[00:08] < Chong-> It seems ok. Done. Sorry for bothering you all, just now.
[00:11] < Chong-> MS-: Just posted the new Review report of private_MPS_JL_IRCSupport to the list. Have a look when you have time. thanks.
[00:11] < Chong-> Gotta sleep. night, all.
[00:11] *** Chong- has parted #kamaelia
[01:37] *** vmlemon_ has joined #kamaelia
[06:54] < Lawouach_> morning
[08:03] *** orphans has joined #kamaelia
[08:09] *** bcarlyon|ubuntu has joined #kamaelia
[08:32] *** mhrd-afk is now known as mhrd
[08:47] *** MS- has joined #kamaelia
[08:47] < MS-> morning
[08:49] < orphans> morning MS-
[08:52] < MS-> 300 spams last night...
[08:54] < orphans> go go Kamaelia Spam Squad
[08:57] < MS-> :-(
[08:58] < orphans> if only you could build concurrency into yourself eh :)
[08:59] < MS-> heh, people are naturally concurrent. It's the body that only has one pair of hands, one pair of eyes, one pair of legs one pair of ears and a largely serialised mind
[08:59] < MS-> The rest of the system is incredibly concurrent :)
[09:00] < MS-> (hence Axon...)
[09:01] *** bcarlyon|laptop has joined #kamaelia
[09:03] < orphans> mm, the mind and body would be the useful bit (although perhaps less effective for, yknow, generally being alive) :)
[09:06] < MS-> All I need to harness is a good split personality disorder, that's all
[09:06] < MS-> ;)
[09:07] < orphans> get on ebay - you can find anything there :)
[09:09] < MS-> heh
[09:09] < MS-> The real problem is that I'm fighting markov chains at the moment
[09:12] < orphans> http://en.wikipedia.org/wiki/Markov_chain#Music < = interesting :)
[09:12] < orphans> might try and write one if I get a moment
[09:14] < MS-> Oh, I've got a kamaelia based markov chain server somewhere
[09:14] < MS-> I should've realised it'd work with music too
[09:15] < MS-> Would be interesting to hook that up with the midi component
[09:15] < MS-> An ever lasting lift music generator :-)
[09:15] < MS-> Mind you, it could be an interesting addition for your Jam project
[09:15] < orphans> I'd never listen to anything else :)
[09:15] < MS-> heh
[09:15] < orphans> yeah - computer players
[09:15] < orphans> you could run a musical turing test!
[09:16] < orphans> I reckon I couldn't tell
[09:16] < MS-> heh
[09:16] < MS-> I think that'd be interesting actually
[09:16] < MS-> After all, music normally has an emotional content to it
[09:17] < MS-> Which *should* be hard to duplicate
[09:17] < orphans> yeah - I do tend to find algorithmic music a bit challenging to listen to
[09:17] < orphans> but that may well be because of the type of person who makes algorithmic music tends to like challenging music
[09:18] < orphans> best part of that article - "These higher-order chains tend to generate results with a sense of phrasal structure, rather than the 'aimless wandering' produced by a first-order system"
[09:18] < orphans> I've played with a fair few people who perfected that aimless wandering :)
[09:18] < MS-> Dumb q which you will hopefully know the answer to - if I got a USB to midi adapter (eg something like: http://www.microanvika.com/product.asp?PNO=RLD51926 ) do you think that'd work with your code?
[09:19] < orphans> mm, not 100% sure
[09:20] < orphans> if the driver creates a real MIDI device (which I'd image it would) then it should do
[09:20] < MS-> "probably" ?
[09:20] < orphans> I'm not really a hardware man though, so I tend to just use midi internally
[09:20] < orphans> yeah, I'd say probably :)
[09:20] < MS-> (which is a conditional word which means if sods law kicks in means "no")
[09:21] < MS-> cool
[09:21] < MS-> (wouldn't be a total loss if it didn't - I know where it could be used)
[09:21] < orphans> I can't imagine how else they'd do it, but sometimes developers have a mind of their own...
