[00:50] *** bcarlyon|laptop has joined #kamaelia

[01:06] *** bcarlyon|ubuntu has joined #kamaelia

[05:59] *** bcarlyon|laptop has joined #kamaelia

[06:15] *** bcarlyon|ubuntu_ has joined #kamaelia

[06:53] *** vmlemon__ has joined #kamaelia

[06:53] *** vmlemon__ is now known as vmlemon_

[06:53] < vmlemon_> Hi

[08:20] *** mhrd-afk is now known as mhrd

[08:24] *** orphans has joined #kamaelia

[08:27] *** mhrd has joined #kamaelia

[08:48] < Lawouach_> morning

[08:55] < orphans> morning Lawouach_

[08:58] *** mhrd has joined #kamaelia

[09:09] < Lawouach_> morning orphans

[09:09] < Lawouach_> feeling better?

[09:10] < orphans> a bit tired but not too bad ta

[09:10] < orphans> slightly lost in my own code at the moment too, but that's my fault :)

[09:11] *** MS- has joined #kamaelia

[09:11] < MS-> morning

[09:12] < orphans> hey MS-

[09:23] < Lawouach_> hi MS-

[09:24] *** Lawouach_ will have to look at the mid-term eval this week closely

[09:24] *** Lawouach_ has the mighty power

[09:26] *** Uraeus has joined #kamaelia

[09:26] < orphans> :o

[09:30] < MS-> Quote of the day: "don't tell me vista is popular because X million users have it. ... lots of people have cancer, both through not having a choice"

[09:32] < Lawouach_> ouch :)

[09:55] *** Davbo has joined #kamaelia

[09:59] *** MS- randomly breaks Pablo's branch (removal/replacement of LikeFile.py - which simply should not even be used any more due to deletion from /trunk)

[09:59] < MS-> He's writing a test framework and sould know better

[10:02] < MS-> It's worse in that he actually appears to have branched off /trunk and added it back in

[10:09] *** MS- adds in evilness

[10:16] < Davbo> morning all

[10:16] < MS-> Anyone know how to use ctypes to cause a segfault?

[10:17] < MS-> Davbo: morning

[10:31] *** MS- thinks this is pretty evil:

[10:31] < MS-> https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/branches/private_PO_Tests/Axon/Axon/LikeFile.py

[10:33] < Davbo> umm

[10:35] < MS-> He shouldn't be using it and knows it.

[10:35] < Davbo> ah

[10:35] < MS-> "Adding private_MPS_Scratch/**/LikeFile.py to the testing branch. Although I should use Handle instead of LikeFile, "

[10:35] < Davbo> I thought you'd just gone crazy

[10:36] < MS-> No :)

[10:41] < MS-> It's a bit harsh, but he's not actually doing anything I've asked so far

[10:42] < MS-> At the moment he appears to just be writing a clone of the "Planet" aggregator code and assuming that in doing so he'll invent something useful for testing things that aren't his KamPlanet code

[10:43] < MS-> Probably is, that is incredibly unlikely to be useful for, for example, testing random graphlines webserver/client systems

[10:43] < MS-> testing input/output with devices (midi, pygame)

[10:43] < MS-> testing code components (dirac, vorbis)

[10:43] < MS-> etc

[10:44] < MS-> Which is his project was "Build a clone of 'Planet'" would be OK

[10:44] < MS-> but it isn't

[10:45] < Davbo> Huh, a "Planet" as in a sort of Python Planet ? the blog aggregating stuff?

[10:45] < orphans> ew, just remembered I need to do a load of work on MIDI support

[10:45] *** orphans alters schedule

[10:45] < MS-> Davbo: yep

[10:46] < MS-> It's Ok as a test for a 2-3 week period (max)

[10:46] < MS-> But to not talk to _anyone_ else even when prompted repeatedly about what they actually do/need is bizarre

[10:47] < Davbo> Yeah

[10:49] < Davbo> LugRadio are finishing up after this LugRadio live later this month MS- (thought you might have been interested)

[10:50] *** Davbo is just listening to todays episode where they announce it

[10:50] < MS-> finishing up?

[10:50] < Davbo> Yeah they're stopping the show

[10:50] < MS-> really?

[10:50] < Davbo> because they don't want to become Spandau Ballet, or so they say lol

[10:50] < Davbo> Yeah

[10:51] < MS-> heh

[10:51] < MS-> cool

[10:53] < MS-> https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/branches/private_PO_Tests/Axon/Axon/LikeFile.py

[10:53] < MS-> -- will now cause a segfault and crash

[11:29] < Lawouach_> back

[11:33] < orphans> quick question - is there any sort of dirty hack I can do to get a component to start running it's main loop a second or two after the rest of the components??

