[05:35] *** j_baker has joined #kamaelia

[05:36] *** vmlemon has joined #kamaelia

[05:49] < Lawouach_> morning

[06:16] *** Uraeus has joined #kamaelia

[08:14] *** orphans has joined #kamaelia

[08:14] < orphans> morning all

[08:16] *** bcarlyon|laptop has joined #kamaelia

[08:47] *** Chong- has joined #kamaelia

[09:05] < Lawouach> hi orphans

[09:09] < orphans> hey Lawouach

[09:09] < orphans> how's it going

[09:12] < Lawouach> good

[09:12] < Lawouach> could use more sleep but good :)

[09:13] < orphans> cool cool. Was just about to ping you to complain about things not working actually :)

[09:20] < orphans> oh oh, but I think I've sorted it :D

[09:20] < Lawouach> :)

[09:22] < orphans> yay! Only took me about four hours to work out that I was calling a function with the wrong number of args :)

[09:23] < Lawouach> where in the C code

[09:23] < Lawouach> ?

[09:23] < Lawouach> isn't it what I asked you yesterday?

[09:24] < orphans> it was the C code calling the python code - I was chucking python the wrong number of args

[09:24] < orphans> thought it'd error out explicitly, but it just returns none and sets a flag

[09:25] < orphans> if that's what you said yesterday then sorry, I thought you were talking about a different bit of code

[09:25] < mhrd-afk> "thought it'd error out explicitly, but it just returns none and sets a flag" ... it is a dynamically typed interpreted language

[09:25] < Lawouach> :)

[09:26] < orphans> when I said error out, I really meant segfault horribly :)

[09:26] < mhrd-afk> did you consider pyrex?

[09:26] < orphans> which to be fair it does if I use a different function which does the same thing

[09:27] < mhrd-afk> heh

[09:27] < Lawouach> mhrd-afk: pyrex doesn't do binding does it?

[09:27] < mhrd-afk> I used pyrex to write the dirac bindings; plus the dvb bindings are pyrex too

[09:28] < mhrd-afk> I believe its what it was written for

[09:29] < Lawouach> okay

[09:30] < mhrd-afk> if you're getting on okay with writing directly in c/c++ then I guess probably stick with that :)

[09:30] *** mhrd-afk shuts up now and goes back to being afk

[09:30] < orphans> mhrd-afk, I'm using sip which takes some of the pain out of it

[09:30] < mhrd-afk> sip?

[09:30] < orphans> it's the bit's I'd have to manually write anyway which are fun

[09:30] < orphans> http://www.riverbankcomputing.co.uk/software/sip/intro

[09:31] < orphans> I'd never heard of it before Lawouach said about it either :)

[09:31] < Lawouach> I know sip because I'm using PyQt4 which is built using sip. I gather it's pretty good :)

[09:32] < mhrd-afk> cool. I've not seen it before either

[09:32] < Lawouach> The nice aspect is that it's a very small dependency for a project which is important.

[09:32] < Lawouach> When consider how painful SWIG may be to deploy correctly

[09:33] < orphans> aside from me being an idiot it's surprisingly nice to work with - for simple stuff it's really easy to get it up and running

[09:34] < mhrd-afk> looks to be quicker than pyrex :-) what I liked about pyrex was that it let me easily write glue code to convert a C API into something more pythonic

[09:34] < orphans> yeah, the glue stuff is I guess a bit more manual in sip

[09:35] < orphans> so for example if a function takes a std::vector you have to do the grunt work of converting a list into it before calling the function

[09:36] < Lawouach> I tend to prefer explicit over implicit so I'm fine with it :)

[09:36] < Lawouach> funny I say that in the context of a dynamic language :)

[09:36] < mhrd-afk> ah, ok, same in pyrex

[09:37] *** ian_brasil has joined #kamaelia

[10:21] < Lawouach_> back

[10:25] *** vmlemon_ has joined #kamaelia

[10:27] < vmlemon_> Hi

[10:38] < Lawouach_> hi

[10:40] *** Davbo has joined #kamaelia

[10:41] < vmlemon_> Hi Davbo

[12:01] < orphans> Lawouach, ping?

[12:05] < Lawouach_> pong

[12:05] < orphans> you got a moment?

[12:06] < Lawouach_> yes

[12:07] < orphans> ok, I think I'm kinda close with getting sound out, but am lost a bit in where the memory is allocated and deallocated

[12:08] < orphans> hang on, and I'll svn ci - should make it a bit clearer

[12:09] < Lawouach_> I won't be able to svn up at work though but I'll be able to read through the code online

[12:09] < orphans> mhm, just getting links for you

[12:10] < orphans> http://code.google.com/p/kamaelia/source/browse/trunk/Sketches/JT/RtAudioBindings/RtAudio.sip - the function streamCallback is the important bit

[12:11] < Lawouach_> k

[12:11] < Lawouach_> what about it?