[09:22] < MS-> Well, yes, indeed :)
[09:22] < MS-> Mainly asked because you have more clue about midi than me :-)
[09:23] < orphans> downloading the manual now - that should make it a bit clearer
[09:23] < MS-> I'd be half tempted to label the keys on a (musical) keyboard A-Z, shift, return etc and use that as a computer keyboard
[09:23] < MS-> Oh I was using that one for illustration :)
[09:24] < orphans> I'd imagine they are all pretty much the same though :)
[09:24] < MS-> Indeed - I suspect that one would for example though given this comment:
[09:24] < MS-> Supports both advanced and OS-standard drivers (Win/ Mac OS X)
[09:24] < MS-> specifically the "OS-standard" bit
[09:25] < orphans> yeah, it gives you a standard midi out - should work fine
[09:26] < orphans> funny you mention labelling the keyboard - I've just done the reverse. Cannibalising old PC keyboards to make foot switches and converting the keypresses to midi messages
[09:26] *** Davbo has joined #kamaelia
[09:28] < MS-> orphans: Yeah, but people having been playing music on computer keyboards for aaages. Writing code on a musical keyboard would make many people blanche
[09:28] < MS-> Which'd be fun
[09:29] < MS-> Probably be a real pain to "type" on
[09:29] < orphans> I dunno, if you hook it up for notes and key presses you could at least provide accompaniment to your coding :)
[09:30] < MS-> heh
[09:30] < MS-> I guess I ought to stop that, gets back to doing what I should be doing
[09:31] < Davbo> Mornin' all
[09:32] < MS-> Davbo: morning
[09:33] *** Uraeus has joined #kamaelia
[17:07] *** kamaeliabot has joined #kamaelia
[17:16] *** j_baker has joined #kamaelia
[17:31] < ms-afk> j_baker: evening :)
[17:31] < j_baker> afternoon :)
[17:32] < j_baker> For me, it's the morning though. Went to a concert last night.
[17:32] < ms-afk> cool
[17:37] *** Davbo has joined #kamaelia
[17:43] < Davbo> hi all
[17:43] < ms-afk> evening
[17:48] < Davbo> Good news ms-afk, exams are over :)
[17:56] < ms-afk> Congratulations :-)
[17:56] < ms-afk> First year ones IIRC ?
[17:56] < ms-afk> The sort where you just need to pass them, but they don't count towards anything specific ?
[17:57] < ms-afk> Beyond "you got to the next set of hurdles"
[17:57] < j_baker> Congrats Davbo
[17:58] < Davbo> Yeah, I just need to pass them
[17:58] < Davbo> should be fine
[17:59] < ms-afk> cool
[18:01] < Davbo> It wouldn't be a major problem if i did mess up really
[18:01] < Davbo> but i'm hopeful :)
[18:02] < ms-afk> OK, since 17:24:47 I've gained 90 spams.
[18:02] *** ms-afk nods
[18:06] < Davbo> Do the BBC not have any kind of anti-spam setup?
[18:06] < ms-afk> Davbo: yes, they do
[18:07] < ms-afk> however by collecting it all (and automating it) rather than just playing catchup deleting it, I'm discovering that my mail seems to be falling between the cracks
[18:07] < ms-afk> At least one system appears to be forwarding mail to another BBC server unchecked - the other one trusts it and doesn't scan it
[18:08] < ms-afk> However that only accounts for maybe 10-20 %
[18:08] *** ms-afk is now known as ms-
[18:08] < Davbo> Yeah
[18:08] < Davbo> you could try piping it through GMail?
[18:08] < ms-> My work email?
[18:08] < Davbo> Yeah
[18:09] < ms-> I very much doubt that would be a) appropriate b) work
[18:09] < ms-> I'm getting somewhere with this deletion tool though
[18:09] < Davbo> I've heard of it being done
[18:09] < Davbo> with work email
[18:09] < ms-> It's not taking anywhere near as much time to deal with now as it was
[18:09] < ms-> and I may end up with a client side greylister as a result
[18:09] < ms-> which is a very strange concept
[18:10] < Davbo> http://mail.google.com/support/bin/answer.py?ctx=%67mail&hl=en&answer=21288
[18:10] < Davbo> Could try that
[18:11] < Davbo> Actually i'm going to set that up on my university email right now :)
[18:13] < Davbo> (Obviously I don't know if it's appropriate for your needs but GMail has a very robust spam algorithm IMO would be nice if you could harness that rather than having to create tools for clearing spam)
[18:14] < ms-> No, to do that I'd need a punch a hole in the BBC all the way through to the POP3 server
[18:14] < ms-> I'd probably get sacked for that
[18:14] < ms-> which would be less than ideal :)
[18:14] < Davbo> Agreed
[18:16] < Davbo> oh well, you can't win 'em all ;-)
[18:19] < Davbo> Ah.
