Logs for jdev

Show join/part/nick changes:

[05:16:18] * MattJ left the chat.
[05:41:56] * ermine joined the chat.
[05:47:36] * Lance Stout left the chat.
[05:48:13] * Lance Stout joined the chat.
[06:07:22] * damian joined the chat.
[06:08:59] * Lance Stout left the chat.
[06:22:51] * smoku left the chat.
[06:35:57] * teo1 left the chat.
[06:35:58] * teo1 joined the chat.
[06:36:19] * teo1 left the chat.
[06:36:20] * teo1 joined the chat.
[06:46:25] * mlundblad_laptop joined the chat.
[06:46:57] * damian left the chat.
[07:10:48] * nabatt joined the chat.
[07:58:20] * smoku joined the chat.
[08:06:43] * Alex joined the chat.
[08:08:41] <dwd> Kanchil, There is a [xep 277].
[08:08:42] <Kanchil> dwd: Sorry, I don't think there is a XEP-0277
[08:08:54] <dwd> Kanchil, Yes there is. We're up to [xep 288] now.
[08:08:54] <Kanchil> dwd: Sorry, I don't think there is a XEP-0288
[08:09:02] <dwd> Kanchil, Well, you're wrong.
[08:09:11] <dwd> Kanchil, Nothing to say to that, now, have you!
[08:13:40] * Kev joined the chat.
[08:13:45] * Tobias joined the chat.
[08:15:03] <johnny> lol
[08:15:14] <johnny> where is that dude HAL
[08:15:23] <johnny> his memory seems longer
[08:15:26] <Kev> HAL reads the same list that Kanchil is, I think.
[08:15:41] <Kev> s/is/does/
[08:16:15] <johnny> hmmm, pretty sure he knew about apis past 277
[08:16:19] <johnny> err xeps*
[08:16:48] <johnny> well in any case, guess the list should be updated
[08:16:51] * Xificurk left the chat.
[08:19:25] * Hermitifier joined the chat.
[08:21:09] * Guus joined the chat.
[08:37:31] * luca tagliaferri joined the chat.
[09:00:29] * petermount joined the chat.
[09:46:17] * Tobias left the chat.
[09:49:21] * julm joined the chat.
[09:49:31] * julm left the chat.
[09:59:25] * Alex left the chat.
[10:00:48] * mazzachre joined the chat.
[10:00:54] <mazzachre> Hi, anyone alive?
[10:01:09] <petermount> possibly
[10:01:55] <mazzachre> undead is good enough :)
[10:01:59] * gnufs joined the chat.
[10:02:35] <mazzachre> Does XMPP support prefixes and prefixed namespaced attributes? I don't see it in any of the specs or XEPs (though I have not looked through everything)
[10:03:25] <petermount> hmmm, good question... I remember seeing something about that a few months ago...
[10:07:53] <mazzachre> It is a good question... We are implementing a library for XMPP and this is a complex part of it if we are to support prefixes and namespaced attributes...
[10:10:37] <petermount> tbh the only time I've seen this recently has been with xml:id attribute and even then I'm not sure i've seen that within an xmpp stanza
[10:11:19] * gnufs left the chat.
[10:12:07] * Hermitifier left the chat.
[10:13:27] <mazzachre> Ya, I also think it is an advanced part of XML that XMPP strive not to use...
[10:14:19] <mazzachre> It will both make the parsing more complex and make the system more fraigle I think.. And it is not really needed since most stanzas are ment to be simple and not complex documents...
[10:15:08] <mazzachre> I was thinking about making a json implementation of the protocol and a converter thingie... but more as an intelectual trial more than an actual system...
[10:19:17] * gnufs joined the chat.
[10:22:12] * dbanes joined the chat.
[10:23:02] <dbanes> Good evening, morning, and afternoon!
[10:23:21] <petermount> I think that json would work for the basic parts of XMPP, but you'd have to think of how to handle an inbound stanza that it wouldn't recognise
[10:23:39] <dbanes> we're looking for a Windows Mobile Jabber client - any recommendations, esp open source w/TLS
[10:23:40] <petermount> i.e. if the a child element of say message was in a namespace you don't support
[10:24:38] <petermount> dbanes: sorry I don't touch windows these days :-) Saying that I've not seen many OS mobiles clients out there...
[10:24:51] <petermount> the one I use on the blackberry is free but not os :(
[10:26:26] <dbanes> Actually, free, commercial, no matter as we have a client, but ability to whte label would be good - asking here as I'm assuming there are some vendor people here :)
[10:27:23] <petermount> there are, but it's probably extreme early morning for most of them
[10:28:05] <dbanes> you mean b4 10am :-)
[10:28:38] <petermount> hehe probably, a lot of timezones in here
[10:38:27] * gnufs left the chat.
[10:50:54] * alfeberlin joined the chat.
[10:51:27] * alfeberlin left the chat.
[10:56:48] * smoku left the chat.
[10:56:52] * smoku joined the chat.
[11:07:33] * Guus left the chat.
[11:13:57] * Florob joined the chat.
[11:19:03] * Neustradamus left the chat.
[11:19:25] * smoku left the chat.
[11:19:27] * smoku joined the chat.
[11:24:18] * alfeberlin joined the chat.
[11:24:23] * alfeberlin left the chat.
[11:31:33] * Guus joined the chat.
[11:31:33] * mlundblad_laptop left the chat.
[11:44:38] * smoku left the chat.
[11:44:39] * smoku joined the chat.
[11:55:07] <mazzachre> As I said earlier :) My json thoughts was only ment as an intelectual excersise... Since json is not very extendible by design, I don't think it would fit very well here...
[11:55:50] <mazzachre> But is the answer to my question that our new library should not bother with prefixes and namespaced attributes? Or must we implement it to be compatible?
[11:56:09] * dbanes left the chat.
[11:56:20] <Kev> Prefixes are needed within the stanza - you're not allowed to define them within the stream header.
[11:56:27] <Kev> Namespaced attributes are allowed.
[11:56:33] <Kev> [from memory]
[11:58:19] <Kev> But it's a bad idea to /send/ either because you'll find software that doesn't cope with it.
[11:58:40] <Kev> [opinion]
[11:59:07] <mazzachre> So I would need to have <xxx:something yyy:xmlns="oh:noes" what="now" yyy:this="is bad"/> ?
[11:59:49] <mazzachre> But everything is not supporting it,. so basically it is not used?
[12:00:27] <Kev> Well, that's not allowed because you've not defined xxx
[12:00:39] <Kev> Some things use it, but not many.
[12:01:51] <Florob> I've seen namespaced attributes in the roster pushes of a social network. Clients didn't seem to mind
[12:03:59] <mazzachre> I know xxx is not defined... something like <iq xxx:xml="pron"><xxx:something ....></iq>
[12:04:52] <mazzachre> I am not certain what would happen inour library if prefixed stuff gets in there...
[12:05:32] <mazzachre> I guess it would be ignored because we don't look for it.. and no one then understand xxx:xmlns as a namespace attribute but as a strange attribute with a : in it's name...
[12:07:21] <Florob> mazzachre, you need to support this. e.g. <iq foo:xmlns='urn:xmpp:ping'><foo:ping/></iq> is valid and should be handled.
[12:07:30] * nabatt left the chat.
[12:07:45] * nabatt joined the chat.
[12:07:53] <Florob> I agree with Kev that some implementations would fail though
[12:08:35] <mazzachre> Hmmm... it adds yet another layer to our XML handling... that would normally be <iq><ping xmlns='urn:xmpp:ping'/></iq> right?
[12:08:47] <mazzachre> Like it have been described in docs now...
[12:09:04] <Kev> Yes.
[12:09:43] <smoku> if you want to be correct you have to support namespace prefixes
[12:09:51] <smoku> there is a rise in usage lately
[12:10:10] <mazzachre> I wonder if it is worth adding the extra complexity for being correct.. or if it is just being correct for being correct and not being used...
[12:10:14] * teo1 left the chat.