[12:11] < orphans> so for the output basically it gets a pointer to a buffer

[12:12] < orphans> then it gets a pointer to another buffer (which was a numpy array) from python in the PyEval_CallFunction bit

[12:13] < orphans> and then it sets the pointer which it got at the start to be equal to the pointer to the buffer created in python

[12:14] < orphans> but after the function has been called the original pointer just gives zeros

[12:14] < orphans> so I think what happens is the buffer created in python get's deallocated - does that sound sensible?

[12:15] < Lawouach_> possible, you could look at the gc module for getting more feedback about that

[12:15] < Lawouach_> careful as it is really verbose :)

[12:15] < orphans> ok, didn't know I could do that :)

[12:16] < orphans> if that was the case what would be the right thing to do? Would I have to do something like memcpy the data in the python buffer over to the original buffer?

[12:16] < Lawouach_> I'm not sure to be honest. I don't do bindings :)

[12:16] < Lawouach_> Either that or increase the refcount on the original object

[12:16] < Lawouach_> so that it isn't deallocated

[12:16] < Lawouach_> Not sure really

[12:17] < orphans> ok, I'll carry on playing :)

[12:17] < orphans> the ideas sound good though :)

[12:18] < Lawouach_> you learn the inside of Python there :)

[12:18] < Davbo> Woo! code.google checkout finished on all the Branches, it only stalled 4 times.

[12:19] < Davbo> Memory allocation fun orphans?

[12:20] < orphans> of course :)

[12:20] < Davbo> Could be worse...

[12:21] < Davbo> could be ASM

[12:21] < Davbo> ;-)

[12:22] *** orphans doesn't laugh :D

[12:28] < orphans> I hear noise! The wrong noise, but noise none the less :D

[12:32] < Davbo> it's all good

[12:34] *** mhrd-afk congratulates orphans

[12:34] < Davbo> Right I have one svn thing for branches and one for the trunk, can I svn cp between them?

[12:35] < Davbo> I'll just try and see what happens

[12:37] *** mhrd-afk thinks it'll probably work

[12:38] < Davbo> looks like it worked

[12:42] < Lawouach_> orphans: nice job mate

[12:44] < orphans> now to make it make the sound I want :D

[12:47] < vmlemon_> It comes pre-circuit bent? ;)

[13:39] < vmlemon_> Hah, just got the e-mail pasted to http://pastebin.org/58183 a few moments ago...

[13:39] *** vmlemon_ wonders how that scheme will turn out

[14:39] *** vmlemon_ has joined #kamaelia

[15:12] < vmlemon_> Hmm, http://joi.ito.com/weblog/2003/05/13/sputniks-rock.html, a Wi-Fi router with Jabber/XMPP support...

[15:25] *** vmlemon_ has joined #kamaelia

[15:40] *** Lawouach has put on the fantastic Urban Dance Squad - Persona non grata

[15:40] < Lawouach> great rock album

[15:41] < Lawouach> vmlemon_: interesting

[15:51] *** Davbo is listening to Led Zeppelin IV (again :-))

[15:58] < Davbo> via reddit: http://ubuntuforums.org/showthread.php?t=863049 the employer who set the questions replies on the second page

[16:15] *** bcarlyon|laptop has joined #kamaelia

[18:40] < vmlemon_> Hmm, anyone thought of building a speech-to-text system, or a binding to one with Kamaelia?

[19:07] < j_baker> I think Michael was working on one.

[19:08] < vmlemon_> Aah

[19:20] < Lawouach> copying files over network using Nautilus is so frustrating

[19:20] < Lawouach> file transfers just stop randomly

[19:21] < Lawouach> so KDE4 is buggy and Gnome is not getting any better

[19:21] < Lawouach> time for some Window Maker fu!

[19:22] < vmlemon_> smbmount and cp -v is so much nicer, sometimes...

[19:22] < vmlemon_> That, or KDE 3.5.x's NFS and SMB stuff

[19:22] < Lawouach> it's a ftp server

[19:22] < vmlemon_> Aah

[19:23] < vmlemon_> Haven't tried the client in Nautilus or KDE 4 :|

[19:39] *** MS- has joined #kamaelia

[19:39] < MS-> evening

[19:39] < vmlemon_> Hi MS-

[19:41] < MS-> Thanks for the posts on the list. Regarding the file appender, the idea situation is to have a contributor agreement in place, since they tend to eradicate potential problems later on (in the case of a license breach primarily). It'll also mean I can add you to the contributors list :-)

[19:41] < MS-> s/idea/ideal/

[19:41] < vmlemon_> No problem

[19:41] < vmlemon_> I take it that the code is pretty much sane, as far as being a component is concerned?

