[05:50] *** Lawouach has joined #kamaelia

[06:13] *** vmlemon_ has joined #kamaelia

[06:26] *** vmlemon_ has joined #kamaelia

[06:42] *** vmlemon_ has joined #kamaelia

[06:55] *** Lawouach_ has joined #kamaelia

[06:55] < Lawouach_> morning folks

[07:00] *** vmlemon__ has joined #kamaelia

[07:01] *** vmlemon__ is now known as vmlemon_

[07:28] *** vmlemon_ has joined #kamaelia

[07:48] *** mhrd-afk is now known as mhrd

[07:48] < mhrd> morning

[09:06] *** Uraeus has joined #kamaelia

[09:31] *** Uraeus_ has joined #kamaelia

[09:40] < vmlemon> Hi

[10:06] *** bcarlyon|laptop has joined #kamaelia

[10:27] *** bcarlyon|laptop has joined #kamaelia

[12:11] < Lawouach_> back

[13:37] *** orphans has joined #kamaelia

[14:13] *** vmlemon_ has joined #kamaelia

[14:35] *** Uraeus_ is now known as Uraeus

[15:27] *** vmlemon_ has joined #kamaelia

[15:27] < vmlemon_> kamaeliabot: dance

[15:27] Reply: does the macarena

[15:32] *** Davbo has joined #kamaelia

[15:50] < vmlemon_> Hi Davbo

[15:52] < Davbo> hi

[15:55] < mhrd> orphans: a guy called chris lyon (aka wyleu) on kamaelia-list is expressing an interest in MIDI functionality - he might also be interested in what you're up to if you feel like mentioning it to him :)

[15:57] < orphans> just read it - was gearing up for a reply :) Cheers for pointing it out mhrd

[15:57] < mhrd> np

[15:59] < vmlemon_> kamaeliabot: ecky

[15:59] Reply: Ptang!

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

[16:14] *** Lawouach_ has joined #kamaelia

[16:16] < Lawouach> back

[16:16] < mhrd> just in time for me to head off :-)

[16:16] *** mhrd is now known as mhrd-afk

[16:25] *** vmlemon_ has joined #kamaelia

[16:27] < Davbo> Lawouach: what was the name of the Web Host you use?

[16:29] < Lawouach> webfaction

[16:29] < Davbo> Thanks :)

[16:29] < Lawouach> http://www.webfaction.com/?affiliate=platom

[16:30] < Lawouach> If you feel like using my affiliate user

[16:30] < Davbo> It's not me, my brother is looking for some

[16:32] < j_baker> webfaction is awesome

[16:32] < Lawouach> I see

[16:53] < Davbo> Lawouach: I think that is more than he needs really, so he's gone with a different one. Thanks though :)

[16:53] < Lawouach> fair enough :)

[16:56] *** MS- has joined #kamaelia

[16:56] < MS-> Afternoon

[17:00] < j_baker> Hello

[17:04] *** vmlemon__ has joined #kamaelia

[17:10] *** vmlemon__ is now known as vmlemon_

[17:27] < vmlemon_> Ugh, 387 packets transmitted, 387 received, 0% packet loss, time 391136ms

[17:30] *** vmlemon_ has joined #kamaelia

[17:32] < vmlemon_> Hah, "utils/o/op gaping security hole"

[17:36] < Davbo> Afternoon MS-

[17:39] < MS-> Afternoon

[17:45] < Lawouach> going to get some food

[17:45] < Lawouach> brb

[17:50] < Davbo> Wow, you really see some interesting things on the GSoC students list

[17:50] < Davbo> on the subject of "Gender ratio in open source" someone said this:

[17:51] < Davbo> > If I might suggest some new point of view on this here? This is war

[17:51] < Davbo> > time economy, where lots of the men are out fighting and in most

[17:51] < Davbo> > countries women moved in to take over "male" jobs at the home front.

[17:51] < Davbo> :-/

[17:51] < MS-> What country are they from?

[17:52] < MS-> And, incidentally, what century are they living in ?

[17:52] < MS-> Sounds rather 19th century

[17:52] < Davbo> I know, reply reads "This isn't 1942.  Women go to war too -- at least in the US military. "

[17:53] < MS-> Why not make the suggestion. "Guys, why not just not kill someone today. Do something more productive"

[17:53] < MS-> A lot of the world's problems would be alot better if that simple basic rule guided behaviour

[17:53] < MS-> Well, and "doing my best not to hurt other people"

[17:53] < MS-> But I think world peace is a little beyond my personal scope of influence

[17:54] < Davbo> Hehe :-)

[17:55] *** vmlemon_ is trying to get the code at http://paster.dazjorz.com/?a=rawpaste&p=2983 compiling on Linux (it's modified slightly from the original, to get slightly further in the compilation phase) :|

[17:57] < Davbo> best of luck vmlemon_

[17:57] < vmlemon_> How useful (pun not intended)... http://paster.dazjorz.com/?p=2984

[17:57] *** vmlemon_ has no idea where that stuff would be declared

[17:58] < orphans> evening all

[17:58] < vmlemon_> Hi orphans

[17:58] < Davbo> Evenin' orphans

[17:58] < Davbo> Anyone used NetBeans for JUnit before?