[18:19] < Davbo> is now parsing all my university email
[18:19] < Davbo> hmm
[18:19] *** Davbo has second thoughts
[18:21] < ms-> If you don't have to punch a hole in a firewall, its not a problem
[18:22] < ms-> If I wanted to do that, I'd have to punch holes, to the outside world, to a BBC server. That would be a very bad thing to do
[18:23] < Davbo> Yeah I imagine it would be, who knows who could be reading George Alagiah's email.
[18:23] < Davbo> ;-)
[18:25] *** ms- finds the bug he's been hunting for ages 
[18:25] < ms-> wooo
[18:32] < vmlemon_> Yay
[18:35] < Davbo> Well that's year 1 over
[18:35] < Davbo> time to delete some email
[18:35] < Davbo> University email caps at 30mb
[18:39] < j_baker> ms- could you not make a greylister that pulls the email from a pop3 server, stores it on its own pop3 or imap server and then pull your email from there?
[18:39] < ms-> j_baker: greylisting works in a rather subtle way
[18:39] < ms-> at the smtp level
[18:40] *** vmlemon_ has 81MB of mail in his GMail account, now...
[18:40] < ms-> The traditional way is a triple
[18:40] < ms-> (ip, sender, recipient)
[18:40] < ms-> This is all known before you get to the SMTP DATA command
[18:40] < j_baker> Ah, I see. So I suppose what you're saying is that if it makes it onto the first POP3 server, it's already to late?
[18:40] < ms-> So as soon as you get there, you cna make a choice
[18:41] < ms-> (yes)
[18:41] < ms-> your choice is to accept or reject
[18:41] < ms-> If you haven't seen that triple before and haven't greylisted (or whitelisted) it, you reject it
[18:41] < ms-> saying "please resend"
[18:41] < ms-> Spam systems generally don't resend
[18:41] < ms-> if you accept it, you don't get that option
[18:42] < ms-> And so it gets accepted before i have anything to do with it
[18:42] < Davbo> it can be slow though, especially dealing with email from something like GMail where getting a comprehensive list of the mail servers isn't possible afaik
[18:43] < Davbo> right ms-?
[18:43] < ms-> It can slow down your mail, but the person it's a problem for is the sender, not the reciever, generally speaking
[18:43] < Davbo> i see
[18:44] < vmlemon_> fetchHeaders.sh++
[18:44] < j_baker> Pretty interesting. How bad is it about stopping legitimate email? It seems as if there would always be a system that isn't going to resend the email.
[18:45] < ms-> About the only "legitimate" stuff that I miss is a few mailing lists (which I checked for early on) and a few mail systems which people complain about other way, which you can then put exceptions in for
[18:45] < ms-> Amazon spam mails are also rejected ironically, and I've not bothered fixing that
[18:45] < ms-> :-)
[18:46] < ms-> Number of mails rejected over the past 6-8 months or so:
[18:46] < ms-> Michael-Sparks-Computer:/var/log michaelsparks$ grep DEFER greylist.log |wc -l
[18:46] < ms-> 309388
[18:47] < ms-> That's on my server at home. Only 2 mailboxes for 2 people
[18:47] *** vmlemon_ tries to find a nicer way of doing things that having a load of lines like "wget http://www.beatjapan.org/mirror/www.be.com/developers/developer_library/headers/be/$directory/$headerfile.h"
[18:47] < vmlemon_> *than having
[18:47] < ms-> So that's something like 40,000 spams per month, 1000 a day, that I simply never see
[18:47] < j_baker> Do you have to have a static IP ms-?
[18:47] < ms-> let alone care about
[18:47] < ms-> yes
[18:47] < Davbo> I agree with the guy on LUGRadio we should just scrap email :P
[18:48] < j_baker> :( Those are too expensive (at least they are here).