[11:33] < orphans> s/??/?

[11:33] < orphans> s/it's/its

[11:33] < orphans> I fail at writing

[11:35] < Lawouach_> I don't recall any such parameters

[11:35] < Lawouach_> I assume you could put a time.sleep() at the top of the main()

[11:36] < orphans> Ahh yeah, hadn't thought of that. it should do the trick for me

[11:39] < orphans> woop, syncing code finally works :)

[11:51] *** mhrd is now known as mhrd-afk

[11:55] *** Davbo may have to go to LugRadio Live now it's the last one

[11:58] < Lawouach_> MS-: I took the liberty of replying to Pablo

[12:02] *** Davbo hopes Pablo gets stuff sorted

[12:05] < Lawouach_> I hope so too of course.

[12:06] < Lawouach_> That's why i'm insisting so much for GSoC student to consider their commitment as a work placement. A project does expect a certain level of work and result.

[12:07] < Lawouach_> If Pablo was doing stuff on his own outside GSoC things would be easier.

[12:10] < Davbo> Yeah#

[12:12] < Davbo> Humm, should

[12:12] < Davbo> Graphline (

[12:12] < Davbo> blah phone

[12:15] < Davbo> Ahaa.

[12:16] < Davbo> MS-: That code works with a Graphline but not ProcessGraphline

[12:17] < Davbo> For some reason

[12:20] < Davbo> http://davbo.pastebin.com/m4c9f30bd Works

[12:20] < Davbo> http://davbo.pastebin.com/me02e211 Doesn't work

[12:29] < MS-> ProcessGraphline is currently not yet supported

[12:30] < MS-> Though if you want to look at the code, you're more than welcome to do so :-)

[12:30] < MS-> that's a useful test case though - thanks :-D

[12:32] < Davbo> Just letting you know

[12:33] < MS-> much appreciated

[12:33] < MS-> :)

[12:33] *** MS- stressed

[12:33] < Lawouach_> how so?

[12:37] *** MS- certainly doesn't take offense FWIW

[12:38] < Lawouach_> cool

[12:38] < MS-> Oh, there's a meeting today, bit behind on work stuff, meeting is an hour earlier than expected, and someone phoned up for a pre-meeting meeting - when an email would've been quicker/more async to deal with - the usual sort of work crap

[12:38] < MS-> :)

[12:39] < MS-> nothing major

[12:39] < MS-> Nothing a good cup of coffee won't cure later on :)

[12:39] < MS-> I thought it worth mentioning because my comment to Davbo looked curt, and wasn't the intent

[12:39] < MS-> :)

[12:39] < Lawouach_> :)

[12:40] < MS-> Pablo's mail is far too long for me to read at the moment as a result, so just responded to the likefile comment

[12:41] *** Davbo is looking at hotels in Wolverhampton :-)

[12:41] *** MS- stayed at the premier inn (nee premier travel inn, nee travel inn) last year

[12:41] < MS-> was nice

[12:41] < MS-> But then I like those places

[12:42] < MS-> what date is it?

[12:42] < Davbo> Dunno if i'll go, never been to one of these things.

[12:42] < Davbo> 19-20 July

[12:42] < MS-> You should definitely go

[12:43] < MS-> Or to Pycon :)

[12:43] < MS-> (pycon uk)

[12:43] < MS-> Since they're community organised events and generally very cool

[12:44] < MS-> If it wasn't for something happening later that week, I'd be definitely going without thinking. (not sure at this instant if I *can* go basically)

[12:44] < Davbo> I'm gonna go to PyCon but not LugRadio is ending I want to go to this too :-)

[12:44] < MS-> But that's more about timing rather than anything else :)

[12:44] *** MS- nods

[12:45] < Davbo> s/not/now

[12:48] *** MS- tods

[12:57] *** Trun has joined #kamaelia

[12:57] < Trun> hi all

[13:00] < MS-> Afternoon Pablo. I'll need to delay our meeting until 3 since I'm expected to be on a phone conf between 2 & 3

[13:00] < MS-> (localtime is 2pm now)

[13:00] < Trun> ok

[13:01] < MS-> if the person who organised it gives me a number to call that is or phones me

[13:01] < MS-> (still waiting)

[13:01] < MS-> I haven't had a chance to read your long reply justifying stuff

[13:02] < MS-> I'm not particularly interested in justifications though - I am interested in what your project is doing , why it's doing things in a certain way (the most important bit) and where you intend going next (and why)

[13:03] < MS-> The *why* has always been implicit and I never realised I needed to spell it out

[13:03] < MS-> but have been posting to you regularly asking for informtaion, asking you to start talking to the rest of us etc

[13:03] < MS-> And nada, zip, zilch

[13:04] < MS-> I was primarily warning you that a time would come when I stop asking because it's having no effect

[13:04] < MS-> By comparison, chong was having similar problems with his checkin messages (which is why I emailed both of you)

[13:04] < MS-> and he's improved

[13:04] < MS-> Dramatically so in fact

[13:05] < MS-> which is much more fun and useful for both of us

[13:06] < MS-> For example, I have no indication where you are relative to your public schedule here:

[13:06] < MS-> http://edit.kamaelia.org/Developers/Projects/TestingFramework

[13:07] < MS-> And its scary: "Freeze the testing framework API"

[13:07] < MS-> in what, 3 days?