[19:42] < vmlemon_> (i.e. doesn't need any additional 'imports' or cleanup code)

[19:42] < vmlemon_> Although that's my 2p

[19:43] < MS-> Yep, it looks pretty good. I'll probably edit it slightly once it's in, since there's almost always improvements that can be made, but I'll ensure the behaviour remains as you'd expect.

[19:43] < vmlemon_> It does have some bugs, although I'm not sure how I can fix them :(

[19:43] < MS-> One way is to declare what you think the bugs are and request solutions :-)

[19:43] < vmlemon_> Mostly relating to the order stuff gets written in, if you attach it to an inbox and an outbox

[19:44] < vmlemon_> The code looks too simple to me though to have any sorts of visible problems...

[19:44] < vmlemon_> Although I don't know much Python

[19:44] < MS-> In which case it's probably a behavioural thing which is more high level

[19:44] < vmlemon_> (So there's probably stuff that's glaringly obvious to others as faults)

[19:44] < vmlemon_> It has served me fairly well, though

[19:44] < MS-> Explaining what you think should happen clearly, and what actually happens generally solves 90-99% of problems IME

[19:45] < MS-> :)

[19:45] < vmlemon_> Considering that it's based on sample code wrapped in how I thought a Kamaelia component would look...

[19:48] < vmlemon_> Not sure about the state of it licensing wise, though

[19:48] < vmlemon_> (Since it was lumped into a directory with a MIT-licensed file I wrote, and has no headers itself)

[19:48] < vmlemon_> I consider it to be unlicensed/BSD licensed, though

[19:48] < vmlemon_> for all extents and purposes

[19:48] < vmlemon_> Although it's better for everyone (as I mentioned in my 'mails), that I go through the motions and have it properly licensed

[19:48] < vmlemon_> (Seeing as signing the CA is presumably a once-in-a-lifetime event)

[19:50] < MS-> Signing the CA is a one off, however, you apply it to code by explicitly declaring it to be covered by that CA. eg something like:

[19:50] < MS-> # Licensed to the BBC under a Contributor Agreement: RJL

[19:50] < MS-> It's the most lightweight thing I could think of/get the legal people to agree to :)

[19:52] < MS-> Unlike many contributor agreements, but identically to Python's the code remains yours. It's just an explicit grant of usage

[19:57] < vmlemon_> Hmm, the 2 or 3 letter "tag" is the same as used on CVS/SVN?

[19:57] < vmlemon_> (Assuming you have access to it/a working area)

[20:01] *** vmlemon_ has joined #kamaelia

[20:02] < vmlemon_> Damn, just knocked the USB cable out my phone and got disconnected

[20:04] < vmlemon_> It's essentially a "cover our backs"/co-ownership thing, I take it?

[20:07] < MS-> "2 or 3 letter "tag" is the same as used on CVS/SVN" - yep - it also landgrabs you Kamaelia.Community.tag. The tag is usually initials.

[20:07] < MS-> It's primarily a cover everyone's backs thing

[20:09] < vmlemon_> Looking at the contributing page now

[20:11] < vmlemon_> I take it that the next step is to ask on the Group for a form?

[20:12] < vmlemon_> Argh, stupid Firefox

[20:12] < vmlemon_> It'd be nice if it could render a page without locking up and blanking the window :|

[20:14] < vmlemon_> Or do I use the one on /Developers/ContributorAgreement or /Developers/SampleContributorAgreement, after changing the bits specified?

[20:16] < vmlemon_> Meh, looks like /Developers/ContributorAgreement has no content

[20:19] < MS-> "Or do I use the one on /Developers/ContributorAgreement or /Developers/SampleContributorAgreement, after changing the bits specified?"

[20:19] < MS-> You use the one on /Developers/SampleContributorAgreement after changing the obvious bits :)

[20:20] < vmlemon_> I can do that with a text editor and then print it and sign it? Or do I need to handwrite those bits for any particular legal reason? (Excepting my signature)

[20:22] < vmlemon_> I fill the green bits in?

[20:23] < MS-> If you want to handwrite it in copperplate, be my guest, but the thing that matters is the signature should be handwritten - the rest is normally printed/typed/etc :)

[20:23] < MS-> essentially yes

[20:23] < vmlemon_> OK

[20:23] < vmlemon_> Excepting the tag, the purple bits are OK as-is?