[12:10:33] * teo1 joined the chat.
[12:10:43] <smoku> I recently needed to fix some nasty namespace prefixes handling bugs in jabberd2 - incorrect xml was tearing a lot of S2S streams
[12:11:40] <Florob> mazzachre, what answer do you expect. You need to be namespace aware. The core spec uses namespaces already (e.g. stream:error). You certainly won't get any sympathy for being lazy
[12:11:43] <smoku> parse->serialise = crash
[12:12:00] <petermount> mazzachre: I tend to use <iq><foo:ping foo:xmlns='urn:xmpp:ping'><foo:ping/></iq> when generating xml as that matches the examples and in most cases matches what the code is doing - mainly keeps the number of namespaces down.
[12:12:14] <petermount> I've also come across the odd client that doesnt like it any other way
[12:12:41] <petermount> but for parsing I'd accept that or the namespace decl for foo in the iq element - but then that keeps to the xml spec then
[12:14:19] <smoku> on parsing you may just skip the prefixes... although I've seen some standard iq-version etc. unnecesary prefixed when coming from GTalk servers
[12:14:20] <mazzachre> I am namespace aware already, that is ofcause needed... I just don't handle prefixes currently...
[12:14:44] <smoku> IE my GG transport is not prefixes aware and reports these as unknown requests
[12:14:59] <Florob> mazzachre, what do you do with <stream:stream> ?
[12:15:42] <mazzachre> Currently handle it as an element named stream:stream...
[12:18:49] <Florob> *sigh* That is even covered by the RFC if you return <bad-namespace-prefix/> for <foo:stream foo:xmlns='etherxish'/>, but I'm not exactly fond of implementations that choose not to support namespace-prefixes
[12:19:26] * Florob left the chat.
[12:20:00] * Florob joined the chat.
[12:22:51] * Lance Stout joined the chat.
[12:25:40] <mazzachre> So... supporting prefixes... That is not hard to implement in parsing... Since we don't need them as such (we just have to keep track of them while building the stanza DOM and set the correct namespace in designated objects...) Whill we also need support for prefixed namespaced attributes? like <iq xmlns:test="test" test:to="test to" to="jid"/> ?
[12:27:00] * gnufs joined the chat.
[12:27:47] <gnufs> are data forms (jabber:x:data) implemented by im clients like pidgin and gajim?
[12:31:19] <Kev> Yes, you can't get very far without them.
[12:32:23] <louiz’> are you sure for pidgin?
[12:32:50] <gnufs> /me was trying to send data forms to gajim and pidgin, but couldn't succeed in any.
[12:32:59] <gnufs> i guess i gotta work on my data foo :)
[12:33:15] <Florob> pidgin does. The correct question is "where". In AdHoc commands: Definitely, during registration: maybe
[12:33:29] <Florob> in "random" messages: Probably not
[12:33:34] <gnufs> ah
[12:33:45] <gnufs> i was in the random message area :)
[12:34:15] * damianmef joined the chat.
[12:34:31] <louiz’> what does "random" message mean?
[12:35:03] <gnufs> i was basically trying to manually send a data form inside a <message>
[12:35:07] * damianmef left the chat.
[12:35:18] <louiz’> oh, ok
[12:36:26] <gnufs> here's a side question: can one use xmpp:x:data instead of jabber:x:data for non-ancient clients?
[12:38:30] <louiz’> it's jabber:x:data in the XEP. Why would you use something else?
[12:41:05] <gnufs> just curious
[12:41:11] <gnufs> /me is a curious boy
[12:42:13] <Florob> It's a namespace. They are mostly transparent strings, that need to match exactly. Therefore only jabber:x:data is valid
[12:42:25] <Kev> louiz’: Yes, I'm sure for Pidgin, you can't support MUC or Ad-Hoc without it.
[12:42:32] * Lance Stout left the chat.
[12:43:07] <louiz’> ok, didn't know that
[12:43:08] <Kev> gnufs: Right, clients that'll present an arbitrary unsolicited form are few.
[12:44:11] <louiz’> http://books.google.com/books?id=SG3jayrd41cC&lpg=PA163&ots=rzQef8Ku5u&dq=xmpp%3Ax%3Adata&pg=PP1#v=onepage&q&f=false why is this book *entirely* available on google books? (it's strange since google books calls that a "limited version")
[12:44:20] <louiz’> Is this book under a free license ?
[12:44:26] <Kev> It's not.
[12:44:42] <gnufs> hah, i was just reading that book's dead tree version
[12:45:00] <gnufs> in fact, i asked the data form question after readin chapter 6
[12:45:15] <Kev> louiz’: I don't see all the pages when I go to that URL.
[12:45:28] <Kev> Pages 7 and 8 are missing, for example.
[12:45:50] <louiz’> ah
[12:45:59] <gnufs> Kev: hey, you're one of the authors of that book?
[12:46:08] <Kev> gnufs: Yes.
[12:46:24] <gnufs> awesome book
[12:46:31] <Kev> Thanks.
[12:46:45] <louiz’> Kev, page 6 and 7 of the Table of Contents?
[12:47:13] <louiz’> Oh, no, ok, you're right, it starts from the chapter 2
[12:47:16] <Kev> louiz’: No idea, but scrolling through it I get quite a few missing pages. Admittedly far fewer than I'd expect.
[12:47:25] <louiz’> but a lot of pages are available
[12:47:35] <gnufs> Kev: especially liked the disco chapter so far
[12:47:54] <louiz’> I intend to buy this book too
[12:48:08] <Kev> gnufs: Excellent.
[12:48:49] <Kev> louiz’: Also excellent :)
[12:49:27] <gnufs> I could see a hundred ways how an author could turn that topic into an incomprehensible spaghetti of info
[12:50:05] <louiz’> Kev, do you (and the other authors) get any money on the books sells?
[12:50:20] <Kev> louiz’: Yes. Although not a whole load.
[12:50:42] <louiz’> Ok
[12:51:47] <gnufs> /me is combining "xmpp the definitive guide" with "professional xmpp programming with javascript and jquery"
[12:52:06] <Kev> gnufs: A few people have commented that reading them together is good.
[12:53:04] <louiz’> and what about this one http://oreilly.com/catalog/9780596002022/ ?
[12:53:30] <Kev> That's going to be largely irrelevant now, I imagine.
[12:53:37] <Kev> XMPP wasn't standardised until 2004.
[12:53:55] <louiz’> Yeah, I just saw that it's a pretty old book
[12:53:56] <Kev> The two that gnufs mentions are the only books I know that're currently relevant.
[12:54:11] <louiz’> Ok, I'll delete this one from my list :)
[12:56:35] <gnufs> oh, and i first got interested in learning xmpp by randomly stumbling upon http://el-tramo.be/documents/beautiful-xmpp-testing/BeautifulXMPPTesting.pdf
[12:57:06] * Xificurk joined the chat.
[13:00:05] * Lance Stout joined the chat.
[13:04:59] <Florob> Who is moderator for the standards@ list?
[13:06:33] <Kev> Peter's moderator for all the lists.
[13:06:58] <Kev> Have you got a post stuck? I Have The Power if you need me to use it.
[13:08:04] <louiz’> gnufs, thank you, this PDF seems really great. I was wondering how I could test automatically my client
[13:08:06] * waqas joined the chat.
[13:08:27] <Kev> Remko being one of the other TDG authors (and the other Swift dev).
[13:08:36] <Florob> Kev, I have one stuck yes. If you could use some of that power it'd be nice.
[13:10:41] * gnufs left the chat.
[13:11:16] * gnufs joined the chat.
[13:16:25] <petermount> as i've just got back from lunch & going back a bit to the xml namespace topic: it's a common mistake to handle xml namespaces as part of the element, i.e. <stream:stream xmlns:stream='http://etherx.jabber.org/streams'>/</stream:stream> is the same as <ns1:stream xmlns:ns1='http://etherx.jabber.org/streams'>/</ns1:stream>
[13:16:42] <petermount> the number of times I've seen parsers fail because they don't handle that correctly :(
[13:18:46] * gnufs left the chat.