[17:59] < Davbo> The guy testing my program insists on using it, but i use Eclipse

[18:00] *** j_baker has joined #kamaelia

[18:00] < j_baker> Just got my book in

[18:01] < Davbo> Really? I got mine before you lol?

[18:01] < Davbo> They must start by shipping abroad and work inland

[18:06] < j_baker> I suppose so

[18:08] < j_baker> MS- fyi, the link to the SVN web interface on the sourceforge site isn't working.

[18:08] < MS-> ta

[18:08] < MS-> :)

[18:08] < Davbo> Ah I thought that was just me, openDNS shouts at me when i click it

[18:09] < Davbo> sometimes it just likes to do that though.

[18:09] < Lawouach> back

[18:10] < orphans> hey Lawouach

[18:11] < Lawouach> hey orphans, how was your exam?

[18:12] < orphans> erm, a bit strange - came out of it not knowing how I'd done really

[18:12] < orphans> I think I've done enough to get a decent-ish pass

[18:12] < Lawouach> Well good enough

[18:12] < orphans> yeah, it could have been a lot worse :)

[18:13] < Davbo> What University are you at orphans?

[18:13] < orphans> Leicester

[18:13] < Davbo> Ah one of my best mates is at Warwick

[18:13] < Davbo> pretty sure they're dead near each other?

[18:13] < orphans> yeah, not too far at all

[18:14] < Davbo> Cool, best of luck with your exams :)

[18:14] < orphans> cheers :)

[18:14] < Davbo> My first one is next Thursday

[18:14] < orphans> Davbo, what's it on?

[18:15] < Davbo> Artificial Intelligence techniques

[18:15] < Lawouach> orphans: whenever you want to start :)

[18:15] < orphans> sorry Lawouach, go for it :)

[18:15] < Davbo> Search algorithms and such, - I'll leave you guys to your meeting :)

[18:15] < orphans> and good luck Davbo

[18:15] < Davbo> thanks

[18:15] < Lawouach> no worries mate

[18:15] < Lawouach> Wasn't pushing you :)

[18:16] < Davbo> Heh i know, I'm procrastinating from work anyway :)

[18:16] < Lawouach> So last week we concluded you'd needed to do:

[18:16] < Lawouach> 1. a bit of documentation

[18:16] < Lawouach> 2. lay out the directory structure of your application

[18:16] < Lawouach> with a focus on 2.

[18:16] < Lawouach> Had some spare time for it with your exams? :)

[18:17] < orphans> a little - I've done a tiny bit of docs (1 component kinda done), and had a bit of a think about the structure, wanted to run a couple of things by you

[18:17] < Lawouach> Alright. Go for it.

[18:18] < orphans> ok, I wasn't really sure how much I should stick to the structure of Kamaelia. I was kinda thinking of working with UI/Protocol/Util as the main sections where stuff would go

[18:19] < orphans> I guess the majority of the code would live in UI and protocol just by the nature of the project, but things like the timing code are probably a better fit in Util

[18:19] < Lawouach> Well

[18:19] < Lawouach> I think we need MS- for this.

[18:19] < MS-> hm?

[18:20] < Lawouach> Have you made any decision in regards to Kamaelia.Apps ?

[18:20] < Lawouach> You were considering hosting applications in there no?

[18:21] < MS-> You mean in the namespace Kamaelia.Apps ?

[18:21] < Lawouach> As I see it orphans' application should be split into what could be seen as an extension of Kamaelia-the-library and what is specific to his application.

[18:21] < Lawouach> Yes.

[18:22] < MS-> If that's the case, the application should really be developed in a scratch branch for then later compartmentalising stuff out for merge onto trunk

[18:22] < Lawouach> My quetion is whether or not you already have considered of a specific layout or if it would just dropping an application in there.

