35> You talk to the developer of a boat.
[23:02:13] <badlop> you are right, i find absolutely no reference to the expected switch
[23:02:35] <badlop> the only mention is in the process-one announcement blog
[23:02:59] <ermine> it looks if ejabberd was selled silently
[23:03:03] <badlop> nobody said that jira becomec primary bugtrack <--- who should say 'let's move to Jira!' ?
[23:03:36] <legoscia> well, it's GPL, noone can sell it so that it is no longer ours
[23:03:58] <ermine> badlop: nobody said that p-1 becomes owner of ejabberd :) or, at least, vendor
[23:04:31] <ermine> or somewhat like it and everything moved from old hosting to hew one
[23:04:34] <TobiasFar> doesn't has aleksey the copyright on ejabberd?
[23:05:11] <ermine> about aleksey, you can find his words in his interview on website
[23:05:20] <TobiasFar> see CUPS, apple hired michael sweet, which owned CUPS, and it's now owned by apple but still open source
[23:05:26] <TobiasFar> s/which/who
[23:05:34] <ermine> it's all the fact that is known
[23:05:39] <ermine> formally
[23:06:09] <Dimka> Bye
[23:06:27] Dimka leaves the room
[23:07:02] <badlop> <Kev> badlop: what is the story with that? <--- here is the story ;)
[23:07:04] <deryni> Yes, we've established that the move to process-one was not a smooth one. Are parts of the move in-and-of-themselves problematic?
[23:09:35] <ermine> dunno
[23:10:02] Dimka joins the room
[23:10:42] <badlop> ermine: do you know why p1 installed Jira for their projects, instead of bugzilla, flyspray, Trac, or any other bug tracker?
[23:11:58] <ermine> badlop: i think i cannot say something instead of p1's managers. Ask them.
[23:12:46] <badlop> of course, i was curious and i asked mremond: 'why jira? it isn't open source!'
[23:12:55] <ermine> badlop: and your question is partially wrong
[23:13:17] <ermine> it is not important what do they use
[23:13:39] <ermine> the question is that why they want to have everything only own
[23:14:00] <fox_a35> And smtp it is complex to put?
[23:14:44] <legoscia> that bug tracker seems to be integrated with their sales and billing systems
[23:14:59] <legoscia> so they can tell a client that they worked N hours fixing problems A, B and C
[23:15:06] <legoscia> (i guess)
[23:15:15] <badlop> do you mean that, if at least 1 resource (let's say svn repository) is not hosted by process-one, then you accept p1?
[23:15:24] <deryni> It doesn't really matter why they chose it either.
[23:15:51] <badlop> so, if we move svn to sourceforge, jabber.ru, or berlios.de, then you will be happy with p1?
[23:16:09] <TobiasFar> +1 for berlios :P
[23:16:13] Dimka leaves the room: скоро вернусь
[23:16:41] <ermine> legoscia: so they wanted simply suppress everything old existing?
[23:16:54] fox_a35 leaves the room
[23:17:05] <legoscia> they're too lazy to synchronize two trackers
[23:17:12] <aleksey> i periodically synchronize my local darcs repository with this one http://re.jabber.ru/~alexey/ejabberd/
[23:17:23] <legoscia> which i completely understand, since that completely loses the point of a tracker ☺
[23:17:24] Dimka joins the room
[23:17:26] <legoscia> oo, darcs!
[23:17:31] <deryni> Keeping two trackers in sync isn't easy, I don't blame them for not wanting to do that.
[23:17:52] <ermine> aleksey: you should ask me to turn off bugzilla, not?
[23:18:27] Dimka leaves the room
[23:19:00] <ermine> aleksey: well, you can close bugzilla yourself, i dont remember my account :)
[23:19:41] <legoscia> hm… could i use the bugzilla for my PEP hack?
[23:19:50] <legoscia> it has lots of state that i just keep in my head now, which is Not Good
[23:19:54] <ermine> aleksey: выключай багзиллу, чего жмешься, народ же мучается!
[23:20:46] <aleksey> buzilla is used by tkabber, and losing bugs history is no good
[23:21:24] <ermine> aleksey: you can disable new posts for ejabberd
[23:22:18] <ermine> aleksey: i think p1 does not interest in bug history
[23:22:55] <ermine> they have their own c00l bugtracker
[23:39:32] <bomer> when i compile from the source, i should end up with an ejabberd binary in the src directory, correct?
[23:50:13] <ermine> it's enough to compile anywhere
[23:53:25] <bomer> i mean, configure works fine, but when i make, it doesn't error out. when i make install, i still do not have an ejabberd binary
[23:57:57] <legoscia> yes, you just have a lot of beam files
[23:58:03] <legoscia> those are your server ☺
[23:58:29] <bomer> but do i not also need a binary for ejabberd?
[23:58:36] <legoscia> no, just for erlang
[23:58:40] <legoscia> it's called "erl"
[23:59:49] <bomer> oh
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!