[13:19:21] <mazzachre> petermount: Uhm... I don't follow you... Those 2 examples are the same, right? Or am I missing something here?
[13:19:40] * gnufs joined the chat.
[13:23:53] <Kev> Florob: You really needed to send a 500k message to the list?
[13:24:48] <Kev> I'd rather not unblock that, it seems rather antisocial.
[13:26:40] * gnufs left the chat.
[13:26:53] <Kev> Ah, I see what it is.
[13:26:54] <petermount> they are the same, but some parsers look at the namespace in the element not the xmlns, so would see that as <stream:stream/> and <ns1:stream> and think it was two different elements...
[13:26:57] <Florob> Kev, I can put it all on my web-space if need be. XEPs are definitely to large for this world
[13:27:07] <Kev> I think that'd be friendlier :)
[13:27:19] <Kev> Or fork it on gitorious and upload the commit there :)
[13:27:23] * gnufs joined the chat.
[13:27:36] <petermount> mazzachre: I've seen it with a commercial soap service before :-(
[13:28:30] <mazzachre> Oh...
[13:28:48] <Florob> Kev, well, the diff is hardly the problem. The html is and I don't trust anyone to be able to build that
[13:29:06] <Kev> Fair enough, I didn't download it to see which bits were big :)
[13:29:42] * scippio_netbook joined the chat.
[13:30:23] * gnufs left the chat.
[13:31:07] <mazzachre> So... prefixes are to be supported... But prefixed/namespaced attributes? Are they needed?
[13:31:07] * gnufs joined the chat.
[13:31:39] * MattJ joined the chat.
[13:33:02] * nabatt left the chat.
[13:33:38] * gnufs left the chat.
[13:33:59] * gnufs joined the chat.
[13:34:05] <waqas> Ah, the joys of doing your own prefix processing...
[13:34:41] <waqas> mazzachre: What do you mean by needed? It's required by XML NAMES. You can choose to be compliant or not :)
[13:34:56] <Florob> mazzachre, I'd recommend supporting them. OSW is going in that direction IIRC. But they are indeed not widely used.
[13:35:49] <mazzachre> Well.. needed as in, will the universe start burning if I don't? :) Well... it will change my implementation from very simple to somewhat more complicated then...
[13:36:56] <waqas> Your implementation of what? You have a custom XML parser?
[13:37:00] * gnufs left the chat.
[13:37:25] * luca tagliaferri left the chat.
[13:37:26] <mazzachre> something like that ya...
[13:37:51] * gnufs joined the chat.
[13:39:06] <petermount> mazzachre if your parser is generating a DOM model then its easy as you can delegate that to whatever is parsing the DOM - I do that with one of my own parsers as I then leave the namespace handling to JAXB
[13:39:48] <MattJ> Sounds like I dropped in on an XML namespaces discussion, my favourite
[13:40:07] <petermount> mattj: yep it can be a good one ;-)
[13:40:13] <MattJ> My belief is that no-one should ever write their own XML parser, unless they're waqas
[13:40:55] <mazzachre> We have own DOM model also :) And up until now we have used a simple Map structure for attributes (name, value) but with namespaces that becomes a triplet and the map does not quite support that...
[13:40:56] <waqas> mazzachre: You can always stay prefix unaware, and use hardcoded 'stream:stream', etc. You wont be compliant with XMPP, but then many servers in popular use aren't :)
[13:41:07] <petermount> mattj: I've done a non-validating one before but as I just said it generates standard W3 dom - so most issues you pass on to the next framework ;-)
[13:41:50] <mazzachre> waqas: I know... It is not hard to do prefixes... I looked through our code and it seems we are almost there anyway, so we will do that... My problem is namespaced attributes now :)
[13:42:10] <petermount> mazzachre: take a look at the w3 spec, your names should really be QNames, which would then handle both namespaces and localnames
[13:42:38] <waqas> In Prosody we convert <x:y x:z="" xmlns:x="blah"> to <blah|y blah|z="">, and change back when sending.
[13:43:19] <waqas> We use \1 (ascii code 1) instead of |. Works perfectly, without most of the code needing to know about namespaces.
[13:43:22] <mazzachre> Ya... but we have specifically tried to stay as simple as possible in what is needed to support xmpp...
[13:44:10] <waqas> mazzachre: You'll only encounter prefixed attributes in custom payloads. We just pass them through, but none of the defined XEPs (AFAIK) define prefixed attributes.
[13:44:12] <petermount> keeping things simple is good, but sometimes you can accidentally go too far...
[13:45:07] * Neustradamus joined the chat.
[13:47:18] * Zash joined the chat.
[13:47:33] <mazzachre> I know that there is something known as "too simple" :) Which is also why I asked here... The old library we used were... well... hardly.. namespace aware... It started out without namespaces and then they were taped on with the equivalent of yellow stickers... Not a good solution... So now we are refactoring... Also taking out some of the stuff that became very complicated over time...
[13:50:03] <petermount> best place to ask ;-) A good place to start would be http://www.w3.org/TR/DOM-Level-2-Core/ and take a look at how Element and Attr are laid out, you might find some ideas from there
[13:51:25] * luca tagliaferri joined the chat.
[13:52:21] * Florob left the chat.
[13:53:30] <mazzachre> The w3 specification is just so huge... we don't want to support everything :) But figuring out which parts to leave out is not quite easy...
[13:55:15] * gnufs left the chat.
[13:55:31] * gnufs joined the chat.
[13:55:32] <petermount> Yes, mainly Element and Attr (both extend Node)... The Java api might be easier on the eyes, but it follows that spec http://retep.net/9AqtQd
[13:55:33] <mazzachre> Also, I don't belive we have to support mixed content elements?
[13:56:12] * alfeberlin joined the chat.
[13:56:27] <petermount> you shouldn't need to, just giving you a few pointers ;-)
[13:56:33] * alfeberlin left the chat.
[13:58:17] <mazzachre> Ya... I apreciate it...
[13:58:46] <mazzachre> Hmmm... I could hack something with the namespaced attributes.. or I could simply ignore them...
[13:59:20] * Florob joined the chat.
[13:59:53] <mazzachre> Or I could implment a QName structure and use that as naming for "everything" (Elements and Attributes)
[14:00:35] <petermount> yes thats what I would do myself...
[14:00:39] * gnufs left the chat.
[14:00:54] * gnufs joined the chat.
[14:01:01] * scippio_netbook left the chat.
[14:02:40] <waqas> Or you could just do what we do, which should work without changing your current system much :)
[14:03:54] * gnufs left the chat.
[14:04:35] * gnufs joined the chat.
[14:05:11] <mazzachre> Then... searching and hashing... especially serching without namespace would become slow... linear... Unless we done something clever with the hashing...
[14:05:30] <waqas> Erm, why?
[14:06:01] * lancestout joined the chat.
[14:06:01] * lancestout left the chat.
[14:06:29] <waqas> We use a hashtable to store attributes. All attribute names are strings, including the prefixed ones. The search time is constant.
[14:06:30] <mazzachre> Well... getting anything by name, without namespace would be difficult if not itterating through the entire list and comparing parts all the time...
[14:07:13] <waqas> Why would it be difficult? Without namespaces is how we do it all the time. It's just a hash access.
[14:07:25] <MattJ> No
[14:07:34] <MattJ> More importantly, why would you want to get an attribute by name?
[14:07:49] <mazzachre> for ease of use...
[14:08:24] <MattJ> For ease of use your API automatically uses the namespace of the element you're looking on
[14:09:04] <MattJ> Looking up by just name is wrong, that's the whole point of namespaces
[14:09:10] <mazzachre> We get children by name also... when we don't care about the namespace (or rather.. asume that the namespace is correct because we are fairly certain what is what in most XMPP stanzas...