[18:22] < MS-> Ala the er modeller

[18:22] < Lawouach> Hmmm

[18:22] < MS-> Maybe my branch will answer the q better

[18:22] *** orphans looks

[18:22] < MS-> cf - https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/branches/private_MPS_Scratch/Kamaelia/Kamaelia/Apps

[18:23] < MS-> My suggestion is normally this

[18:23] < MS-> develop in /Sketches, and find out what actually works for the app

[18:23] < MS-> get that clear

[18:23] < MS-> then migrate the names to Kamaelia.Apps.Appname/here

[18:23] < MS-> then migrate code out from there to the standard locations

[18:24] < orphans> so I guess doing that it would be best to stick at least roughly to the kamaelia core structure for the app

[18:24] < orphans> for a start at least

[18:24] < Lawouach> I think so yes.

[18:24] < orphans> cool cool

[18:25] < Lawouach> MS-: what I wonder however is how you consider the deployment

[18:25] < Lawouach> I mean that this year's GSoC is about making Kamaelia better known through real working applications.

[18:25] < Lawouach> Let's face it, easy_install is the de facto standard these days

[18:25] < Lawouach> I wonder how Kamaelia.Apps fits into that.

[18:25] < MS-> I wouldn't be a slave to convention. I'd say if you don't have a clear idea of what the names should be or where they live, do that.

[18:26] < MS-> Kamaelia.Apps should, for most apps, actually end up being a stub

[18:26] < MS-> Or pure support code

[18:26] < Lawouach> I agree with the stub idea

[18:26] < Lawouach> Alright that's cool

[18:26] < MS-> However, for short term it's an interim staging zone

[18:27] < MS-> since it simplifies the process of getting something out there

[18:27] < Lawouach> orphans: One thing I'm highlighting here is that we should consider deployment on the long run.

[18:27] < Lawouach> That doesn't conflict with your idea of following K. structure

[18:27] < MS-> Getting all the components in one place makes it easier to work with.

[18:27] < Lawouach> But it's important to keep deployment in mind.

[18:28] < MS-> I'm not sure I buy easy_install as a working standard though, but that's a conversation for another time.

[18:28] < orphans> yeah, I have to do a bit of work to remember all of my setuptools up to scratch :)

[18:28] < orphans> s/remember/get

[18:28] < MS-> (I did many years of perl work as well, and python is just plain broken with regard to problems perl solved years ago IMO)

[18:28] < Lawouach> MS-: I am not a fan of easy_install but it is there

[18:28] < Lawouach> It's hard to ignore :)

[18:29] *** MS- stops interrupting now :)

[18:29] < Lawouach> anyway, whether or not orphans decide to use it in the end is not important for now.

[18:29] < Lawouach> orphans: I want you to simply keep deployment in mind.

[18:30] < orphans> yeah, I will do. Already had to look a bit at the best way to sort out the bindings/libraries etc

[18:30] *** MS- seconds that

[18:30] < Lawouach> Cool. We're in agreement here then.

[18:30] < orphans> yup

[18:30] *** Lawouach wouldn't go against a Kamaelia specific way of deploying applications BTW

[18:31] < Lawouach> But that'll be for another time too :)

[18:31] < orphans> yeah :)

[18:31] < MS-> probably helpful that I'm a pragmatist :)

[18:31] < Lawouach> :)

[18:32] < Lawouach> orphans: how does your week look like with your exams?

[18:33] < orphans> *sigh* internet dying

[18:34] < orphans> Lawouach, I've got one on Wednesday then a whole week off :)

[18:34] < orphans> so not too bad, but still busy with revision

[18:35] < Lawouach> Okay so i'll leave you in peace tomorrow but I wish you to do one thing please

[18:35] < Lawouach> I'd like you to create the directory structure in the svn repository

[18:36] < Lawouach> Based on what you envision will be required to start with the application.

[18:36] < Lawouach> I wish also that you make a simple use case of that structure so that I can check it out and run an example you would have setup based on that structure

[18:36] < orphans> yup, ok - I'll sort that out

[18:37] < orphans> uh huh

[18:37] < Lawouach> In other words, refactor your existing code and start making it focused on your project itself

[18:37] < Lawouach> You see?

[18:37] < orphans> yeah, I do

[18:39] < orphans> should be a nice test of my svn memory :)

[18:39] < Lawouach> :)

[18:39] < Lawouach> That should be a second language to you :)

[18:40] < orphans> I've used so many different vcses over the last couple of years they kinda all merge into one

[18:40] *** Davbo is looking forward to the lectures next year on revision control systems lol - Should be easy after this summer! :-)

[18:41] < Lawouach> What I'll do then is change a few files in your repository and you'll have to use svn to deal with it :)

[18:42] < orphans> svn undo-whatever-Lawouach-just-did-to-my --please?

