Logs for jdev

Show join/part/nick changes:

[00:30:35] * johnny joined the chat.
[00:30:48] * sprite left the chat.
[00:35:49] * sprite joined the chat.
[00:35:49] * sprite left the chat.
[00:40:49] * sprite joined the chat.
[00:40:49] * sprite left the chat.
[00:45:50] * sprite joined the chat.
[00:45:50] * sprite left the chat.
[00:50:50] * sprite joined the chat.
[00:50:50] * sprite left the chat.
[00:55:50] * sprite joined the chat.
[00:55:50] * sprite left the chat.
[01:00:50] * sprite joined the chat.
[01:00:50] * sprite left the chat.
[01:05:51] * sprite joined the chat.
[01:05:51] * sprite left the chat.
[01:10:51] * sprite joined the chat.
[01:10:51] * sprite left the chat.
[01:15:51] * sprite joined the chat.
[01:15:51] * sprite left the chat.
[01:20:52] * sprite joined the chat.
[01:20:52] * sprite left the chat.
[01:24:49] * luca tagliaferri left the chat.
[01:25:52] * sprite joined the chat.
[01:40:10] * Florob joined the chat.
[01:40:54] * sprite left the chat.
[01:45:54] * sprite joined the chat.
[01:45:54] * sprite left the chat.
[01:50:55] * sprite joined the chat.
[01:50:55] * sprite left the chat.
[01:53:05] * evilotto left the chat.
[01:55:55] * sprite joined the chat.
[02:11:06] * deryni joined the chat.
[02:11:06] * sprite left the chat.
[02:11:08] * Florob left the chat.
[02:15:57] * sprite joined the chat.
[02:16:27] * jcea left the chat.
[02:35:13] * Zash left the chat.
[02:35:34] * Zash joined the chat.
[02:40:58] * Zash left the chat.
[02:42:37] * Zash joined the chat.
[02:52:35] * darkrain_ left the chat.
[02:59:06] * Zash left the chat.
[03:03:24] * waqas left the chat.
[04:04:20] * waqas joined the chat.
[04:16:06] * johnny left the chat.
[05:02:46] * johnny joined the chat.
[05:18:18] * MattJ left the chat.
[06:06:18] * sprite left the chat.
[07:15:56] * niekie left the chat.
[07:25:01] * lastsky joined the chat.
[07:26:09] * lastsky left the chat.
[07:34:30] * ermine joined the chat.
[07:43:24] * deryni left the chat.
[07:44:07] * mlundblad_laptop joined the chat.
[08:07:54] * jonas joined the chat.
[08:23:53] * lastsky joined the chat.
[08:24:30] * lastsky left the chat.
[08:33:38] * lastsky joined the chat.
[08:33:42] * lastsky left the chat.
[08:37:09] * Hermitifier joined the chat.
[09:28:00] * smoku left the chat.
[09:28:17] * smoku joined the chat.
[10:06:45] * Tobias joined the chat.
[10:37:48] * dwd left the chat.
[10:37:54] * dwd joined the chat.
[10:55:48] * scippio left the chat.
[11:06:48] * dwd left the chat.
[11:06:48] * dwd joined the chat.
[11:39:31] * Florob joined the chat.
[11:51:34] * waqas left the chat.
[12:22:32] * scippio joined the chat.
[12:51:49] * will.thompson joined the chat.
[13:05:55] * will.thompson left the chat.
[13:09:42] * MattJ joined the chat.
[13:13:15] * will.thompson joined the chat.
[13:17:52] * jameschurchman joined the chat.
[13:26:25] * duck1123 left the chat.
[13:43:04] * Neustradamus left the chat.
[13:43:52] * Neustradamus joined the chat.
[13:48:52] * Hermitifier left the chat.
[14:24:14] * scippio left the chat.
[14:26:43] * scippio joined the chat.
[14:37:31] * Zash joined the chat.
[14:47:10] * niekie joined the chat.
[15:15:20] * deryni joined the chat.
[15:25:13] * Lance joined the chat.
[15:25:38] * Lance left the chat.
[15:44:35] * simon joined the chat.
[15:45:59] * darkrain_ joined the chat.