[14:09:20] <MattJ> sigh
[14:09:33] <mazzachre> I know.. but it makes for much faster access :)
[14:09:56] <mazzachre> I know it is wrong.. Like crossing the road for red lights at night... but it beats waiting :)
[14:10:04] <waqas> /me applies his namespace tazer to MattJ. Unprefixed attribute namespace == "", not the element's namespace.
[14:10:14] <MattJ> waqas, whatever
[14:10:32] <waqas> But I still don't see how namespaces add any complexity this way :)
[14:10:42] <petermount> I can see something like 500 messages/s and don't see any slowdowns with the lookups we do
[14:10:53] * bjc joined the chat.
[14:10:58] * lancestout joined the chat.
[14:10:58] * lancestout left the chat.
[14:11:00] <petermount> mattj: agreed, thats what I mentioned about 45mins ago ;-)
[14:12:12] <mazzachre> Well... we do this (ugly... cover your eyes if you are easily frightened) <something><child xmlns="child:xmlns"/></something>... On the something element .getChildren("child"); which returns any element named "child" whatever namespace it is in...
[14:15:12] <MattJ> Sure, I know what you do, you don't do anything regarding namespaces :)
[14:15:58] * lancestout joined the chat.
[14:15:58] * lancestout left the chat.
[14:16:35] <waqas> Which is only to be expected :)
[14:16:59] * Lance Stout left the chat.
[14:17:36] <mazzachre> I do... I save them... and I get also get children by name and namespace which just filters the list more...
[14:20:59] * lancestout joined the chat.
[14:20:59] * lancestout in now known as Lance Stout.
[14:20:59] * Lance Stout left the chat.
[14:26:00] * lancestout joined the chat.
[14:26:00] * lancestout in now known as Lance Stout.
[14:26:00] * Lance Stout left the chat.
[14:31:01] * lancestout joined the chat.
[14:31:01] * lancestout in now known as Lance Stout.
[14:31:01] * Lance Stout left the chat.
[14:36:02] * lancestout joined the chat.
[14:36:02] * lancestout in now known as Lance Stout.
[14:36:02] * Lance Stout left the chat.
[14:41:02] * Lance Stout joined the chat.
[14:41:02] * Lance Stout in now known as lancestout.
[14:41:02] * lancestout left the chat.
[14:41:15] * lancestout joined the chat.
[14:41:15] * lancestout left the chat.
[14:41:15] * lancestout joined the chat.
[14:41:15] * lancestout in now known as Lance Stout.
[14:41:15] * Lance Stout left the chat.
[14:41:15] * lancestout joined the chat.
[14:41:15] * lancestout left the chat.
[14:44:51] * lancestout joined the chat.
[14:44:51] * lancestout in now known as Lance Stout.
[14:44:51] * Lance Stout left the chat.
[14:46:03] * lancestout joined the chat.
[14:46:03] * lancestout in now known as Lance Stout.
[14:46:03] * Lance Stout left the chat.
[14:51:03] * lancestout joined the chat.
[14:51:03] * lancestout in now known as Lance Stout.
[14:51:04] * Lance Stout left the chat.
[14:56:04] * lancestout joined the chat.
[14:56:04] * lancestout in now known as Lance Stout.
[14:56:04] * Lance Stout left the chat.
[14:58:35] * Guus left the chat.
[14:58:36] * stpeter joined the chat.
[15:01:05] * Lance Stout joined the chat.
[15:01:05] * Lance Stout in now known as lancestout.
[15:01:05] * lancestout left the chat.
[15:03:29] * teo1 left the chat.
[15:03:46] * smoku left the chat.
[15:03:49] * smoku joined the chat.
[15:06:05] * lancestout joined the chat.
[15:06:05] * lancestout in now known as Lance Stout.
[15:06:06] * Lance Stout left the chat.
[15:09:39] * mazzachre left the chat.
[15:11:06] * lancestout joined the chat.
[15:11:06] * lancestout in now known as Lance Stout.
[15:11:06] * Lance Stout left the chat.
[15:11:58] * smoku left the chat.
[15:11:59] * smoku joined the chat.
[15:16:06] * lancestout joined the chat.
[15:16:06] * lancestout in now known as Lance Stout.
[15:16:06] * Lance Stout left the chat.
[15:16:20] * Lance Stout joined the chat.
[15:17:31] * julm joined the chat.
[15:17:40] * julm left the chat.
[15:18:59] * lancestout joined the chat.
[15:18:59] * lancestout left the chat.
[15:21:07] * lancestout joined the chat.
[15:21:07] * lancestout left the chat.
[15:26:08] * lancestout joined the chat.
[15:26:08] * lancestout left the chat.
[15:31:08] * lancestout joined the chat.
[15:31:09] * lancestout left the chat.
[15:36:09] * lancestout joined the chat.
[15:36:09] * lancestout left the chat.
[15:36:29] * hawke joined the chat.
[15:41:09] * lancestout joined the chat.
[15:41:09] * lancestout left the chat.
[15:44:39] * Lance Stout left the chat.
[15:46:10] * Lance Stout joined the chat.
[15:46:10] * Lance Stout in now known as lancestout.
[15:46:10] * lancestout left the chat.
[15:48:19] * Tobias joined the chat.
[15:49:44] * deryni left the chat.
[15:51:11] * lancestout joined the chat.
[15:51:11] * lancestout in now known as Lance Stout.
[15:51:11] * Lance Stout left the chat.
[15:53:28] * waqas left the chat.
[15:53:44] * teo1 joined the chat.
[15:56:08] * luca tagliaferri left the chat.
[15:56:12] * lancestout joined the chat.
[15:56:12] * lancestout in now known as Lance Stout.
[15:56:12] * Lance Stout left the chat.
[15:59:17] * Tobias left the chat.
[16:01:05] * petermount left the chat.
[16:01:12] * lancestout joined the chat.
[16:01:12] * lancestout in now known as Lance Stout.
[16:01:12] * Lance Stout left the chat.
[16:01:22] * petermount joined the chat.
[16:04:29] * Tobias joined the chat.
[16:05:46] * deryni joined the chat.
[16:06:08] * jkhii joined the chat.
[16:06:12] * lancestout joined the chat.
[16:06:12] * lancestout in now known as Lance Stout.
[16:06:13] * Lance Stout left the chat.
[16:11:13] * Tobias left the chat.
[16:11:13] * lancestout joined the chat.
[16:11:13] * lancestout in now known as Lance Stout.
[16:11:13] * Lance Stout left the chat.
[16:11:13] * Tobias joined the chat.
[16:16:14] * lancestout joined the chat.
[16:16:14] * lancestout in now known as Lance Stout.
[16:16:14] * Lance Stout left the chat.
[16:17:41] * lancestout joined the chat.
[16:17:41] * lancestout in now known as Lance Stout.
[16:17:41] * Lance Stout left the chat.
[16:21:03] * Florob left the chat.
[16:21:05] * Florob joined the chat.
[16:21:15] * lancestout joined the chat.
[16:21:15] * lancestout in now known as Lance Stout.
[16:21:15] * Lance Stout left the chat.
[16:26:15] * lancestout joined the chat.
[16:26:15] * lancestout in now known as Lance Stout.
[16:26:15] * Lance Stout left the chat.
[16:31:16] * lancestout joined the chat.
[16:31:16] * lancestout in now known as Lance Stout.
[16:31:16] * Lance Stout left the chat.
[16:31:43] * jkhii left the chat.
[16:36:17] * Lance Stout joined the chat.
[16:36:17] * Lance Stout in now known as lancestout.
[16:36:17] * lancestout left the chat.
[16:41:17] * Lance Stout joined the chat.
[16:41:17] * Lance Stout in now known as lancestout.
[16:41:17] * lancestout left the chat.
[16:43:04] * damian joined the chat.
[16:46:18] * lancestout joined the chat.
[16:46:18] * lancestout in now known as Lance Stout.
[16:46:18] * Lance Stout left the chat.
[16:46:48] * damian left the chat.
[16:51:18] * lancestout joined the chat.