[18:42] < Lawouach> :)

[18:43] < orphans> s/my/my-files

[18:43] < Lawouach> The point of the next meeting will be to:

[18:43] < Lawouach> 1. Have the application initiated so that we expand on that base

[18:44] < Lawouach> 2. Ensure you have subversion's most common commands in mind

[18:44] < Lawouach> We will not forget obviously to keep 1. from last meeting as well

[18:44] < Lawouach> Documentation as you go

[18:44] < Lawouach> You'll be happy about that later on

[18:44] < Lawouach> Trust me :)

[18:45] < orphans> yup, I'm fitting it in where I can :)

[18:45] < Lawouach> That's cool.

[18:45] < Lawouach> I won't keep you any longer if you want to relax before more revisions.

[18:46] < orphans> k, cool. Cheers Lawouach

[18:46] < orphans> and MS-

[18:46] < MS-> yw

[18:46] < Lawouach> Thanks MS- indeed

[18:46] *** MS- is about to get some food

[18:48] < j_baker> Food definitely sounds good.

[18:49] < MS-> j_baker: May be a little delayed as a result, but should be back soon :)

[18:49] < j_baker> np... I'll prolly be here

[18:49] < MS-> cool

[18:50] *** Davbo just received an email which means I'll have to miss the meeting on Thursday

[18:50] < Davbo> hmm, I might be there somehow actually

[18:51] < MS-> Thanks for the warning. If you'll still pass in the DONE/BLOCKED/TODOs by email, that'll be fine

[18:51] < Davbo> EEE-PC might save me.

[18:51] < MS-> The structure for meetings does allow for skipping a week if things crop up .

[18:51] < Davbo> Will do MS-

[18:51] < MS-> OK, that is a very food smell.

[18:51] *** MS- is now known as ms-away

[18:51] < Davbo> hehe

[18:52] < j_baker> brb

[18:52] *** j_baker has parted #kamaelia

[18:53] *** vmlemon_ wonders how anyone could manage to make such a convoluted SVN tree, as QNX have done... (It makes it neigh-on impossible to find the file you want, without getting the while tree)

[18:54] < vmlemon_> Yay, complexity...

[18:55] < Davbo> My group has to give a presentation on Thursday at 16:40.

[18:55] < Davbo> 10 minute presentation for 6 people lol

[18:55] < Davbo> actually they say 2 minutes for questions so that makes it 8 minutes

[18:55] *** Davbo sighs

[18:55] < Lawouach> :)

[18:56] *** j_baker has joined #kamaelia

[18:56] < Davbo> 3 people in the groups English isn't that good either so making it so short isn't very nice

[19:09] < vmlemon_> Hah,

[19:09] < vmlemon_> [20:03] < Doc> cia-50: A/S/L ?

[19:09] < vmlemon_> [20:06] < oninoshiko> o.O

[19:09] < vmlemon_> [20:06] < oninoshiko> Doc: getting lonely? gonna date our bot?

[19:24] < j_baker> Quesadillas = pure awesomeness

[19:32] *** bcarlyon|laptop has joined #kamaelia

[19:43] *** vmlemon__ has joined #kamaelia

[19:57] < j_baker> ms-away: I've got to run to the UPS store to drop off a package. I'll be back in a little while.

[20:03] < ms-away> k no problem

[20:03] < ms-away> I'll be here later

[20:03] < ms-away> (back now)

[20:04] < ms-away> (but sorting through stuff, I'll here this go "bing" though)

[20:13] < bcarlyon|laptop> bing

[20:14] < vmlemon__> Bong? ;)

[20:14] < bcarlyon|laptop> bing

[20:14] < vmlemon__> Ping!

[20:14] < bcarlyon|laptop> pong

[20:15] < vmlemon__> Byng?

[20:15] < bcarlyon|laptop> Im sorry the host could not be reached

[20:16] *** bcarlyon|laptop has joined #kamaelia

[20:16] *** ms-away notes that bing is a bizarrely useful programme when you need it

[20:16] < bcarlyon|laptop> Ha vista made hydrairc crash lol

[20:16] < ms-away> much like ping is, but for different reasons

[20:17] < bcarlyon|laptop> hmm

[20:17] < ms-away> http://linux.maruhn.com/sec/bing.html

[20:17] < bcarlyon|laptop> nice

[20:20] < vmlemon__> Hmm, interesting

[20:20] *** vmlemon__ wonders what happened to 'dog' (the utility)

[20:29] < vmlemon__> ("Enhanced replacement for cat " according to Debian's package page for it)

[20:30] *** mhrd-home has joined #kamaelia

[20:30] < vmlemon__> Aah, I've just found a copy

[20:31] *** vmlemon__ was curious about how useful it was

[20:31] < vmlemon__> *is

[20:32] *** vmlemon__ downloads bing

[20:41] < Davbo> Evenin mhrd-home

[20:42] < mhrd-home> hi

[20:42] < ms-away> evening

[20:42] < ms-away> private_JMB_AddedIndexCapabilityToMimimal is merged, yes?