[15:48:10] <simon> I would like ad-hoc chatrooms i.e. start chatting with two users automatically creates a room and makes them join. If I understand correctly this is currently impossible because the standard has no functionality to make a user join a room, correct?
[15:48:11] * justin left the chat.
[15:49:40] <jonas> simon, there is a protocol for it, and at least gajim supports it
[15:50:18] <darkrain_> It's an *invitation*
[15:51:13] <darkrain_> "When the client...receives the invitation, it SHOULD auto-join the room or prompt the user whether to join..."
[15:51:32] <MattJ> You can't force a client to do anything in a protocol
[15:51:57] <darkrain_> Well, you can MUST something, in which case you can argue the client isn't actually implementing the protocol unless they do it :)
[15:52:12] <MattJ> My client would still prompt the user :)
[15:52:36] <darkrain_> Right, auto-joining chat rooms that the user hasn't (in some way) configured to be autojoined is evil
[15:52:40] <darkrain_> (IMHO)
[15:52:54] <deryni> MSN does that, right?
[15:53:32] <Zash> MattJ: Skype forces you into channels, and you won't notice untill someone says something.
[15:53:34] <darkrain_> I don't know how the official client actually treats it, but I suspect so
[15:53:47] <darkrain_> Google Talk chat (via the gmail web interface) does the same thing, IIRC
[15:54:04] <deryni> I love it.
[15:54:15] <Zash> http://xmpp.org/extensions/xep-0045.html#continue
[15:54:23] <Zash> simon: ^
[15:54:26] <jonas> gajim prompts the user as well
[15:55:12] <deryni> Transitioning an IM conversation into a muc isn't what he was asking for exactly though, that I could tell.
[15:55:13] <jonas> deryni, in MSN, every chat is a MUC
[15:55:37] * mlundblad_laptop left the chat.
[15:55:46] <deryni> That's true but I was more specifically talking about the new groups with that, though it is likely also true (to an extent) with the old ones.
[16:00:40] <simon> well, I think there should be an option like "auto-join group chats invited to by people in this group"
[16:01:02] <MattJ> Sure
[16:01:09] <simon> but apparently some of what I want does exist...
[16:01:15] <simon> I just haven't found it yet ;)
[16:01:16] <MattJ> If the users are in your control, you can also make a client that does that without a UI option
[16:01:20] <deryni> Right, but that's entirely a client option.
[16:01:33] <MattJ> but the point is that you can't say XMPP doesn't have a way to do that :)
[16:02:10] <Zash> You might also make the server/muc think that the user already joined
[16:02:12] <simon> well, yeah you could create a client that automatically creates a room, gives it a random name and invites people
[16:02:22] <Zash> But client behaviour in that case is undefined
[16:02:43] <simon> but it would be nice if this was defined so that some sort of conformity could be expected
[16:03:24] <deryni> There's nothing to define here.
[16:03:49] <dwd> You can, with XEP-0223 based bookmarks, have the server publish a new auto-join bookmark.
[16:04:00] <dwd> Which in principle might have the client auto-join it.
[16:04:08] <deryni> Invites already exist. Clients can send them and receive them. Whether clients have the option (or don't have the option) to auto-join base on invites is up to the client.
[16:04:42] <simon> deryni: I agree.
[16:05:03] <simon> dwd: auto-join bookmark... never heard of that :)
[16:05:15] <Zash> !xep bookmarks
[16:05:16] <Kanchil> Zash: XEP-0048: Bookmarks is Standards Track (Draft, 2007-11-07) See: http://xmpp.org/extensions/xep-0048.html
[16:05:26] <darkrain_> !xep 223
[16:05:26] <Kanchil> darkrain_: XEP-0223: Persistent Storage of Private Data via PubSub is Informational (Active, 2008-09-08) See: http://xmpp.org/extensions/xep-0223.html
[16:05:31] <deryni> Yeah, 223 might be interesting.
[16:06:24] <deryni> I wonder if clients support auto-joining on publish notifications.
[16:06:30] <Zash> Plain ol' iq:private would work to, but it's only read when the client starts/logs in
[16:06:47] <Zash> deryni: What clients support 223?