[16:51:18] * lancestout in now known as Lance Stout.
[16:51:18] * Lance Stout left the chat.
[16:53:20] * damian joined the chat.
[16:56:19] * Lance Stout joined the chat.
[16:56:19] * Lance Stout in now known as lancestout.
[16:56:19] * lancestout left the chat.
[16:56:49] * damian left the chat.
[17:00:35] * damian joined the chat.
[17:01:20] * lancestout joined the chat.
[17:01:20] * lancestout in now known as Lance Stout.
[17:01:20] * Lance Stout left the chat.
[17:04:20] * damian left the chat.
[17:04:59] * damian joined the chat.
[17:06:08] * smoku left the chat.
[17:06:20] * lancestout joined the chat.
[17:06:20] * lancestout in now known as Lance Stout.
[17:06:20] * Lance Stout left the chat.
[17:06:38] * damian left the chat.
[17:07:31] * damian joined the chat.
[17:11:21] * Lance Stout joined the chat.
[17:11:21] * Lance Stout in now known as lancestout.
[17:11:21] * lancestout left the chat.
[17:16:21] * lancestout joined the chat.
[17:16:21] * lancestout in now known as Lance Stout.
[17:16:21] * Lance Stout left the chat.
[17:19:54] * damian left the chat.
[17:21:22] * lancestout joined the chat.
[17:21:22] * lancestout in now known as Lance Stout.
[17:21:22] * Lance Stout left the chat.
[17:26:22] * lancestout joined the chat.
[17:26:22] * lancestout in now known as Lance Stout.
[17:26:22] * Lance Stout left the chat.
[17:28:17] * Tobias left the chat.
[17:31:23] * Lance Stout joined the chat.
[17:31:23] * Lance Stout in now known as lancestout.
[17:31:23] * lancestout left the chat.
[17:32:56] * damian joined the chat.
[17:34:41] * damian left the chat.
[17:35:38] * damian joined the chat.
[17:35:40] * damian left the chat.
[17:35:56] * evilotto joined the chat.
[17:36:24] * lancestout joined the chat.
[17:36:24] * lancestout in now known as Lance Stout.
[17:36:24] * Lance Stout left the chat.
[17:37:46] * petermount left the chat.
[17:41:24] * lancestout joined the chat.
[17:41:24] * lancestout in now known as Lance Stout.
[17:41:24] * Lance Stout left the chat.
[17:44:56] * asdasda joined the chat.
[17:45:01] * asdasda left the chat.
[17:46:25] * lancestout joined the chat.
[17:46:25] * lancestout in now known as Lance Stout.
[17:46:25] * Lance Stout left the chat.
[17:47:31] * mef joined the chat.
[17:51:25] * lancestout joined the chat.
[17:51:25] * lancestout in now known as Lance Stout.
[17:51:25] * Lance Stout left the chat.
[17:55:31] * mef left the chat.
[17:56:25] * lancestout joined the chat.
[17:56:25] * lancestout in now known as Lance Stout.
[17:56:26] * Lance Stout left the chat.
[18:01:26] * lancestout joined the chat.
[18:01:26] * lancestout in now known as Lance Stout.
[18:01:26] * Lance Stout left the chat.
[18:06:27] * lancestout joined the chat.
[18:06:27] * lancestout in now known as Lance Stout.
[18:06:27] * Lance Stout left the chat.
[18:10:41] <Zash> Kev: +1 on MattJ in on the muc status 100 thing
[18:11:01] <Kev> Zash: Yes, that's my opinion too.
[18:11:27] * lancestout joined the chat.
[18:11:27] * lancestout in now known as Lance Stout.
[18:11:27] * Lance Stout left the chat.
[18:16:28] * Lance Stout joined the chat.
[18:16:28] * Lance Stout in now known as lancestout.
[18:16:28] * lancestout left the chat.
[18:20:30] * mlundblad joined the chat.
[18:21:29] * Lance Stout joined the chat.
[18:21:29] * Lance Stout in now known as lancestout.
[18:21:29] * lancestout left the chat.
[18:24:13] * Tobias joined the chat.
[18:26:29] * lancestout joined the chat.
[18:26:29] * lancestout in now known as Lance Stout.
[18:26:29] * Lance Stout left the chat.
[18:31:30] * lancestout joined the chat.
[18:31:30] * lancestout in now known as Lance Stout.
[18:31:30] * Lance Stout left the chat.
[18:31:58] <louiz’> and mine too
[18:36:30] * lancestout joined the chat.
[18:36:30] * lancestout in now known as Lance Stout.
[18:36:31] * Lance Stout left the chat.
[18:37:19] * hawke left the chat.
[18:37:57] * hawke joined the chat.
[18:41:31] * lancestout joined the chat.
[18:41:31] * lancestout in now known as Lance Stout.
[18:41:31] * Lance Stout left the chat.
[18:43:22] * SJrX joined the chat.
[18:43:59] * SJrX left the chat.
[18:46:32] * lancestout joined the chat.
[18:46:32] * lancestout in now known as Lance Stout.
[18:46:32] * Lance Stout left the chat.
[18:51:32] * lancestout joined the chat.
[18:51:32] * lancestout in now known as Lance Stout.
[18:51:33] * Lance Stout left the chat.
[18:56:33] * lancestout joined the chat.
[18:56:33] * lancestout in now known as Lance Stout.
[18:56:33] * Lance Stout left the chat.
[19:01:20] * evilotto left the chat.
[19:01:34] * lancestout joined the chat.
[19:01:34] * lancestout in now known as Lance Stout.
[19:01:34] * Lance Stout left the chat.
[19:06:32] * justin joined the chat.
[19:06:35] * lancestout joined the chat.
[19:06:35] * lancestout in now known as Lance Stout.
[19:06:35] * Lance Stout left the chat.
[19:11:35] * lancestout joined the chat.
[19:11:35] * lancestout in now known as Lance Stout.
[19:11:35] * Lance Stout left the chat.
[19:16:19] * evilotto joined the chat.
[19:16:36] * Lance Stout joined the chat.
[19:16:36] * Lance Stout in now known as lancestout.
[19:16:36] * lancestout left the chat.
[19:17:06] * evilotto left the chat.
[19:17:25] * evilotto joined the chat.
[19:18:03] * waqas joined the chat.
[19:18:06] * evilotto left the chat.
[19:18:39] * evilotto joined the chat.
[19:21:36] * Lance Stout joined the chat.
[19:21:36] * Lance Stout in now known as lancestout.
[19:21:36] * lancestout left the chat.
[19:25:27] <MattJ> Zash, I thought you figured out how to post to mailing lists? :)
[19:25:46] <MattJ> Come on, it took me long enough to convince waqas - and now he possibly posts more than me
[19:26:04] <Tobias> ahh, apple is looking for Sr. iChat Software Engineer with strong knowledge of XMPP
[19:26:25] <Tobias> maybe this finally means the'll update their iChat to the new millenium
[19:26:37] * lancestout joined the chat.
[19:26:37] * lancestout in now known as Lance Stout.
[19:26:37] * Lance Stout left the chat.
[19:26:53] <Zash> MattJ: +1 seemed small to justify an entire post
[19:27:56] <MattJ> Tobias, I thought they had darco?
[19:29:08] <Tobias> MattJ: they do?
[19:29:42] <waqas> Tobias: Are you going to apply? :)
[19:30:28] <Tobias> waqas: nah, i don't have time for two church visits on sundays
[19:31:18] * smoku joined the chat.
[19:31:38] * lancestout joined the chat.
[19:31:38] * lancestout in now known as Lance Stout.
[19:31:38] * Lance Stout left the chat.
[19:36:38] * lancestout joined the chat.
[19:36:38] * lancestout in now known as Lance Stout.
[19:36:39] * Lance Stout left the chat.
[19:36:40] <stpeter> /me laughs
[19:37:25] <Zash> So did I!