[18:48] < vmlemon_> IMAP/SMTP to Jabber "gateway", anyone?
[18:48] < ms-> It's just an extra here which was a free when I signed up.
[18:48] < vmlemon_> (Seems like an interesting idea to me)
[18:48] < ms-> That was when I signed up about 7 years back mind
[18:48] < ms-> and the t&c on that haven't changed
[18:49] < ms-> If I got myself an address block from RIPE (tempting) I'd be able to get that routed to me for either very little or not much as well
[18:49] < ms-> It's ok for a home server, but for anything more it's suck since it's just a DSL uplink
[18:50] < Davbo> My networking lecturer owns a class B i think :|
[18:51] < Davbo> he's had the block for years
[18:51] < ms-> an entire class b?
[18:51] < ms-> not c
[18:51] < ms-> ?
[18:51] < Davbo> Maybe C
[18:51] < Davbo> I know it was ridiculous when he told us though
[18:51] < vmlemon_> Dunno why one person would want an entire block of IP addresses for themself...
[18:52] < Davbo> it's one that is put aside for "Corporations and Small Governments" something like that lol
[18:52] < vmlemon_> Yes, there is the argument of having a home (W)LAN, although it's overkill, even for that
[18:52] < vmlemon_> (Especially when there's the whole of 192.168.*.* (and maybe another non-routeable/"private" block) to play with)
[18:52] < Davbo> There is a use for it vmlemon_, I don't think i'm allowed to really talk about it though
[18:53] < vmlemon_> Aah
[18:53] < vmlemon_> I was talking in general, as opposed to how one specific person would use it
[18:53] < Davbo> hehe :)
[19:17] < Davbo> (after conferring, apparently my lecturer claimed his brother had a class C - now he's just lost all credibility :-))
[19:20] *** bcarlyon|laptop has joined #kamaelia
[19:38] *** mhrd is now known as mhrd-afk
[19:44] *** vmlemon_ has joined #kamaelia
[19:59] < Lawouach> burp
[20:01] *** bcarlyon|laptop has joined #kamaelia
[20:35] < Davbo> Going to watch "The Onion" film, back later.
[20:46] < ms-> oooh, excalibur is on
[20:46] < ms-> The nicest thing being it has a welsh merlin (which is about right since he would've been celtic)
[20:52] *** ms- ponders whether XMPP, pubsub, the issues with http post for uploads (resumption etc), various desires for different encryption levels and auth will lead to a need for BEEP finally
[20:53] < ms-> Always did rather look like a solution looking for a problem
[20:54] < Lawouach> BEEP was never has foxy as xmpp but it has its usage
[20:54] < Lawouach> I had meant to write bucker as an implementation of BEEP at first
[20:54] < Lawouach> Couldn't be bothered in the end :)
[20:55] < ms-> I can understand why
[20:56] < ms-> I got the book on it speculatively a few years back because I was curious, and figured it might be more explanatory about the why/how as well as the what than the RFC, and it was
[20:56] < ms-> but it did strike me as a protocol waiting for a problem
[20:56] < ms-> The book almost said as much as well
[20:57] < Lawouach> :)
[20:57] *** bcarlyon|laptop has joined #kamaelia
[20:57] < ms-> (I'm reading the POP3 RFC, which was written by Marshall Rose as well, hence why it made me think of beer)
[20:58] < ms-> beep even
[21:10] *** ms- notes that he tends to use IRC the way others use twitter
[21:10] < ms-> especially when using /me
[21:16] *** ms- ponders signing up to twitter and modifying kamaeliabot to forward these /me things to it
[21:30] < ms-> Oh, for Davbo, if he reads the logs
[21:30] < ms-> https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/trunk/Sketches/MPS/Systems/Paint
[21:30] < ms-> I'd forgotten completely about that
[21:42] *** ms- has parted #kamaelia
[21:43] *** j_baker has joined #kamaelia
[21:56] *** vmlemon__ has joined #kamaelia
[21:56] *** vmlemon__ is now known as vmlemon_
[22:02] *** vmlemon has joined #kamaelia