[20:42] < ms-away> (I'm tidying up /branches, so just checking before moving to merged :) )

[20:45] *** vmlemon_ has joined #kamaelia

[20:57] *** Davbo should really stop filling up his hard drive with .FLAC music.

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

[21:26] < j_baker> Ok, I'm finally here ms-away

[21:26] < j_baker> (had internet problems)

[21:27] *** mhrd-home gets pointed at yet another standards doc describing DVB PSI data types :-/ :-)

[21:28] < vmlemon_> Hmm, "No reason we couldn't get it to work. Anything UNIXy is particularly easy for us. I personally find it rather distastful, but LINA itself is blind to legal and political matters."

[21:32] < Davbo> Can anyone access http://svn.sourceforge.net/viewcvs.cgi/kamaelia/ ?

[21:33] < vmlemon_> I think they changed it so that it doesn't work anymore

[21:33] < vmlemon_> (It's now svn.projectname.sourceforge.net for the host, as far as I know)

[21:34] < Davbo> oh

[21:34] < Davbo> http://kamaelia.svn.sourceforge.net/viewvc/kamaelia/

[21:34] < Davbo> should be that

[21:34] < Davbo> I was following the link from the kamaelia front page, thought it had been changed

[21:35] *** vmlemon_ assumes that there are plenty of little "gotchas" like that on the present Kamaelia #site :(

[21:35] < vmlemon_> *site

[21:36] < Davbo> we'll sort it out vmlemon_ :-)

[21:38] *** ms-away is now known as MS-

[21:38] < vmlemon_> I'm sure that that stuff will be sorted when/if the new site gets rolled out (with added Hype, of course)...

[21:38] < MS-> when

[21:38] < MS-> Hi j_baker

[21:38] < j_baker> Hello

[21:39] < MS-> So, I don't have anything specific to raise this week - I can see that you've been pushing on.

[21:39] < MS-> Which is appreciated.

[21:39] < MS-> Your branch was also merged, which was great :)

[21:39] < j_baker> Awesome. For the most part, I've just been working on setting up some file structuring.

[21:41] < MS-> Indeed. Refactoring is an often under-estimated task, and worth doing.

[21:41] < j_baker> One question I had, though: is there any reason why the server code relies so much on factory functions? It seems like it could be component-ized more.

[21:41] < MS-> Which server code?

[21:41] < MS-> The HTTP server?

[21:42] < MS-> Or MoreComplexServer (which I don't think does)

[21:42] < j_baker> Well, for example, you use a factory function to create the routing table.

[21:42] < j_baker> Or to pull a component out of said table.

[21:42] < MS-> That's a yes for the HTTP server :)

[21:43] < MS-> The reason is because the person who wrote that code originally worked that way when writing it

[21:43] < MS-> (which was Ryan)

[21:43] < MS-> Since the code worked & worked sufficiently well, that's why it was merged

[21:44] < j_baker> Ok. I was just curious if there was an underlying reason for it. I was thinking about changing a few of those things into components.

[21:44] < MS-> As you say, it could be componentised much more, and would benefit from doing so

[21:44] < MS-> For example, repurposing it to be an RTSP server and extending it for more HTTP methods would be easier

[21:45] < j_baker> Ok. I have a few other TODOs to work on atm, but that will be a priority for me.

[21:46] < j_baker> (most importantly, I want to get a logger up and running)

[21:47] < MS-> Sounds like a sensible plan

[21:47] < j_baker> For the most part though, I think the WSGI code is working. I've gotten most of the "musts" out of the way. But there are a few strong "shoulds" that need to be worked on.

[21:48] < MS-> One thing I would like to consider is how to start moving code from /Sketches into a /branches/Scratch/Apps/Kamaelia-Webserver location

[21:48] < MS-> That would then make it easier to test the WSGI code against real WSGI applications.

[21:49] < MS-> It would also make a start to deal with some of the deployment issues your project will have

[21:49] < MS-> Since it is expected to be a client app

[21:49] < j_baker> Yeah. I think the biggest problem will be resolving dependencies sensibly.

[21:49] < MS-> So getting it into a position for py2exe and py2app to be able to work with it sooner rather than later is a good idea :)

[21:50] < MS-> The WSGI compatibility (musts) is really good to see

[21:51] < j_baker> Oh, one somewhat minor question: Is there a reason why when I use sys.path(), the first string in the list is an empty string?