[19:37:29] <Zash> Probably not about the same thing
[19:38:27] <stpeter> I laughed about "<Tobias> waqas: nah, i don't have time for two church visits on sundays"
[19:38:48] <Zash> /me watched american dad
[19:41:39] * lancestout joined the chat.
[19:41:39] * lancestout in now known as Lance Stout.
[19:41:39] * Lance Stout left the chat.
[19:45:17] <louiz’> I didn't get it :(
[19:46:40] * lancestout joined the chat.
[19:46:40] * lancestout in now known as Lance Stout.
[19:46:40] * Lance Stout left the chat.
[19:51:41] * Lance Stout joined the chat.
[19:51:41] * Lance Stout in now known as lancestout.
[19:51:41] * lancestout left the chat.
[19:55:26] * bjc left the chat.
[19:56:41] * lancestout joined the chat.
[19:56:41] * lancestout in now known as Lance Stout.
[19:56:41] * Lance Stout left the chat.
[19:59:12] <stpeter> louiz’: the joke is that Apple is a religion
[19:59:38] <louiz’> ah, yeah, ok. Thanks :)
[20:01:42] * lancestout joined the chat.
[20:01:42] * lancestout in now known as Lance Stout.
[20:01:42] * Lance Stout left the chat.
[20:02:12] <stpeter> /me posts updated versions of the bis drafts
[20:06:43] * Lance Stout joined the chat.
[20:06:43] * Lance Stout in now known as lancestout.
[20:06:43] * lancestout left the chat.
[20:11:44] * lancestout joined the chat.
[20:11:44] * lancestout in now known as Lance Stout.
[20:11:44] * Lance Stout left the chat.
[20:14:56] * ermine left the chat.
[20:15:03] <justin> some years back i spoke with an apple recruiter, and the person said i would have to cease working on psi if i were to take a job there
[20:15:20] <justin> i'm pretty sure that wasn't true, but it left a bad taste
[20:16:10] <MattJ> Sounds about right :)
[20:16:20] <Tobias> doesn't sound nice though
[20:16:44] * lancestout joined the chat.
[20:16:44] * lancestout in now known as Lance Stout.
[20:16:44] * Lance Stout left the chat.
[20:17:10] <waqas> Does anyone know of any Apple developers with active personal open source projects?
[20:18:12] <waqas> I see quite some stuff from Google folks, and even Microsoft. But never really saw anything from Apple devs that I recall.
[20:18:34] <justin> hmm.. i /think/ there are... somewhere. which is why i didn't fully believe the requirement
[20:18:54] <justin> but it is surely rare. apple is definitely a very closed society
[20:18:54] <deryni> I think one of the Adium devs got a job with Apple and needed to largely stop working on Adium or something like that.
[20:19:08] <Tobias> yup..all they to is put some stuff out with some free license, but that doesn't make that stuff an open source project
[20:19:52] <waqas> Tobias: Technically it does. Open source has a definition: http://en.wikipedia.org/wiki/Open_Source_Definition
[20:20:22] <Tobias> waqas: they define what open source is, not what active open source projects are :D
[20:20:44] <waqas> WebKit is fairly open and active, and accepts contributions :)
[20:21:04] <Tobias> waqas: yeah..right, that might be the only open source projects where apple folk workon
[20:21:07] <Tobias> *work on
[20:21:30] <waqas> There are also LLVM and Clang.
[20:21:45] * lancestout joined the chat.
[20:21:45] * lancestout in now known as Lance Stout.
[20:21:45] * Lance Stout left the chat.
[20:21:59] <waqas> And that I think sums up the Apple open source projects I'm interested in :)
[20:21:59] <Tobias> waqas: ahh, right..forgot about those
[20:22:45] <justin> yes, although, those are run by apple
[20:23:04] <waqas> Yep
[20:23:19] <justin> the original question is if there is anyone doing anything personal outside of work. and i don't mean a webkit guy at apple going home and working on webkit :)
[20:23:53] <stpeter> :)
[20:25:08] * hawke left the chat.
[20:25:38] * hawke joined the chat.
[20:26:45] * lancestout joined the chat.
[20:26:45] * lancestout in now known as Lance Stout.
[20:26:45] * Lance Stout left the chat.
[20:31:45] * lancestout joined the chat.
[20:31:45] * lancestout in now known as Lance Stout.
[20:31:46] * Lance Stout left the chat.
[20:36:46] * lancestout joined the chat.
[20:36:46] * lancestout in now known as Lance Stout.
[20:36:46] * Lance Stout left the chat.
[20:41:47] * Lance Stout joined the chat.
[20:41:47] * Lance Stout in now known as lancestout.
[20:41:47] * lancestout left the chat.
[20:46:48] * lancestout joined the chat.
[20:46:48] * lancestout in now known as Lance Stout.
[20:46:48] * Lance Stout left the chat.
[20:51:48] * lancestout joined the chat.
[20:51:48] * lancestout in now known as Lance Stout.
[20:51:48] * Lance Stout left the chat.
[20:59:59] * scippio_netbook joined the chat.
[21:03:34] * joaoblue joined the chat.
[21:04:07] * Treebilou left the chat.
[21:04:36] <joaoblue> alguém do BRASIL pra dar uma dica?
[21:07:12] * dwd left the chat.
[21:07:50] * dwd joined the chat.
[21:09:45] <joaoblue> I can send and receive e-mails login@jabber.org?
[21:10:15] <Zash> Wouldn't think so
[21:11:52] * dwd left the chat.
[21:13:03] * dwd joined the chat.
[21:16:35] <gnufs> maybe i should donate to xsf just for maintaining such a beautiful and sane protocol..
[21:17:06] <Tobias> gnufs: that's always an option :)
[21:17:26] <Tobias> feed the evil at its root, or how goes te saying? :)
[21:18:07] * mlundblad left the chat.
[21:20:37] * julm joined the chat.
[21:21:11] <louiz’> But in what way will money help? :o
[21:21:42] <Tobias> it'll feed all the nice XSF folk
[21:22:28] <louiz’> ah, yeah, good reason. They'd die, otherwise :(
[21:22:43] <MattJ> louiz’, http://xmpp.org/about-xmpp/xsf/xsf-financial-summary/ :)
[21:22:47] <Tobias> exactly :D
[21:22:47] <waqas> /me gives Tobias a cookie
[21:23:12] <MattJ> Hmm, that page is out of date
[21:23:31] <waqas> No revenue since 2008 ^^
[21:24:02] <Tobias> it's been out of date for months
[21:24:03] <Tobias> :)
[21:24:25] <Tobias> apparently there's more important stuff to care about that $s
[21:24:37] * joaoblue left the chat.
[21:25:30] <louiz’> in my opinion, the Marketing expense is kind of a waste :/
[21:25:40] <louiz’> $1600 for a logo, really?
[21:25:58] <gnufs> "Total Revenues 5,000.00"
[21:26:10] <gnufs> wow it's in a much smaller range than i would have guessed
[21:26:22] <gnufs> i guess individual donations would make a difference
[21:26:59] <gnufs> i guess google doesn't come in with its gazillions to every project it sponsors :D
[21:27:33] <MattJ> :)
[21:30:09] <Zash> Woah, 330 pages!
[21:30:52] <waqas> 330 pages?
[21:30:57] <louiz’> in ?
[21:31:33] <Zash> XMPP Core + IM + address
[21:32:33] * Lance Stout joined the chat.
[21:32:34] <Zash> Linked from https://stpeter.im/index.php/2010/10/06/xmpp-wg-update/
[21:32:48] <dwd> Zash, As I said on the WG list, we can't make it any bigger or people will assume it's a SIP spec.
[21:34:03] <dwd> Okay, so possibly very silly question, but does anyone know of a live XMPP server running on a revoked certificate?
[21:34:50] <Zash> dwd: I did, some day ago, when I talked about that .. somewhere
[21:35:12] <dwd> Zash, Thanks for the extraordinary amount of detail and help, there.
[21:35:31] <Zash> :D
[21:38:24] <dwd> Zash, Any idea where, or what server, or anything?