[13:07] < MS-> With *zero* discussion with the users of the framework?

[13:07] < MS-> That's pointless

[13:08] < Trun> yeah, I know

[13:09] < MS-> You may wish to read todays log as well

[13:09] < MS-> Since you don't bother hanging around on IRC

[13:09] < MS-> http://yeoldeclue.com/logs/kamaelia2008-07-01_log.html

[13:09] *** MS- feels like he's wasting his time

[13:10] < MS-> (doesn't help that still haven't been phone by the people running the meeting)

[13:11] < Trun> (I'm reading the logs)

[13:12] < Trun> done

[13:12] < MS-> You'll find that the version of LikeFile in our branch will now bomb out

[13:12] < Trun> yeah, I saw it in the commits list

[13:12] < MS-> It was deleted from /trunk for good reasons

[13:13] < Trun> yeah, I know

[13:13] < Trun> I just thought that I could start writing tests if I added it temporarily

[13:13] < MS-> You could use handle

[13:13] < MS-> That works

[13:13] < Trun> since last week I had problems using Handle

[13:13] < MS-> The examples in the distribution were rewritten

[13:13] < MS-> Specifically to use Handle

[13:13] < Trun> yeah, I saw the examples

[13:13] < MS-> In which case you should post the problems to the list

[13:14] < MS-> or discuss on IRC

[13:14] < Trun> but still, the way it works is different

[13:14] < MS-> Yes

[13:14] < MS-> It is

[13:14] < MS-> Slightly

[13:14] < MS-> not radically

[13:14] < MS-> slightly

[13:14] < Trun> yeah, I should have asked in the list

[13:14] < MS-> This isn't a problem - it was the first release with the code in.

[13:15] < MS-> Anyway, at present I can't use your testing framework, and see no reason to as a result

[13:16] < MS-> And we're 6 weeks in

[13:16] < Trun> last week I couldn't get KamTestCase working with it. That's why I commited a KamTestCase "using it" but broken, and then I commited the old version again

[13:16] < MS-> 5 and a bit

[13:16] < MS-> No, I mean more fundamentally

[13:16] < MS-> you've not bothered documenting it

[13:16] < Trun> why do you say you can't use it?

[13:16] < MS-> you've not bothered creating small examples

[13:16] < Trun> you mean no documentation?

[13:16] < MS-> (14:23:29) Trun: why do you say you can't use it?

[13:16] < MS-> (14:23:23) MS-: you've not bothered documenting it

[13:16] < MS-> (14:23:30) MS-: you've not bothered creating small examples

[13:17] < MS-> You do realise that without any docs you wouldn't stand a chance of using any part of kamaelia?

[13:17] < MS-> You've created a branch for unknown reasons

[13:18] < MS-> You've not explained anywhere AFAICT what the code in where does what

[13:18] < Trun> I did a simple sample for it: https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/trunk/Sketches/PO/KamPlanet/test/test_SimpleSample.py

[13:18] < MS-> (IRC Is not appropriate)

[13:18] < MS-> Oh right

[13:18] < MS-> How silly of me to miss that

[13:18] < MS-> ~/kamaelia/trunk/Sketches/PO> find |grep py$|wc

[13:18] < MS-> 48 48 1568

[13:18] < MS-> 48 files

[13:18] < Trun> and I thought that the tests of kamplanet were good "samples"

[13:18] < MS-> one example

[13:19] < MS-> Trun: That's like me saying the Whiteboard code is a good example of how to use a backplane

[13:19] < Trun> ok

[13:20] < MS-> 12,225 lines

[13:20] < MS-> ~/kamaelia/trunk/Sketches/PO> wc `find |grep py$`

[13:20] < Trun> I created the branch trying to do what you told me to do in an e-mail to the list

[13:20] < MS-> Why not actually TALK then?

[13:20] < MS-> Seriously

[13:20] < Trun> * After you're *relatively* happy that you can test kamplanet "enough", I

[13:20] < Trun> would suggest that you move on to creating a branch[1] and working in:

[13:20] < Trun> /branches/< yourbranch>/Kamaelia/Kamaelia/Test

[13:21] < MS-> Subject?