[16:06:51] <simon> since I've never written a client this discussion is over my head now ;)
[16:07:04] <deryni> Zash: Also a relevant question. =)
[16:07:33] <Zash> support for '48 in iq:private is in most clients i think
[16:08:57] <deryni> Right, but doesn't get you "anytime" joining.
[16:09:05] <Zash> No
[16:09:49] <Zash> But, 223+48+autojoin pushed woul probably be tricky to implement properly
[16:10:15] * bjc joined the chat.
[16:10:28] <Zash> Eg, if you manualy leave, and then the bookmark is changed from another client, and pushed, what do you do?
[16:10:42] <deryni> Why would you do both 223 and 48? But even if you did the tricky part would be keeping the iq:private and pep data in sync especially if not all your clients support both.
[16:10:59] <deryni> Yeah, which is sort of why I asked about push-join to begin with.
[16:11:16] <deryni> I'm not sure that's something I think is reasonable to do at all let alone automatically.
[16:11:17] <Zash> 48 doesn't specify storage, only format
[16:11:23] <Zash> iirc
[16:11:44] <deryni> Right. I was making an assumption.
[16:11:54] <Zash> it mentions both iq:private and pubsbu
[16:11:57] <Zash> pubsub*
[16:12:04] <Zash> and 223
[16:12:13] <deryni> Yeah. I was looking at it the other day.
[16:12:43] <simon> Apparently it's a complicated issue.
[16:14:09] <deryni> Auto-joining on invites as a client option isn't complicated.
[16:14:27] <deryni> The fancier stuff gets complicated when you try to figure out interaction issues.
[16:14:46] * Ludovic joined the chat.
[16:15:55] <simon> Is there a commonly preferred clients amongs devs that supports lots of stuff?
[16:16:56] <Zash> Gajim :)
[16:19:12] <simon> /me ponders having an X server running all the time.
[16:19:21] <Kev> Or Psi.
[16:19:31] <Kev> Gajim/Psi are the usual ones.
[16:24:31] <simon> Yeah, psi I'm using now because of the native osx support.
[16:24:56] <Zash> Oh, OS X :
[16:24:57] <Zash> :(
[16:25:38] <simon> yeah...
[16:25:48] <simon> need a new laptop...
[16:26:11] * jonas left the chat.
[16:26:23] <simon> Gajim needs a terminal only mode. :)
[16:26:32] <Zash> simon: meet gajim-remote ;)
[16:26:47] <simon> /me looks it up.
[16:26:54] <Zash> Needs dbus thou
[16:27:19] <simon> yuck <o>
[16:28:30] <simon> well maybe dbus isn't so bad...anyway I will try that
[16:28:40] <simon> /me adds more stuff to his todo list.
[16:28:49] <simon> my todo list for free time is longer than my todo list for work
[16:29:39] <deryni> I'm assuming gajim-remote controls an already running gajim the way purple-remote controls an already running pidgin, right?
[16:29:44] <Zash> .. yes
[16:30:12] <Zash> So not exactly "terminal only"
[16:36:56] * stpeter joined the chat.
[16:37:29] <simon> still worth trying
[16:37:30] * jameschurchman left the chat.
[16:39:25] <Zash> Still, Iduno how maintained Gajim on OS X is. *pokes Asterix*
[16:40:01] <simon> /me shrugs. I will try later and tell you.
[16:45:19] <Florob> it isn't
[16:45:46] <Florob> js got bored of trying to support it. And I don't think someone else did work on it
[16:48:00] * Yagiza joined the chat.
[16:48:11] <simon> hmhmhm
[16:48:13] <Yagiza> Hello!
[16:49:58] <Yagiza> /me wonders if PEP notification may contain more than one item...
[16:50:08] * Ludovic left the chat.
[16:50:34] <simon> /me wonders why Yagiza talks in orange while everybody else talks in black.
[16:51:09] <Yagiza> Errr... orange?
[16:51:15] <Zash> because you don't ignore xhtml?
[16:51:21] <simon> Well, my Psi makes it orange.
[16:51:26] <deryni> iChat?
[16:52:05] * jameschurchman joined the chat.
[16:52:33] <Yagiza> It seems simon is annoyed by "/me" messages...