[21:39:15] <Zash> My personal server, untill I replaced the cert and restarted :P
[21:40:10] <dwd> Zash, Right. :-) I can run up a CA and issue and revoke a cert - and I have - but it'd be nice to find a commercial CA doing the same.
[21:40:46] <dwd> Zash, StartCom haven't got any current, but revoked, certs. Not that I could tell which they were even if they had.
[21:40:50] <Zash> Oh, well. CAcert might not be "commersial"
[21:41:07] <Zash> but they allow you to revoke certs yourself
[21:41:45] <dwd> Oh, CA Cert is good enough. I suppose I could get, and revoke, a cert from them.
[21:42:29] <dwd> /me has one kitten asleep on his lap, and the other stalking about the office trying to sneak up on me/it to attack.
[21:44:54] <Zash> Anyone who has the cool version of openssl with -starttls xmpp who could check what cert I have on zash.se?
[21:45:07] <dwd> Zash, Wireshark's good at that too.
[21:45:31] <dwd> As is M-Link, actually. Hang on.
[21:47:08] <dwd> That useful enough?
[21:47:51] <Zash> Hrr
[21:47:56] <gnufs> XMPP.org refuses to give me a paypal link...
[21:48:36] <gnufs> i guess i could try to manually donate to paypal@xmpp.org on paypal but still..
[21:49:19] <dwd> gnufs, Ask the Executive Director - send an email to PSA.
[21:50:26] <stpeter> hmm, it seems that http://xmpp.org/xsf/contact.shtml has a link but it's for paypal@jabber.org :/
[21:50:55] <stpeter> probably forgot to update it at one point
[21:50:56] <stpeter> brb
[21:50:58] <gnufs> dwd: it seems the executive dir is in presence :)
[21:52:27] <gnufs> stpeter: i wish the donate button on xmpp.org linked to paypal instead of http://xmpp.org/?page_id=258#paypal which doesn't even have a link
[21:55:06] <stpeter> hrm
[21:55:34] <stpeter> sounds like a problem with the new website :)
[21:55:41] <gnufs> while we're at it, http://blog.xmpp.org/wp-content/uploads/2010/02/Fabio_Forno_low.mp3 , http://blog.xmpp.org/wp-content/uploads/2010/02/Dave_Cridland_low.mp3 , http://blog.xmpp.org/wp-content/uploads/2010/02/Florian_Jensen_low.mp3 and http://blog.xmpp.org/wp-content/uploads/2010/02/Simon_Tennant_low.mp3 are broken
[21:56:02] <gnufs> basically the podcasts listed on the right side pane
[21:56:10] <dwd> gnufs, Oh, you don't want to listen to mine, it's crap.
[21:57:03] <gnufs> dwd: :) still, dead links never look good for a project
[21:58:59] <gnufs> btw, will there be an xmpp summit and related talks during fosdem 2011?
[21:59:15] <gnufs> /me may attend fosdem this time
[22:01:13] <Florob> /me certainly hopes so
[22:02:15] <dwd> gnufs, And beer.
[22:02:37] <gnufs> /me never says no to delicious belgian beer
[22:05:17] <Zash> dwd: Is this page valid? http://wiki.xmpp.org/web/XMPP_Server_Certificates
[22:06:14] <dwd> I vaguely recall doing something similar ages ago. But to be honest, I normally use Isode's CA software which is all GUI based and alrgely does things like Subject Alternative Names automatically.
[22:06:43] <dwd> It generates CSRs with all the right bits in, which then get universally ignored by comemrcial CAs. :-(
[22:07:42] <Florob> dwd, if you're going to go through this anyway: Could you take a look whether cacert swallows the XMPP related fields or not?
[22:07:56] <Zash> openssly says "othername:<unsupported>" about my cert
[22:08:17] <dwd> Zash, Oh? Hang on a tick, I'll look at what it's got.
[22:11:14] <dwd> http://dave.cridland.net/xeps/zash.se.png
[22:11:36] <dwd> Spiffy. xmppAddr's all there.
[22:11:42] <Zash> Yay
[22:11:49] <dwd> Weirdly, zash.se is duplicated.
[22:11:59] <dwd> No idea why. But Sodium never lies. :-)
[22:13:14] <dwd> No sRVNames, though.
[22:13:48] <dwd> Florob, Erm. I might get around to doing it tomorrow. If I do, I'll give CA Cert a proper, full-on CSR, and see what they do with it.
[22:13:59] <stpeter> gotta run, ttyl
[22:14:14] <dwd> Florob, I mean, with a Subject pointing to a real X.500 directory entry and everything. :-)
[22:14:30] <Zash> http://q.zash.se/7e2a9387.txt
[22:14:37] * stpeter left the chat.
[22:14:46] <Zash> That's the config I used
[22:14:55] <Florob> dwd, splendid
[22:16:00] <Zash> :D
[22:16:39] <Zash> They do eat O and OU fields
[22:16:52] <Zash> If you have below some number of points
[22:17:20] <dwd> Zash, Not fields. "RDNs of the Subject".
[22:17:50] <dwd> Zash, RDN being "Relative Distinguished Name".
[22:18:16] <dwd> Zash, I get lectured about this stuff, working somewhere that does all of X.500, not just the X.509 bit.
[22:19:05] <MattJ> Tobias, did we upset you in the Prosody MUC? :)
[22:19:24] <Tobias> MattJ: nope :)
[22:19:45] <MattJ> I'll add Apple to the list of cool things if you come back
[22:19:51] <MattJ> ...or maybe I won't
[22:19:52] <Zash> MattJ: NOOOO!
[22:19:59] <dwd> MattJ, Apples are nice.
[22:20:06] <dwd> MattJ, If you ferment them into Cider.
[22:20:08] <waqas> I had one today.
[22:20:12] <waqas> Not fermented.
[22:20:25] <Tobias> MattJ: it might be that one day there were too much offtopic/nonsense discussion in there :)
[22:21:05] <MattJ> The LDAP argument you mean? :)
[22:21:06] <Tobias> too much that i just couldn't take it anymore
[22:21:14] <Zash> /me has apple wine with flavor of cinnamon and vanilla
[22:21:24] <Zash> It tastes like liquid apple pie!
[22:21:26] <MattJ> I didn't know LDAP could be a controversial issue until tonight
[22:21:49] <dwd> MattJ, Oh, yes. Isode is split about whether LDAP is better than proper full-on DAP or not, you know.
[22:21:54] <Tobias> i don't remember any LDAP argument
[22:22:08] <Tobias> there is full-on DAP?
[22:22:08] <MattJ> Tobias, you should have been in the room then!
[22:22:13] <Zash> dwd: And I had no idea that X.500 where DAP stuff untill some days ago
[22:22:15] <deryni> And to think there wasn't even really any controversy, just confusion and cross-purpose talking. =)
[22:22:22] <dwd> Tobias, Sure. LDAP is "Lightweight" DAP.
[22:22:30] <Tobias> MattJ: i should have been a lot of things D
[22:22:31] <Tobias> *:D
[22:23:10] <MattJ> deryni, only because in the end I think the conclusion was "there is no right way" :)
[22:23:44] <dwd> MattJ, Don't *ever* suggest that to Steve in conversation. There is only one right way in LDAP, and he probably invented it. :-)
[22:23:54] <deryni> I think it was more there are two right ways, neither of which is what was being done. =)
[22:24:23] <MattJ> dwd, the question is how to auth a user
[22:24:40] <MattJ> Bind as a privileged account, and look up the user/credentials
[22:24:42] <MattJ> or bind as the user
[22:24:46] <dwd> MattJ, Ah. Yes.
[22:24:56] <dwd> MattJ, Well, if you bind as the user, you can only do PLAIN.
[22:25:12] <MattJ> Why? Isn't that dependent on implementation?
[22:25:16] <dwd> MattJ, Whereas if you bind as a priv account, you *might* be able to lookup the credentials.