[13:21] < Trun> "Testing Framework: current state report"

[13:21] < Trun> 11 Jun 2008 10:32:11 +0100

[13:21] < MS-> OFFS

[13:21] < MS-> nearly 3 weeks ago?

[13:22] < Trun> yeah, I started testing HTTP in the KamPlanet test folder

[13:22] < MS-> Why?

[13:22] < MS-> That is the SINGLE most important question

[13:22] < Trun> then I moved to the new branch to add it "more properly" to then add more tests of Kamaelia.Chassis.* and so on

[13:22] < MS-> That you always need an answer for

[13:22] < MS-> I mean that was 3 *week*s ago

[13:23] < MS-> I give up

[13:23] < Trun> because I didn't want to spend time creating the branch back then, and I waited as soon as I had tested something else (not KamPlanet)

[13:23] < MS-> you either build the project you proposed or you don't

[13:23] < MS-> If you don't I'll not bother with any of this

[13:24] < MS-> You need to read up on communications

[13:24] < MS-> and learn some communications skills

[13:24] < Trun> yes, I agree

[13:24] < MS-> rather than coming up for justifications for pissing around with kam planet

[13:24] < MS-> I'm not the most diplomatic of people in the world

[13:24] < Trun> ok, and I really want to move on from kam planet

[13:24] < MS-> To say the least

[13:25] < MS-> But I do aim for clarity

[13:25] < MS-> I'd like your project to be useful to the rest of us

[13:25] < Trun> i also want it to be useful

[13:25] < MS-> That's why you were accepted this year, and because you'd proposed it the previous year and it showed improvment

[13:25] < MS-> However at present you're just creating stuff that's of no use to anyone

[13:26] < MS-> The worlds best system is no use if no-one understands how to use it or why they should

[13:26] < MS-> And you'll just get upset and frustrated at the end of the summer if you end up in that situation

[13:26] < Trun> I know

[13:26] < MS-> 12 weeks, 3 months, and a minimum of 12,225 lines (since that's what you've done so far) and you'll be left with no-one knowing how to use it

[13:27] < MS-> That's perhaps the single harshest lesson we all learn from open source

[13:27] < MS-> that reality sucks :)

[13:27] < Trun> yeah

[13:27] < MS-> But also that communication is important

[13:27] < MS-> So, you can either carry on on your current path

[13:28] < MS-> or you can rummage around and find out how we have already been doing tests for the past 5 years

[13:28] < MS-> (because we do and have

[13:28] < MS-> )

[13:28] < MS-> And work with the project

[13:28] < MS-> or work in a quiet dark corner and have a very well tested kam planet

[13:28] < MS-> I was nearly that blunt 2 weeks ago

[13:29] < Trun> ok, I get it

[13:29] < MS-> k

[13:29] < MS-> good

[13:30] < MS-> I get to choose whether to cut my losses shortly, and would rather us have a disagreement now (where you probably think I'm annoying :) ) and end up with you happier at the end of the summer and something we can all use

[13:30] < MS-> than to just let it drift

[13:31] < MS-> But bear in mind that I didn't post this link to you : http://blog.red-bean.com/sussman/?p=96

[13:31] < MS-> for no reason

[13:31] < Trun> yeah, I have it printed on top of my laptop

[13:31] < MS-> :)

[13:31] < Trun> s/laptop/desktop table/

[13:31] < MS-> I actually think CS courses are to blame for this behaviour BTW

[13:32] < MS-> ie something you've learnt rather than how you are - since it seems to be a problem with more experienced people

[13:32] < MS-> So in some respects it's unlearning and going back to basics

[13:32] < Lawouach_> They certainly don't encourage for communication across student projects (at least in many instances in the Uni I went to)

[13:33] < MS-> Lawouach_: Indeed. They also heavily encourage the "don't show anyone anything until it's perfect" mode of behaviour as well

[13:33] < MS-> The combination is a real pain

[13:35] < Lawouach_> indeed

[13:35] < Lawouach_> Hi Pablo BTW :)

[13:35] < Trun> Hi :-)

[13:35] < Trun> (answering your e-mail)

[13:36] < MS-> Trun: Just trying to say, this isn't necessarily just your natural behaviour - it may be reinforced by training

[13:36] < MS-> The manta of TDD is after all - "what's the minimal thing needed to make the test pass"

[13:37] < MS-> The minimal thing in TDD terms for creating a test framework (which is a little meta) would be "to collect all the existing tests in one place and make them run"

[13:37] < MS-> Then "to extract out the code that makes the run"

