Logs for jdev

Show join/part/nick changes:

[01:19:54] * Zash left the chat.
[01:40:03] * MattJ left the chat.
[01:46:24] * MattJ joined the chat.
[02:13:08] * Florob left the chat.
[02:21:11] * MattJ left the chat.
[05:12:40] * Xificurk left the chat.
[05:24:07] * Xificurk joined the chat.
[07:27:06] * Tobias joined the chat.
[07:36:18] * Alex joined the chat.
[08:00:34] * scippio joined the chat.
[08:04:41] * jonas joined the chat.
[08:33:15] * jonas left the chat.
[08:37:16] * dwd joined the chat.
[08:55:52] * guus joined the chat.
[09:05:04] * jonas joined the chat.
[09:17:36] * luca tagliaferri joined the chat.
[09:27:41] <dwd> Anyone know if the suggestion of preserving child elements of <item/> ion roster sets was ever codified in RFC 6121? I thought it was, but I can't find it...
[09:32:13] <mathieui> Any thoughts about adding xml:lang attributes in each stanza a client sends?
[09:36:47] <dwd> You probably should ensure that the effective value for xml:lang in each stanza remains the same in each stream, despite any changes in the default provided by the stream.
[09:38:50] * Treebilou joined the chat.
[09:38:52] <mathieui> Would that cause a problem?
[09:42:36] <dwd> In theory, or in practise?
[09:43:08] <dwd> In theory, error messages may not be in the right language. In practise, those implementations that send textual errors tend to ignore the xml:lang anyway.
[09:43:37] <dwd> I don't *think* anything else checks it - although (again in theory) there could be language translation services, or MUC traffic enforcement, that would.
[09:44:25] <mathieui> dwd, wouldn’t the service use the latest value it received?
[09:44:42] <dwd> Maybe. But for S2S, that could easily be wrong.
[09:46:52] <Alex> as dwd said it makes sense to add it to teh stream or the stanzas for error message, registrations forms and other stuff
[09:47:23] <Alex> in theory you could have content with muliple languages liek described here: http://xmpp.org/extensions/xep-0171.html
[09:47:58] <Alex> in pratise I have not seen anybody using it. And I have concerns with it, because most existing libraries and client will fail
[09:52:05] <mathieui> (would be nice, though)
[09:53:33] <Alex> yes, can be nice if you need it
[09:54:24] <Alex> I have done some stuff with it some years ago. At this time most frameworks were not designed for multiple languages
[09:55:18] <Alex> so many libs completly ignore all xml:lang stuff, and for example when they get the body of a message they just fetch the first child with the tag body
[10:11:26] * Treebilou left the chat.
[10:14:25] * Treebilou joined the chat.
[10:48:08] * Zash joined the chat.
[11:17:31] * Zash left the chat.
[11:25:51] * Xificurk left the chat.
[11:27:24] * Xificurk joined the chat.
[11:29:15] <dwd> Next question - is the empty string legal as a roster group name?
[11:35:45] <Hermitifier> No, it is not legal.
[11:37:25] <Hermitifier> The server MUST return a <not-acceptable/> stanza error to the client if the roster set contains any of the following violations: 2. The XML character data of the <group/> element is of zero length
[11:57:24] <dwd> Ah! Yes, found that now. Thanks!
[11:58:35] <Hermitifier> Also the RFC suggests treating group names like resource names.
[12:00:49] <dwd> Yes, saw this, the general description of groups I'd read.
[12:00:56] <dwd> I'd assumed anything would be mentioned there.
[12:04:37] <Alex> dwd: you got my mail about b64 in vcards?
[12:06:04] <dwd> Alex, Yes, thanks. I'll look at this today - 99% sure I know what it is, but surprised nobody noticed before.
[12:14:13] <Alex> dwd: adding line feeds fixed it for now. RFC4648 sais: Implementations MUST NOT add line feeds to base-encoded data unless the specification referring to this document explicitly directs base encoders to add line feeds after a specific number of characters.