[16:53:06] <deryni> Ah.
[16:53:28] <simon> I'm stupid so I easily get confused.
[16:53:53] <Yagiza> (^_^)
[16:54:32] <Florob> Interestingly my Gajim choose the same colour :D
[16:55:35] <Yagiza> Well... I never saw multiple events in one PEP message, but I still didn't find statement that forbids that.
[16:57:22] <Yagiza> So, I wonder if my client must look for more than one event in the message.
[16:58:12] * jameschurchman left the chat.
[16:58:13] <Florob> Yagiza, the schema actually allows any number of <item/>s or <retract/>s but not both at the same time. Not sure that's intentional though...
[17:00:04] <Yagiza> I mean different nodes. But node specified in <event /> element, so I'm asking 'bout multiple <event /> elements in one message.
[17:00:40] <Florob> ah...
[17:00:43] <Florob> /me looks again
[17:00:45] <Yagiza> For example, is it possible, that one message can contaim both mood and activity event?
[17:02:52] <Florob> you mean multiple <items/> in one <event/>, right?
[17:03:07] <Florob> which is not allowed
[17:03:55] <Yagiza> No. I mean multiple <event /> in one <message />
[17:06:21] <Zash> Florob: It's not?
[17:07:34] <Florob> Zash, Actually... my XML Schema-foo is not strong enough to be sure.
[17:07:50] <Zash> Oh, multiple itemS
[17:08:00] <deryni> That's what it looks like to me too. Multiple <item/>s are allowed in an <items/> but only one <items/> in an <event/>.
[17:08:26] <Zash> I'd find it weird if it only allowed one <item> in a <items> :P
[17:08:40] <Florob> Multiple <event/>s in a <message/> are allowed in theory I guess... But I doubt anyone generates that
[17:08:55] <Yagiza> Ok
[17:09:21] <deryni> I'm not sure allowing or disallowing that is a pubsub/pep level thing even, is it?
[17:10:02] <Florob> Zash, actually. I had originally expected that <items xmlns="pubsubns#events"/> allows only one, but <items xmlns="pubsubns"/> allows multiple, because that is what's actually generated/described AFAICT
[17:10:32] <Yagiza> Yeah. It's a pubsub level thing.
[17:10:35] <Florob> deryni, that is what I'm wondering. I guess that is RFC level, but that allows any number of any element it doesn't know about AFAICT...
[17:12:44] <Yagiza> Some XEP's disallows multiple elements of specific type, but I'm not sure 'bout this case.
[17:14:19] <Florob> At any rate, generating this would require caching etc., which defeats PubSubs purpose IMHO
[17:15:06] <Yagiza> Why caching?
[17:15:12] <deryni> You mean delaying notifications to send in a burst, right? Not caching. Since caching of old notifications can happen as is.
[17:15:38] <Yagiza> deryni, you're right
[17:16:35] <Florob> yes, what deryni said
[17:16:38] * Treebilou left the chat.
[17:20:59] <Asterix> Zash: Gajim is not maintained on OSX. if we could replace the objc code to pyobjc, that would help a lot: no more compilation, so it would be included in trunk
[17:24:01] <deryni> pep on-presence notifications could need to send multiple events, right?
[17:24:55] <MattJ> /me wonders why they they didn't make TCP message oriented
[17:25:31] <Zash> You have UDP for that
[17:25:48] <MattJ> Not quite, UDP messages have a finite length
[17:26:05] <MattJ> You can do the same with TCP if you keep your messages short enough too
[17:26:13] * scippio left the chat.
[17:27:08] <MattJ> I'm questioning why they decided a single TCP unit of transfer should fit into an IP packet
[17:27:12] * Florob left the chat.
[17:27:15] * Florob joined the chat.
[17:28:19] <Florob> what the.... Did I just rejoin multiple times?
[17:28:44] <MattJ> [17:27:12] Florob has left [17:27:15] Florob has joined the group chat
[17:29:44] <Florob> That's not multiple times then.... I got history many times though. Or at least Gajim displays it many times :/
[17:30:04] <MattJ> Client?
[17:30:13] <Florob> deryni, right, auto-subscribe can subscribe you to multiple nodes simultaneously...