[13:37] < MS-> It also has the fringe benefit of being useful immediately

[13:38] < MS-> an "a priori" implementation would perhaps be useful

[13:38] < Trun> yeah, I really want to change the behaviour. I'm not used at all to this kind of communication etc. (my advisor doesn't even know python; and in the lab we all tend to work on our own most of the time)

[13:39] *** MS- nods

[13:39] < MS-> I guessed that

[13:39] < Trun> (no review, I usually don't even know in what my labmates are exactly working on)

[13:39] < MS-> But the point is watch the mailing list, and copy the beahviours

[13:39] < MS-> I think that's a work thing as well

[13:40] < Trun> yeah

[13:40] < MS-> I've got no idea what my colleague who is sharing my office is actually working on

[13:40] < MS-> :)

[13:40] < MS-> Looks fun though

[13:40] < MS-> Involves cameras and finding stuff in images AFAICT

[13:41] < MS-> But the difference is that your code should be used by everyone in the project sooner or later - if its done even half right

[13:41] < MS-> Even if its just for core sections

[13:42] < MS-> Anyhow, I need you to go away, think about what you're doing, how you're doing it, and *why* you're making certain decisions

[13:42] < MS-> The why is very important

[13:43] < Davbo> MS-: quick question

[13:43] < MS-> Davbo: go

[13:43] < MS-> :)

[13:43] < Davbo> can you point me at the docs for DisplayConfig(width=270, height=600)

[13:43] < Trun> ok

[13:44] < Trun> prior to all this stuff, I was going to keep testing more code even with the "old" LikeFile.py, which is obviously wrong

[13:45] < MS-> Davbo: do you mean this: http://edit.kamaelia.org/Components/pydoc/Kamaelia.UI.Pygame.Display.html

[13:46] *** mhrd-afk is now known as mhrd

[13:46] < Trun> should I focus on documenting and making the testing framework ready for other uses with current Handle?

[13:46] < MS-> Trun: I'd focus on building something that we can actually use

[13:47] < Trun> ok

[13:47] < MS-> That is after all the core

[13:47] < Trun> yep

[13:47] < MS-> Now that Chong for example has his code to the stage it can be used, he's gotten feedback which is useful

[13:47] < MS-> It doesn't have to be pretty (helps)

[13:47] < MS-> Doesn't have to be large (large hinders)

[13:47] < MS-> But actually usable would be great

[13:47] < Lawouach_> I'll be honest but personally I would be more interested in seeing one dummy Axon component under test than a complete application like KamPlanet. At least on the first stage.

[13:48] < MS-> Ditto

[13:48] < Trun> yeah, I was answering the e-mail just a minute ago

[13:48] < MS-> I suspect also that given the amount of effort already expended that should be a relatively easy task

[13:48] < MS-> Trun: Rule #1 of meetings - you're either in it or not

[13:48] < Trun> yeah, I'm in

[13:48] < MS-> If you're in it, and doing emails, the meeting is pointless

[13:49] < Trun> ok

[13:49] < MS-> If you're doing emails then you shouldn't be in the meeting (that's a general opinion I have with all meetings tbh)

[13:50] < MS-> Anyway, that's actually the point of the examples director

[13:50] < MS-> y

[13:50] < Trun> ok, sorry for that (I haven't written even three lines, I started when I saw the e-mail before we talked)

[13:50] < MS-> all the examples there are small (generally) to the point of triviality

[13:50] < MS-> perhaps too trivial really

[13:51] < MS-> but they serve as basic intros

[13:52] < MS-> Have you read test driven design BTW/

[13:52] < MS-> ?

[13:52] < MS-> By Kent beck?

[13:52] < Trun> you mean the kent beck book?

[13:52] < MS-> I keep giving my copy away to people

[13:52] < MS-> to encourage better code

[13:53] < Trun> yeah, I also have it on my desktop

[13:53] < MS-> cool

[13:53] < Trun> yeah, I really like Kent Beck and Martin Fowler books

[13:53] < MS-> just that one

[13:54] < MS-> no, looks like I've given my one away (again)

[13:54] < Trun> :-)

[13:54] < MS-> Anyway, the point about it is that you'll not he spends about a third of the book building up how TDD systems work

[13:54] < MS-> and building a testing framework to go with it

[13:54] < Trun> the "refactoring: improving the design of existing code" by Martin Fowler (and contributions of Kent Beck) is also really really nice

[13:54] < MS-> deriving the latter from the former

[13:55] < MS-> May get work to buy that one

[13:55] < MS-> anyway, the point is that it starts small

[13:55] < MS-> and in his case it was with something dummy

[13:55] < MS-> but actually in the context of really recommending something real

[13:56] < MS-> It's worth considering

[13:56] < MS-> You could for example declare everything to date experimentation and effectively restart, but from a much greater knowledgebase/starting point