[12:20:10] <dwd> Alex, Yup, we certainly have a bug here.
[12:21:56] <Alex> I think it would be good to add a note somewhere how b64 data should be handled. My parser doesn't care, but some do.
[12:23:05] * scippio left the chat.
[12:24:15] * scippio joined the chat.
[12:32:18] * user1@localhost joined the chat.
[12:32:32] * user1@localhost left the chat.
[12:54:35] <Alex> I am doing some work on disco items caching right now. Are there any "best practices"? or doesn't male it sense at all`?
[13:04:45] <Kev> Caching of what and why, Alex?
[13:05:16] <Alex> caching disco items of the server to reduce traffic on each login
[13:06:03] <Kev> That sounds like it's probably not a great idea.
[13:06:04] <Alex> they normally don't change that often
[13:06:21] <Kev> They don't - but they may be different for different users.
[13:06:21] <Alex> we need caps on them, or TTL like in DNS
[13:06:29] <Kev> Oh, wait.
[13:06:35] <Kev> You mean caching them clientside so they don't get requested again?
[13:06:46] <Alex> yes
[13:07:12] <Alex> currently I disco the server on each login to find conference and pubsub services
[13:07:46] <Kev> Right. And that's too expensive?
[13:08:41] <Alex> its probably not, but I try to optimize where possible
[13:09:28] <Alex> caps on disco items would be great, then I could at least get rid of some info requests
[13:09:40] <Alex> BRB
[13:15:05] * lastsky joined the chat.
[13:15:28] * Patrick Stadler joined the chat.
[13:15:45] * weibelm joined the chat.
[13:16:07] * Patrick Stadler left the chat.
[13:16:08] * Patrick Stadler joined the chat.
[13:17:31] * Patrick Stadler left the chat.
[13:17:34] * pstadler joined the chat.
[13:23:43] <Alex> back
[13:32:40] * user1@localhost joined the chat.
[13:32:58] * user1@localhost left the chat.
[13:41:11] * naw joined the chat.
[13:52:18] * weibelm left the chat.
[13:55:48] * Florob joined the chat.
[14:31:58] * naw left the chat.
[14:49:30] * Zash joined the chat.
[15:01:04] * deryni left the chat.
[15:09:26] * Destructhor joined the chat.
[15:17:39] <Zash> I wonder if the gtalk guys are ever going to deploy TLS on s2s
[15:18:35] <dwd> Zash, Ask on operators@? But I'd imagine the short answer is not until DNA - the problem is all the hosted apps domains, as I understand things.
[15:19:32] <Zash> Yeah, I remember something like that
[15:20:50] <Tobias> yeah...and IIRC we're currently waiting for larger deployment of DNSSEC, right?
[15:21:21] <dwd> FWIW, I think that they could reasonably generate certs off a private CA in their case. It'd give sufficient security, I think - or at least substantially better security than nothing.
[15:21:48] <Zash> They have their own CA already
[15:22:08] * Destructhor left the chat.
[15:23:03] <Tobias> so everybody wanting to interop with gtalk or a google hosted domain would add their cert to the list of trusted Root CAs and that's it
[15:24:35] <Kev> Tobias: Trusting the root is rarely relevant in current-day S2S interop.
[15:24:59] <Tobias> Kev, yeah..in current day S2S interop
[15:25:23] <Tobias> but i think they are waiting for a trusted solution
[15:27:23] <dwd> Tobias, Well, I assume they are.
[15:28:19] * Alex left the chat.
[15:28:35] <Tobias> dwd, what's google hosted stuff's stand on the SMTP site? are they doing TLS connections there? do you know?
[15:28:59] <Tobias> they would have similar problems there
[15:29:03] <dwd> Tobias, I'm not sure. I don't think so - I think I'd have noticed.
[15:29:16] <Zash> Received: from mail-ey0-f172.google.com (mail-ey0-f172.google.com [209.85.215.172]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority" (verified OK)) by mail.zash.se (Postfix) with ESMTPS id 5AF371B3C6 for <zash@zash.se>; Tue, 18 Oct 2011 00:36:59 +0200 (CEST)
[15:29:24] <dwd> Tobias, Except that MTA TLS is only very vaguely specified.