[20:25] < vmlemon_> Or is that just an example, even in the appended version?

[20:25] < vmlemon_> (the '"Licensed to the BBC under a Contributor Agreement: JLB"' bit)

[20:26] < vmlemon_> Seeing as JLB is Joe in the example...

[20:26] < aa_> MS-: yay! http://www.pyconuk.org/talk_abstracts.html#999

[20:28] < vmlemon_> Cool

[20:28] < vmlemon_> You got accepted to do a talk?

[20:30] < MS-> aa_: :-)

[20:31] < aa_> vmlemon_: yeah

[20:31] < aa_> I think MS here is one of the bosses of pyconuk

[20:31] < vmlemon_> So I read on the talks page...

[20:33] < aa_> vmlemon_: you going to be there?

[20:33] < vmlemon_> Unfortunately not :(

[20:34] < aa_> aww, it's so good

[20:34] < vmlemon_> I'd like to go though... one day

[20:34] < aa_> it's like returning to the mothership

[20:35] < vmlemon_> It sounds interesting though

[20:35] *** vmlemon_ wonders if there's going to be another Mashed or similar, next year

[20:37] < MS-> aa_: Yep, I'm on the committee for Pycon UK - I've known John for many years now, so when he was first looking for help I offered :)

[20:37] < MS-> vmlemon_: It wouldn't surprise me, this year went really well

[20:37] < MS-> But dunno yet :)

[20:38] < vmlemon_> Hmm, wasn't there another event? Although it's name escapes me now

[20:38] < MS-> new boss and all that - but this new boss seems to really get the things that are seem to be important from a devs perspective at present - open source, open standards, developer communities/events etc

[20:40] < aa_> MS-: good for you, it was organized beautifully last year

[20:41] < vmlemon_> Shame PyCon is so expensive, although you do get a fair amount of stuff included in the price...

[20:41] < aa_> vmlemon_: our pycon or the big pycon?

[20:41] < vmlemon_> (Although that's just my view, as a currently unemployed ex-college student at the moment)

[20:41] < vmlemon_> the UK one

[20:42] < aa_> vmlemon_: to be honest, I think 65£ early-bird rate is an absolute bargain

[20:42] < aa_> costs more to fill my bloody car up with petrol

[20:42] < vmlemon_> Ouch

[20:42] < aa_> but yeah, unemplyed ex-college I might think differently

[20:43] < MS-> vmlemon_: It's about the lowest break even cost tbh

[20:43] < vmlemon_> I'm sure getting a train to Birmingham or wherever it is might be fun, that said

[20:43] < vmlemon_> Aah

[20:43] < MS-> The £65 includes meals

[20:44] < MS-> Lunch on sat/sun, evening dinner on saturday. The etap is pretty basic, but nice, and the bookings through that last year went flawlessly

[20:44] < vmlemon_> Didn't it include a subsidised hotel room or something like that, for those that booked early enough?

[20:45] < vmlemon_> (At least for those that paid a little extra for the privilege)

[20:45] *** vmlemon_ can never seem to spell that word, for some reason

[20:46] < vmlemon_> Of course, I'm not complaining, or yelling "give us a free conference", since that would be pointless...

[20:46] < vmlemon_> (Not to mention that it does cost to provide the thing in the first place)

[20:47] < MS-> John foraged round Birmingham and found the etap, and I think it's a negotiated rate

[20:47] < vmlemon_> Aha, £11.67 gets you a room shared between two others...

[20:47] < vmlemon_> So the FAQ reckons

[20:48] < MS-> You can never please everyone really - the closest you get to completely gratis is bar camps, but even then people have travel & accommodation

[20:48] < vmlemon_> Although I don't know if that's a rotation of any 2/3 people, or 3 people in a group

[20:48] < vmlemon_> Aha, that's the one I was thinking of

[20:48] < MS-> mashed tried to deal with that with buses & stuff, but the busses were insane times

[20:51] < vmlemon_> Still, it's a long way off yet

[20:51] < vmlemon_> (September)

[20:51] < aa_> MS-: I really plan to use the open space this year

[20:55] < aa_> MS-: I think something like "bring your cool dev tools, and we'll plug them into an IDE"

[20:55] < aa_> not sure if that is the right theme

[21:02] < MS-> That's a good use of open space IMO :)

[21:39] *** vmlemon_ registers for the Backstage mailing list

[21:40] < vmlemon_> kamlogbot: dance

[21:40] Reply: does the macarena

[22:17] < vmlemon_> Night

[22:51] *** vmlemon_ has joined #kamaelia