[22:25:18] <deryni> With the added confusion that prosody currently binds as a priviledged account and then does a search for the user/pass pair. =)
[22:26:01] <dwd> deryni, You can't do that. You need to transform the user into a DN, first. The bind-as-user will do that for you, otherwise you need to do it yourself.
[22:26:14] <dwd> /me knows sadly far too much about this kind of thing.
[22:26:33] <deryni> I believe the dn creation was being done before the search.
[22:26:37] <dwd> MattJ, Most SASL mechanisms preclude a MITM, which is what bind-as-user is.
[22:27:06] <dwd> deryni, Right, but once you *have* a DN (possibly through searches), then you just lookup the DSE and yank the userPassword.
[22:27:28] <deryni> DSE?
[22:27:37] <MattJ> dwd, example? I don't see how they can detect a MITM, excepting SCRAM w/ channel binding
[22:27:41] <dwd> deryni, Directory Service Entry.
[22:28:04] <dwd> MattJ, Or DIGEST-MD5, which will want a service of "ldap" rather than "xmpp".
[22:28:05] <waqas> "...and yank the userPassword." - Which is what our current mod_auth_ldap is doing.
[22:28:16] <waqas> I think ^^
[22:28:17] <deryni> And then presumably you can store anything you want in userPassword?
[22:28:28] <deryni> Like SCRAM hashes?
[22:28:29] <dwd> deryni, Well, it's the password, traditionally.
[22:28:39] <dwd> deryni, No, that'd prevent anyone binding directly.
[22:28:46] <MattJ> deryni, there's a spec for storing SCRAM hashes
[22:28:47] <dwd> deryni, SCRAM hashes get their own attribute.
[22:28:58] <deryni> Yeah, remembered that as I hit <enter>.
[22:29:20] <deryni> I'm trying to see how you can do more than PLAIN with this method that binding doesn't allow, but I'm not seeing it.
[22:29:22] <dwd> deryni, One problem is that you may not be able to read userPassword at all.
[22:29:40] <deryni> Granted.
[22:30:37] <deryni> The prosody module currently does a search for a uid/userPassword pair using a configured basedn.
[22:30:58] <MattJ> http://code.google.com/p/prosody-modules/source/browse/mod_auth_ldap/mod_auth_ldap.lua#25
[22:31:20] <dwd> deryni, Well, you could allow any SASL mechanism which was susceptible to an MITM attack, I suppose, but you'd need to actually perform the bind operation at low-level yourself.
[22:31:21] <Tobias> couldn't one auth XMPP clients supporting SCRAM via proxy-auth at LDAP?
[22:31:28] <Tobias> or doesn't that make sense?
[22:31:33] <dwd> deryni, Doing PLAIN is *much* easier.
[22:32:23] <deryni> Assuming PLAIN the backend process is: bind-as-priv, search-for-dn, bind-as-dn-with-PLAIN-password?
[22:32:35] <dwd> deryni, Basically.
[22:32:56] <dwd> deryni, You can possibly do the search-for-dn as anonymous, of course.
[22:33:23] <deryni> I should have just asked you in the first place and saved myself most of an hour. =)
[22:33:57] <MattJ> "bind-as-dn-with-PLAIN-password" - this is on a new connection, or?
[22:34:04] <dwd> MattJ, That code won't work for quite a few LDAP deployments I've seen. It'll work with the M-Vault setup jabber.org uses, mind.
[22:34:20] <dwd> MattJ, IIRC, LDAP supports an Unbind operation.
[22:34:28] <MattJ> I see
[22:34:36] <dwd> MattJ, So in principle... But I dunno for sure.
[22:34:57] <MattJ> I don't see the point in the PLAIN step
[22:35:24] <dwd> MattJ, That's okay. Tell them to use M-Link instead for LDAP setups.
[22:35:38] <deryni> Which step?
[22:35:42] <MattJ> I'm assuming they can't afford it ;)
[22:35:56] <dwd> MattJ, Can they afford not to?
[22:35:58] <MattJ> deryni, why binding as the user is necessary
[22:36:21] <MattJ> dwd, afaik they're using ejabberd right now, so...
[22:36:30] <deryni> Because you can't always find the user by password from the user you bound as, AIUI.
[22:37:12] <MattJ> deryni, if you bound with a privileged account, could you not?
[22:37:13] <dwd> deryni, Aside from anything else, not all user DSEs will have a uid attribute, or a readable userPassword. Consider Active Directory.
[22:37:33] <dwd> MattJ, Not with AD, for sure. Not that you'd have a uid either, I think.
[22:37:41] <deryni> You don't need to convince me. I'm not the one questioning the need for the second bind.
[22:37:41] <MattJ> This stuff hurts my head... it's SQL but worse
[22:38:03] <MattJ> deryni, I'm asking because an hour ago you didn't understand it, and now you seem to :)
[22:38:06] <MattJ> and I still don't
[22:38:18] <dwd> MattJ, That's why we charge actual money. :-P
[22:38:46] <deryni> The problem before was less that I didn't understand it and more that I believed other people understood something about it I didn't, that turns out not to have been the case. =)
[22:38:56] <deryni> And of course dwd does understand it better than I do.
[22:39:18] <dwd> deryni, ... as the person who understands least about LDAP in Isode...
[22:39:20] <MattJ> It's his job to make it seem that way
[22:39:23] <deryni> Heh.
[22:39:49] <deryni> In a world of blind men the man with one eye is king.
[22:40:24] <dwd> Actually, Kev understands slightly less about LDAP than I do, come to think of it.
[22:40:30] <deryni> Anyway it seems to me that the idea is that you bind as the user since that's the accepted way to test your credentials, since you may not have the ability to find the user with the credentials any other way.
[22:40:45] <deryni> Especially if you can search anonymously.
[22:41:01] <deryni> Which, to bring this all back around, may suit Robot101 just fine.
[22:41:07] <deryni> But you'd have to ask him.
[22:41:07] <dwd> deryni, Oh, you don't *have* to. You *can* fish out the userPassword from the DSE, once you've located it.
[22:41:22] <dwd> deryni, We have a mode for each.
[22:42:06] <dwd> deryni, Robot101 as in Collabora, or someone else?
[22:42:19] <MattJ> Same Robot101, he brought up this issue
[22:44:04] <dwd> Right. Well, anyway, now that I've talked about LDAP and CRLs, I must be ready for bed. Cures insomnia, those two, like a dream, if you'll pardon the pun.
[22:46:16] * Tobias left the chat.
[22:48:44] * johnny left the chat.
[23:13:12] * johnny joined the chat.
[23:17:07] * scippio_netbook left the chat.
[23:18:12] * scippio_netbook joined the chat.
[23:18:21] <dwd> =-]
[23:22:18] <waqas> What kind of face is that?
[23:23:45] * gnufs left the chat.
[23:24:46] <evilotto> Mr Roboto?
[23:30:02] * wiretap left the chat.
[23:30:05] * wiretap joined the chat.
[23:32:50] * MattJ left the chat.
[23:57:41] * hawke left the chat.
[23:58:25] <louiz’> http://blog.bluendo.com/ff/facebook-user-location-patent-vs-xep-80 oh :o
[00:08:40] * whatever left the chat.
[00:08:40] * whatever joined the chat.
[00:18:55] * evilotto left the chat.
[00:35:22] * whatever left the chat.
[00:35:22] * whatever joined the chat.
[01:11:13] * whatever left the chat.
[01:11:13] * whatever joined the chat.
[02:09:24] * MattJ joined the chat.
[02:35:21] * whatever left the chat.
[02:35:21] * whatever joined the chat.
[02:48:16] * waqas left the chat.
[02:57:34] * justin left the chat.
[02:59:43] * johnny left the chat.
[03:06:23] * MattJ left the chat.
[03:23:03] * Florob left the chat.
[03:29:15] * Lance Stout left the chat.
[03:34:52] * johnny joined the chat.
[03:38:01] * Zash left the chat.
[04:05:31] * smoku left the chat.