[21:51] < MS-> the wsgi.input support would suffer from a DOS attack vulnerability, but as a first pass your approach looked sensible. (based on reading the checkins)

[21:51] < MS-> Current directory

[21:51] < MS-> At a guess it's equivalent to saying "." in a path

[21:52] *** vmlemon_ wonders how easy it would be to make a "dual-mode" HTTP/RTSP server that runs both a single port, and checks the protocol string for either "RTSP/$version" or "HTTP/$version" and then serves as appropriate

[21:52] < MS-> Which seems to be less used these days

[21:52] < vmlemon_> *on a

[21:52] < j_baker> Oh, so should that be left in the path environment variable?

[21:52] < MS-> vmlemon_: It shouldn't be difficult really

[21:52] < MS-> sys.path is the python path

[21:53] < MS-> the path environment variable isn't that

[21:53] < MS-> >>> os.environ["PATH"]

[21:53] < MS-> '/opt/kde3/bin:/home/zathras/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/X11R6/bin:/usr/games:/usr/lib/jvm/jre/bin:/usr/lib/mit/bin:/usr/lib/mit/sbin:/usr/lib/qt3/bin'

[21:53] < MS-> >>> sys.path

[21:53] < MS-> ['', '/usr/lib/python25.zip', '/usr/lib/python2.5', '/usr/lib/python2.5/plat-linux2', '/usr/lib/python2.5/lib-tk', '/usr/lib/python2.5/lib-dynload', '/usr/lib/python2.5/site-packages', '/usr/lib/python2.5/site-packages/Numeric', '/usr/lib/python2.5/site-packages/PIL', '/usr/lib/python2.5/site-packages/gtk-2.0', '/usr/lib/python2.5/site-packages/wx-2.8-gtk2-unicode', '/usr/local/lib/python2.5/site-packages']

[21:53] < j_baker> Ah, I see.

[21:54] < j_baker> That won't be difficult to change.

[21:54] < MS-> :)

[21:54] < MS-> Shifting the code over as a simple Kamaelia.App will also make code review and system testing easier

[21:54] < vmlemon_> Hell, it might even stretch to do Jabber and SIP, too

[21:54] < j_baker> Also, how exactly is wsgi.input vulnerable?

[21:54] < vmlemon_> Depending on how far someone would be willing to take the approach

[21:55] < MS-> I only gave it cursory look, so I'll check

[21:55] < MS-> Ah yes, it is

[21:55] < j_baker> I'm guessing that if you sent enough requests it would cause too many sockets to open. Is that what you were getting at?

[21:55] < MS-> It's because you do this: cStringIO.StringIO(full_request)

[21:56] < j_baker> s/sockets/file handles

[21:56] < MS-> No, normally if I deal with stdin inside a CGI, I'm not necessarily storing the full object into memory

[21:56] < MS-> whereas the above line will do

[21:57] < MS-> Which means all I do is post a moderate number of large files at the server and I can exhaust memory

[21:57] < MS-> But the key thing though is "it works"

[21:57] < j_baker> Ah. That makes sense.

[21:57] < MS-> :)

[21:57] < j_baker> As of right now, I can't think of any other questions I had.

[21:58] < MS-> Cool.

[21:58] < MS-> This evening I've been going through branches tidying them up

[21:59] < MS-> since as well as sorting the website, I want to merge Kamaelia-Grey - which would result in the code you're relying on being merged onto /trunk

[21:59] < Davbo> MS-: ping me when you're not busy if that's okay. Might be better to just leave it for our meeting tomorrow actually :)

[21:59] < MS-> Making that a good time to make your own branch

[21:59] < j_baker> Ok.

[22:00] < j_baker> Also, one thing that I thought about suggesting (may be best for the meeting or the mailing list):

[22:00] < MS-> If you need something specific and specific feedback, please do shout :)

[22:00] < MS-> Go for it

[22:00] < Davbo> File "/home/dave/Kamaelia/kamaelia-trunk/Sketches/MPS/pprocess/MultiPipeline.py", line 6, in < module>

[22:01] < Davbo> import pprocess

[22:01] < Davbo> ImportError: No module named pprocess

[22:01] < j_baker> Why don't we use a system sort of like twisted does for versioning? Instead of trying to version everything into one big megabundle, why don't we version components as groups?

[22:01] < Davbo> Have I got the wrong version of Kamaelia installed?