[13:57] < MS-> But progress rapidly by pulling in stuff that DOES do what you need/want

[13:57] < MS-> That DOES match with how we actually test things and assists with automating them

[13:58] *** Davbo feels like an idiot

[13:58] < Trun> hmm... I hadn't even considered it

[13:58] < Davbo> found what i wanted MS-

[13:58] < Davbo> :-)

[13:59] < MS-> Trun: We often don't

[13:59] < MS-> Also for example

[13:59] < MS-> https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/trunk/Sketches/MPS/GSOC08/MultiPipeTest.py

[13:59] < MS-> is specifically created for testing multi process stuff

[13:59] < Trun> do you think it would worth it? you mean by copying some code fragments when useful?

[14:00] < MS-> now I'm not saying you should make *that* test suite run automagically

[14:00] < MS-> but you'll note that it does this:

[14:00] < MS-> if 0: # This stuff works

[14:00] < MS-> < test>

[14:00] < MS-> if 0: # Basic test of Source/Expecter

[14:00] < MS-> < another>

[14:00] < MS-> if 1: # Basic test of Source/Expecter

[14:00] < MS-> < another>

[14:00] < MS-> if 0:

[14:00] < MS-> < another>

[14:01] < MS-> which isn't uncommon

[14:01] < MS-> And is a *very* manual way of testing something and ripe for automation

[14:02] < MS-> Heck it could be even that you do something like this

[14:02] < MS-> SchedulerKiller(timeout=...).activate()

[14:02] < MS-> Pipeline(TestSource(),

[14:02] < MS-> Component(),

[14:02] < MS-> TestSink(),

[14:02] < MS-> ).run()

[14:03] < MS-> and after it exits (cleanly or otherwise)

[14:03] < MS-> you check the results inside a testsink

[14:03] < Trun> oh, sounds really good

[14:03] < MS-> That's just one option

[14:03] < MS-> Your work so far may've opened up others

[14:04] < MS-> Build one to throw away after all

[14:04] < MS-> and all that

[14:04] < MS-> Even Davbo's done that with his paint app :)

[14:06] < MS-> (though I'm not sure if that was conscious good behaviour of just a good behavior by accident :) )

[14:06] < MS-> :)

[14:07] < Trun> ok, I'll look how to do this

[14:08] < MS-> OK good

[14:09] < MS-> Can I have daily (short) status updates to the list for the next week in that case as well?

[14:09] < MS-> (Or at minimum every other day?)

[14:09] < Trun> ok

[14:09] < MS-> ta

[14:10] < MS-> (the more clear/frequent updates (==short & includes why) , the better chance we have of pushing things back on track before next friday...)

[14:10] < MS-> In a way that gives us both confidence

[14:11] < Trun> that's good

[14:11] < Davbo> Believe it or not! that was conscious behaviour MS-! ;-)

[14:14] < Davbo> on a related note to earlier I've just managed to get a segfault

[14:14] < Davbo> Fatal Python error: (pygame parachute) Segmentation Fault

[14:27] < MS-> Davbo: cool

[14:27] < MS-> :)

[14:27] < MS-> (re behaviour)

[14:27] < MS-> Re the segfault that's pretty impressive

[14:27] < MS-> Can you make it reprodicible?

[14:27] < MS-> If so, separating it out as a test case would be useful

[14:28] < Davbo> I'm not sure why it happens tbh, if it work it out I'll let you know

[14:28] < Davbo> if it's a pygame error i'll post it to their list

[14:34] < MS-> It could be a threading issue which is why I ask

[14:35] < MS-> If it is, it's a kamaelia problem not a pygame one

[14:38] < Davbo> i'll checkin maybe you could have a look

[14:47] < MS-> ta

[14:48] < MS-> i'll or i've ?

[14:51] < Davbo> oh sorry, it's up now :)

[15:08] < Davbo> http://groups.google.com/group/google-summer-of-code-announce/web/student-survey-questions

[15:09] < Davbo> I can't even answer the first question :-(

[15:09] *** Davbo first contacted Kamaelia after GSoC was announced but before Kamaelia was accepted.

[15:25] < MS-> FWIW, it looks like the problem will be a kamaelia thing

[15:25] < MS-> I have to go now but back later

[15:25] < MS-> I'll explain later

[15:25] *** MS- does svn update for the train

[15:26] < Davbo> alright MS-

[15:26] < MS-> Regarding first Q - I'd say B in your case

[15:26] < MS-> Since it's not A

[15:26] < MS-> :)

[15:27] < Davbo> Yeah I will do

[15:27] < MS-> OK biab

[15:27] *** MS- has parted #kamaelia

[15:37] *** vmlemon__ has joined #kamaelia

[15:37] *** vmlemon__ is now known as vmlemon_

[15:37] < vmlemon_> Hi

[15:43] < vmlemon_> kamaeliabot: dance

[15:43] Reply: does the macarena

[17:55] < Lawouach> back

[17:55] < orphans> hey Lawouach

[17:58] < Lawouach> hey mate

[17:59] < orphans> how's it going?