[15:29:38] <dwd> Ah, there we are.
[15:30:44] * MattJ joined the chat.
[15:31:30] <Tobias> dwd, are there plans within the IETF or else to specify it more clearly?
[15:32:16] <Tobias> Zash, the "verified OK" means what exactly? that the cert you got, got signed by a trusted root CA? or more
[15:33:22] <Zash> Tobias: That something matched something, and that the chain was trusted... I think
[15:33:41] <dwd> Tobias, RFC 3207 says: The decision of whether or not to believe the authenticity of the other party in a TLS negotiation is a local matter.
[15:33:51] <dwd> Tobias, Which is, in my view, vague.
[15:35:32] <Tobias> yeah..that soundls like you could do anything and would still comply with the standard
[15:36:31] <Tobias> well..it interops and allows secure connections but without correct asserted authenticity...
[15:37:33] * lastsky left the chat.
[15:41:43] <Zash> But then, you can auth as whatever and send MAIL FROM some completly different domain
[15:42:06] <Tobias> bbl
[15:43:14] * deryni joined the chat.
[15:48:17] * Tobias left the chat.
[15:49:43] <Zash> X509v3 Basic Constraints: critical CA:TRUE, pathlen:0
[16:13:30] * Zash left the chat.
[16:15:45] * Zash joined the chat.
[16:33:02] * pstadler left the chat.
[16:57:07] * luca tagliaferri left the chat.
[17:08:26] * waqas joined the chat.
[17:16:07] * psa joined the chat.
[17:21:55] * Florob left the chat.
[17:21:56] * Florob joined the chat.
[17:23:14] * Tobias joined the chat.
[17:38:36] * waqas left the chat.
[17:38:37] * guus left the chat.
[17:38:44] * waqas joined the chat.
[17:42:21] * jonas left the chat.
[17:54:53] * waqas left the chat.
[17:54:59] * waqas joined the chat.
[17:55:08] * waqas left the chat.
[17:55:30] * waqas joined the chat.
[18:15:28] * scippio left the chat.
[18:16:31] * waqas left the chat.
[18:17:57] * evilotto joined the chat.
[18:27:47] * waqas joined the chat.
[18:28:48] * jonas joined the chat.
[18:50:13] * Zash_ joined the chat.
[18:55:48] * Zash_ left the chat.
[18:56:38] * Zash left the chat.
[19:06:23] * Alex joined the chat.
[19:07:09] * Alex left the chat.
[19:15:50] * naw joined the chat.
[19:17:19] * waqas left the chat.
[19:17:26] * waqas joined the chat.
[19:42:31] * Alex joined the chat.
[19:49:25] * Zash joined the chat.
[19:55:07] * waqas left the chat.
[19:55:23] * waqas joined the chat.
[19:57:40] * waqas left the chat.
[20:01:03] * Florob left the chat.
[20:01:30] * psa left the chat.
[20:05:25] * Alex left the chat.
[20:13:18] * Treebilou left the chat.
[20:51:41] * psa joined the chat.
[20:57:50] * Zash_ joined the chat.
[20:58:21] * Zash left the chat.
[21:11:18] * deryni left the chat.
[21:37:15] * Alex joined the chat.
[21:51:49] * Zash joined the chat.
[21:57:19] * Zash left the chat.
[21:58:47] * Zash_ left the chat.
[21:59:05] * naw left the chat.
[22:02:36] * Alex left the chat.
[22:07:05] * Alex joined the chat.
[22:09:19] * jcea joined the chat.
[22:11:23] * Zash joined the chat.
[22:12:38] * Alex left the chat.
[22:17:10] * deryni joined the chat.
[22:24:52] * darkrain left the chat.
[22:30:10] * darkrain joined the chat.
[22:35:27] * darkrain left the chat.
[22:36:38] * darkrain joined the chat.
[22:44:25] * jonas left the chat.