[22:01] < MS-> Davbo: http://pypi.python.org/pypi/pprocess

[22:02] < Davbo> I should have googled that lol :)

[22:02] *** Uraeus has joined #kamaelia

[22:02] < Davbo> Thanks MS-

[22:03] < MS-> j_baker: There's a few issues there really

[22:04] < MS-> 1) The megabundle was/is a relatively nice idea, but also causes issues, so that's why for the apps I've investigated better packaging means

[22:04] < MS-> For Axon & Kamaelia then we have a relatively simple version number

[22:04] < MS-> Axon was initially released as 1.0.0

[22:04] < MS-> since that was an API stable release

[22:05] < Davbo> My answer j_baker: lots of components use each other, easier to bundle what is compatible rather than keep track of which version of which group of components works with others

[22:05] < MS-> since then it has been extended, and 1.5 was a large jump in functionality

[22:05] < MS-> which was the core point of that jump

[22:05] < MS-> However, Kamaelia as a single bundle is still as far as I'm concerned in flux, since it doesn't cover the functionality I want before I'm ready to declare it feature complete

[22:05] < MS-> As a result, that's why it's less than 1.0

[22:06] < MS-> However, each step moves it closer to where it should be for being basic feature complete

[22:06] < MS-> And Davbo's answer is spot on.

[22:07] < MS-> The version number says "this lot works together"

[22:07] < MS-> The other point is twisted changed their versioning mechanism recently, so I'm not sure actually what their rationale is at the moment

[22:08] < MS-> I do know it's related by date.

[22:08] < MS-> For the megabundle, that does make sense.

[22:08] < Davbo> Humm, I can do Pipeline(MagnaDoodle(), MagnaDoodle()).run() and works fine, - well it's interesting what happens anyway.

[22:08] < j_baker> I see. I suppose that does make things easier.

[22:09] < MS-> I'll take another look at what they're doing though, and revisit things

[22:09] < MS-> I *have* been tempted to allow for *sub-distributions* as well though

[22:09] < MS-> Such as "just internet server core related"

[22:10] < MS-> Which could be useful

[22:10] < j_baker> Oh, I don't know exactly what they're doing now either. I suppose we're using the same model as django and it seems to work for them.

[22:10] < Davbo> MS-: Should MultiPipeline work with similar syntax to Pipeline?

[22:10] < vmlemon_> Hmm, not meaning to be negative, but won't that create a big mess for people who won't know which version to choose?

[22:10] < MS-> I've not looked at what Django does :)

[22:10] < MS-> Davbo: ProcessPipeline is the same as Pipeline for a user

[22:10] < vmlemon_> And then they'll screech later about it being "broken", when they were stupid enough to make the wrong choice

[22:10] < MS-> vmlemon_: That's the problem I want to avoid

[22:11] < vmlemon_> (Just my 2p)

[22:11] < MS-> And it's the rationale behind the megabundle

[22:11] < j_baker> I do know that every time I go to install it, it tells me that I'm better off downloading the current development version than installing the packaged version.

[22:11] < vmlemon_> The MegaBundle is a sound idea, though

[22:11] < MS-> j_baker: That's a symptom that packaging is bust

[22:11] < vmlemon_> Assuming that you can persuade people to install everything in it ;)

[22:12] < MS-> yeah

[22:12] < vmlemon_> (Which doesn't always happen)

[22:12] < vmlemon_> Short of saying, "INSTALL THIS... OR ELSE!" ;)

[22:14] < Davbo> Where is ProcessPipeline MS-?

[22:16] < MS-> Davbo: in /Sketches/MPS/pprocess/MultiPipeline.py ...

[22:16] < MS-> You *have* looked in there ? :)

[22:17] < Davbo> YAY!

[22:17] < Davbo> :)

[22:17] < Davbo> It's alive

[22:17] < MS-> :)

[22:17] < MS-> Cool :)

[22:19] < Davbo> Wow eating my CPU#

[22:19] < vmlemon_> Just remember not to get it wet, and not to feed it after midnight ;)

[22:19] < MS-> Step 1, make it work

[22:20] < MS-> Probably worth checking your code in BTW

[22:20] < Davbo> Step 2, make it work without the CPU fan taking off

[22:21] < MS-> :)

[22:23] < Davbo> commited MS-

[22:23] < MS-> cool

[22:24] < MS-> Works here too :)

[22:24] < Davbo> Custom size circles now too

[22:25] < MS-> and happily using 2 CPUs, for simply adding one import and changing one line of code...

[22:25] < Davbo> Yeah, it's really cool! :)

[22:25] < Davbo> I'm thinking about how to get them talking to each other now

[22:25] < vmlemon_> Hmm, anyone got a PlayStation 3 yet, to test Linux and Kamaelia on?