[17:59] *** Davbo has joined #kamaelia

[18:01] < Lawouach> good

[18:01] < Lawouach> give me a minute please :)

[18:02] < orphans> never! no-one escapes the English inquisition

[18:03] < orphans> or erm, yeah, cool

[18:09] < Lawouach> right-o

[18:09] < Lawouach> sorry

[18:09] < orphans> np :)

[18:10] < Lawouach> you can start

[18:11] < orphans> ok, so since I last spoke to you the main thing that I've done is sort out getting components to tempo-sync properly

[18:11] < Lawouach> yes

[18:11] < orphans> which turned out to be a bit harder than I though, but is OK (if a little hacky) now

[18:13] < orphans> but with that in place I can start work on the piano roll, and should be able to make real music :)

[18:14] < Lawouach> good

[18:14] < Lawouach> you seem to be on schedule according to your plan

[18:14] < orphans> yeah. I had to change it a bit today cause I remembered I need to sort out MIDI later on

[18:15] < orphans> but yeah, I think I'm pretty much on schedule now

[18:15] < Lawouach> right

[18:17] < Lawouach> I liked your StepSequencer code

[18:17] < Lawouach> It was well written

[18:17] < orphans> thanks

[18:18] < orphans> I've put a bit more docs in there now too (over the last few of hours)

[18:18] < orphans> s/of//

[18:20] < Lawouach> You'll have update some areas of the MusicTiming module so that it doesn't refer to the sched module any longer.

[18:20] < orphans> yeah, I added that to the TODO at the top

[18:20] < orphans> there's a couple of bits of doc I need to catch up on - I think I'm going to sort that out tomorrow morning

[18:21] < Lawouach> I've been able to start the jam application without problem

[18:22] < Lawouach> However I still don't get any sound... I assume it's because I haven't configured my MIDI output properly

[18:22] < Lawouach> Also, on the jam application I don't understand why there are two boxes but only one of them supports mouse events.

[18:22] < orphans> ooh, I've got a couple of things to commit for that. I don't know if you remember but I switched to pure-data for testing stuff

[18:22] < Lawouach> Yeah

[18:23] < Lawouach> But I've installed PureData from Synaptic

[18:23] < Lawouach> but not sure that helped: )

[18:23] < orphans> I'll get the stuff committed then update the getting started document and ping you

[18:24] < orphans> it should be really simple now, but I've got to sort out my setup files a bit

[18:24] < Lawouach> okay

[18:24] < Lawouach> no worries

[18:24] < orphans> I forgot I hadn't committed it actually

[18:24] < Lawouach> I'm quite pleased how it moves forward, code wise I don't have much comment so far actually.

[18:24] < orphans> and the two boxes with one controllable is really a test of the network stuff - not really necessary now the step sequencer is running

[18:25] < orphans> cool :) touch wood it seems to be going fairly smoothly at the moment - things are working almost easier than I expected

[18:26] < Lawouach> That's what happens when you set the right bricks at the beginning

[18:26] < orphans> :)

[18:27] < Lawouach> Did you modify the TransformSplitter based on our discussion a couple of weeks ago?

[18:27] < Lawouach> Remember about the filters

[18:27] < orphans> yeah

[18:27] < orphans> oh, wait, maybe not

[18:27] < orphans> what was I supposed to do? Convert everything to using backplanes?

[18:29] < Lawouach> Well it's a bit rusty in my head

[18:29] < Lawouach> :)

[18:29] *** orphans reads back the logs

[18:30] < orphans> < Lawouach> It might be better if the filter was a component

[18:30] < orphans> [18:32] < Lawouach> rather than a function < = that?

[18:31] < Lawouach> yes

[18:31] < Lawouach> I think I explained then how I would do it, didn't I?

[18:31] < orphans> yeah, reading through I've done what you suggested, using a SubscribeTo component for any data coming in through the socket then filtering it through various components

[18:32] < Lawouach> alright

[18:32] < orphans> yup - if you look in jam (which needs splitting up some) you can see how that works

[18:35] < Lawouach> right

[18:35] < Lawouach> you're not using the TransformerSplitter module anymore then?

[18:35] < orphans> no

[18:36] < Lawouach> alright, in that case could you remove it?