[17:30:35] <Florob> MattJ, are you asking me?
[17:30:54] <MattJ> I am, just in case :)
[17:31:05] <Florob> MattJ, "Or at least Gajim displays it many times :/"
[17:31:42] <dwd> Florob, From here? COuld be that somehow the join request got repeated.
[17:32:08] <Florob> dwd, yes from here
[17:32:24] <Florob> Which... would've been a good thing to check. Yes just from this MUC
[17:54:13] <dwd> Florob, M-Link reacts to join requests specifically, so if it sees multiple "joinish" presences, then it'll send history for each of them.
[17:54:56] <Florob> dwd, but where would it get multiple of them from?
[17:55:10] <dwd> Florob, That I don't know.
[18:00:44] <darkrain_> Google retransmits?
[18:02:39] * evilotto joined the chat.
[18:36:52] * mlundblad joined the chat.
[18:47:20] <Florob> darkrain_, google?
[18:48:20] <darkrain_> Google Talk retransmits of <presence/> stanzas? Maybe I misunderstood what was going on
[18:49:00] <stpeter> http://www.ietf.org/id/draft-moffitt-xmpp-over-websocket-00.txt
[18:51:41] <Florob> darkrain_, I guess you didn't. But it doesn't apply to the my Prosody to jabber.org's M-Link ;)
[18:51:49] <Florob> case
[18:52:05] <deryni> I would hope Google doesn't retransmit the join bits.
[18:52:15] * deryni left the chat.
[18:55:09] * scippio joined the chat.
[18:56:38] <Zash> Didn't they just disable websockets in all browsers due to a security bug?
[19:00:08] <Tobias> yeah
[19:00:10] <Tobias> :)
[19:08:13] * lastsky joined the chat.
[19:38:26] * deryni joined the chat.
[20:04:12] * lastsky left the chat.
[20:22:55] * will.thompson left the chat.
[20:27:50] * Ludovic joined the chat.
[20:35:18] * max joined the chat.
[20:39:35] * simon left the chat.
[20:39:35] * simon joined the chat.
[20:46:58] * darkrain left the chat.
[20:49:47] * simon left the chat.
[20:49:47] * simon joined the chat.
[20:51:26] * max left the chat.
[20:57:39] * Tobias left the chat.
[21:05:29] * Tobias joined the chat.
[21:21:50] * Ludovic left the chat.
[21:25:47] * duck1123 joined the chat.
[21:35:37] * simon left the chat.
[21:35:45] * simon joined the chat.
[21:36:03] <simon> what are websockets anyway?
[21:36:55] <MattJ> There's a whole IETF WG dedicated to finding the answer to that question
[21:37:03] <MattJ> and they still haven't succeeded
[21:37:56] <simon> :)
[21:38:02] * mlundblad left the chat.
[21:38:47] <Florob> Yes, I'm always wondering how people implement open questions...
[21:39:04] <MattJ> simon, essentially they want a web page to be able to communicate efficiently with a web server
[21:39:24] <MattJ> current solutions are built on top of HTTP, which works, but is fairly inefficient
[21:39:42] * Yagiza left the chat.
[21:52:02] <simon> hm...
[21:52:02] * simon left the chat.
[21:52:02] * simon joined the chat.
[21:53:06] <simon> man connecting took forever
[21:53:27] * evilotto left the chat.
[21:58:08] * lastsky joined the chat.
[22:24:42] * teo1 left the chat.
[22:24:44] * teo1 joined the chat.
[22:37:58] * ermine left the chat.
[22:39:01] * simon left the chat.
[22:48:42] * louiz’ left the chat.
[22:48:47] * louiz’ joined the chat.
[22:49:43] * simon joined the chat.
[22:50:42] * lastsky left the chat.
[22:52:02] * darkrain_ left the chat.
[22:52:02] * darkrain_ joined the chat.
[22:55:07] * darkrain_ left the chat.
[22:55:08] * darkrain_ joined the chat.
[22:55:45] * evilotto joined the chat.
[22:58:24] * duck1123 left the chat.
[23:01:13] * stpeter left the chat.
[23:07:23] * evilotto left the chat.