[22:25] < vmlemon_> (Out of curiosity)

[22:26] < Davbo> No but j_baker has used YellowDog Linux with Kamaelia i think vmlemon_

[22:26] < vmlemon_> Aah

[22:26] < Davbo> which is what most people put on there PS3's

[22:27] < Davbo> Well, if they're going to install Linux on it lol :)

[22:27] < vmlemon_> Would have been interesting to see how it would have performed in it's "natural habitat" (multithreaded/multicore machines)

[22:27] < MS-> vmlemon_: Indeed

[22:27] < MS-> It would be interesting

[22:27] < vmlemon_> I suppose there's always a Cray machine or Blue Gene/L ;)

[22:28] < vmlemon_> Assuming that you could get access to one ;)

[22:29] < MS-> I wish :)

[22:29] < vmlemon_> Or one of the Sun 30 day free trial boxes

[22:29] < vmlemon_> (with 8 cores+, last time I checked)

[22:30] < Davbo> Hmm I wonder what we have at Uni

[22:30] < Davbo> We used to have one of the most powerful supercomputers back in the day

[22:30] < vmlemon_> Rusty old SGI boxes, if you're lucky? Some iMacs?

[22:30] *** vmlemon_ assumes, anyway

[22:31] < Davbo> I think so vmlemon_

[22:32] < vmlemon_> Dunno if it'd work on OpenVMS without reworking parts of it

[22:33] < vmlemon_> (If anyone even has a functional VMS machine, of any architecture)

[22:34] < MS-> VMS was a nice system

[22:34] < MS-> Odd, but nice

[22:34] < vmlemon_> Probably the most obscure thing I've ever tried to run Kamaelia on was QNX, and I failed miserably trying to get a new enough version of Python working

[22:34] < vmlemon_> I have a copy of Novell NetWare and SCO OpenServer somewhere, on another machine

[22:35] < vmlemon_> Dunno if it'd work on them, or how well it would work, if I was lucky enough to get it to install

[22:37] < vmlemon_> It probably functions without trouble on F/O/N BSD, although I don't think anyone has tested

[22:37] < Davbo> MS-: I'll think about how to get them talking for our meeting tomorrow.

[22:37] < Davbo> Night all

[22:37] < j_baker> I see. So the next step is Kamaelia on the NetBSD toaster?

[22:37] < j_baker> night Davbo

[22:37] < vmlemon_> Hah

[22:38] < vmlemon_> See you Davbo

[22:38] < MS-> night Davbov

[22:38] < vmlemon_> Would the NetBSD toaster even have enough RAM and CPU power?

[22:38] < vmlemon_> Kamaelia on a heavily modified Sky Digibox? On a Motorola EZX/Linux-based phone?

[22:39] < j_baker> I have a motorola linux phone. I don't know if it even has python though

[22:40] < MS-> j_baker: is it a series 60 phone?

[22:40] < vmlemon_> It supposedly works on Symbian OS, although I don't know if it was only earlier versions that worked

[22:40] < MS-> If it is, then kamaelia should run there

[22:40] < j_baker> It's a A1200 ming

[22:40] *** vmlemon_ notes that the Series 60 environment doesn't work on Linux...

[22:41] < vmlemon_> Any idea on how to actually install it on a Symbian OS/Series 60 phone?

[22:41] < vmlemon_> Assuming that the Python runtime is already installed

[22:41] < MS-> Been a couple of years since I've done that I must admit, and we actually cheated somewhat - the phone used has a small memory card

[22:42] < MS-> and used that to copy the data on

[22:42] *** vmlemon_ doubts that the very latest version would run, somehow

[22:42] < vmlemon_> (Seeing as the version for S60 is based on an ancient version of Python, or was last time I checked)

[22:42] < MS-> It's based on 2.2.< some late version>

[22:43] < MS-> Which was the first to support generators

[22:44] *** vmlemon_ wonders if it would work on Windows CE/Windows Mobile

[22:44] < MS-> Does python run there?

[22:44] < vmlemon_> Not sure

[22:45] < MS-> I didn't think it did

[22:45] < vmlemon_> Hah, "PCs are like air conditioners, if you open Windows(R) they don't work."

[22:48] *** mhrd-home notices MS- has been busy in /branches/ :)

[22:49] *** mhrd-home calls it a night

[22:49] < MS-> Night

[22:49] < MS-> :)

[22:50] < mhrd-home> :)

[22:50] *** mhrd-home has parted #kamaelia

[22:53] *** vmlemon_ is off for the night...

[23:01] *** MS- notes that the best way to merge Jinna's code onto the mainline will be to split it into a number of sub-branches, and merge those.

[23:01] < MS-> The code is known good & working though - which helps