[18:36] < Lawouach> I mean later on :)

[18:37] < orphans> yeah. actually reminds me, there's another little test thing which could do with removing too

[18:37] < Lawouach> okay

[18:37] < Lawouach> I like the code

[18:37] < Lawouach> It really smells like Kamaelia code :)

[18:37] < Lawouach> That's good.

[18:37] < orphans> cool :)

[18:38] < Lawouach> In fact I won't bother you any longer tonight. You're moving forward and that's good. I will ask you one thing though...

[18:38] < orphans> uh huh

[18:39] < Lawouach> Since it's also mid-term for you as a student, along the GSoC questions you'll have to answer, could you think of what this experience has provided to you so far.

[18:39] < Lawouach> Don't worry, I'm not asking too much.

[18:40] < Lawouach> I'd like your view between now and three months ago.

[18:40] < Lawouach> Do you feel you've improved? Where? Are there areas you'd like to discuss about?

[18:40] < Lawouach> Things like that, don't write me an essay :)

[18:40] < orphans> ok, do you want it in email form or something?

[18:41] < Lawouach> No no.

[18:41] < Lawouach> I think we'll go through that on our next meeting.

[18:41] < Lawouach> It's really informal.

[18:41] < orphans> ahh, ok

[18:42] < Lawouach> so don't put too much time into it, just think a bit about it and draw me the big picture next week :)

[18:43] < orphans> yeah, I will. I've kinda been thinking about it as I've gone along, but I'll try and get the thoughts in some sort of coherent order :)

[18:43] < Lawouach> Thanks. Much appreciated.

[18:43] < Lawouach> Well that's all for me for this evening.

[18:44] < Lawouach> I'll be looking forward having sound from your application :)

[18:44] < orphans> ok, cool

[18:44] < Lawouach> And I'll be even more interested in launching several jam clients at the same time

[18:44] < Lawouach> and see what happens :)

[18:44] < orphans> yeah - it currently sounds (kinda) lovely here

[18:44] < Lawouach> ah :)

[18:45] < orphans> that is if you like listening to drum beats repeating for ever (which I do :))

[18:46] < Lawouach> that's what happens whenever I use hydrogen :)

[18:47] < orphans> I played a gig a couple of weeks ago, had to spend like a day listening to the same beat in hydrogen so I learnt the lyrics :)

[18:48] < orphans> actually, if you're interested - http://www.archive.org/details/FourLiveSongs

[18:48] < Lawouach> http://fr.youtube.com/watch?v=_SGwDhKTrwU < -- did they steal your idea?

[18:48] < Lawouach> :)

[18:48] < orphans> nah, they stole monome's idea

[18:48] < Lawouach> :)

[18:48] < Lawouach> will be listening to it

[18:48] < orphans> going to make a monome emulator when I get a chance

[18:49] < orphans> I think it'll only be a ~10 line hack, which is fun :)

[18:52] < orphans> right I'm off to get some food I think. Cheers Lawouach

[18:56] < Lawouach> np

[19:21] *** Davbo has joined #kamaelia

[20:08] *** MS- has joined #kamaelia

[21:26] *** orphans has joined #kamaelia

[21:30] *** Davbo is eating large amounts of Prawn Crackers

[21:30] *** MS- has found the .tld ".cat"

[21:31] < Davbo> where's that for?

[21:31] < MS-> catalonia

[21:31] < Davbo> huh, makes sense

[21:31] < Davbo> one day I'll get dav.bo! but that damn Bolivia don't like handing out the top top ones

[21:32] < MS-> .cat appears to be quite expensive sadly

[21:32] < MS-> €62

[21:33] < MS-> (for a year)

[21:36] < Davbo> €?

[21:37] < Davbo> Sorry it didn't show up properly, probably because I'm in Windows (to play Team Fortress 2 with some friends :))

[21:47] < j_baker_> Hey MS-: what about having a Plugins directory in the tree? I'm not really sure about putting the WsgiApps in the Apps directory because they're not really standalone apps.

[23:11] < MS-> Plugins is a strange concept - largely because kamaelia code *is* plugins

[23:12] < MS-> The things you're talking about are really support code for a WSGI app.

[23:12] < MS-> But it's a little late here, so I'll need to discuss that another time

[23:13] < MS-> However, my intent/feeling here is that the code in the WSGI server stuff is actually sufficient to mark it as a separate app (and packageable seperately - independently from your main project - which opens up that aspect)

[23:14] < MS-> They also don't make sense incidentally as "shards" which is a concept we explored last summer - since those are about creating "sub-component pieces" that can be aggregated into a real component

[23:55] < Davbo> Just been playing Audiosurf, I bet orphans would like this game (http://www.audio-surf.com/)