[00:20] *** kamaeliabot has joined #kamaelia |
[00:21] < MS-> test |
[00:21] < MS-> http://yeoldeclue.com/logs/kamaelia2008-03-20_log.html |
[00:21] < MS-> We've been accepted for GSOC 2008 | ideas/info page: http://kamaelia.sourceforge.net/SummerOfCode2008 | Don't ask to ask, just ask | Multicore Kamaelia : http://yeoldeclue.com/cgi-bin/blog/blog.cgi?rm=viewpost&nodeid=1205626569 | logs: http://yeoldeclue.com/logs/ |
[00:22] *** MS- changed the topic to We've been accepted for GSOC 2008 | ideas/info page: http://kamaelia.sourceforge.net/SummerOfCode2008 | Don't ask to ask, just ask | Multicore Kamaelia : http://yeoldeclue.com/cgi-bin/blog/blog.cgi?rm=viewpost&nodeid=1205626569 | logs: http://yeoldeclue.com/logs/ |
[00:23] < MS-> OK, That's the logger packaged up and tested on linux and mac os |
[00:23] < MS-> https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/branches/private_MPS_Scratch/Apps/Kamaelia-Logger has the packaging |
[00:24] < MS-> Specifically https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/branches/private_MPS_Scratch/Apps/Kamaelia-Logger/DistBuild/Logger.build.sh is the package build script |
[00:24] < MS-> (based on the packaging approach for kamaelia grey) |
[00:25] < MS-> http://yeoldeclue.com/t/Kamaelia-Logger-1.0.0.tar.gz is the tar ball |
[00:26] < MS-> install is the usual |
[00:26] < MS-> tar zxvf Kamaelia-Logger-1.0.0.tar.gz |
[00:26] < MS-> cd Kamaelia-Logger-1.0.0 |
[00:26] < MS-> sudo python setup.py install |
[00:26] < MS-> dance |
[00:26] < MS-> Config file in /etc |
[00:27] < MS-> init script in /etc/init.d under linux and in /System/Library/StartupItems/KamaeliaLogger on Mac OS X |
[00:27] < MS-> (tested on both, running on the latter) |
[00:28] < MS-> and on that note, I'm off :) |
[00:28] *** MS- is now known as ms-away |
[00:30] < ms-away> Oh, before I go |
[00:30] < ms-away> all lines preceded with [off] |
[00:30] < ms-away> aren't logged |
[00:30] < ms-away> (kinda handy) |
[01:35] < Davbo> Looks good ms-away |
[01:59] *** robertofaga has joined #kamaelia |
[07:52] *** vmlemon__ has joined #kamaelia |
[08:45] *** ms-away has parted #kamaelia |
[09:07] *** mhrd has joined #kamaelia |
[09:37] *** Uraeus has joined #kamaelia |
[10:38] *** MS- has joined #kamaelia |
[11:36] *** Lawouach has joined #kamaelia |
[11:46] *** Lawouach has joined #kamaelia |
[12:36] < MS-> kamaeliabot: help |
[12:36] Reply: Name: kamaeliabot Channel: #kamaelia |
[12:36] Reply: I do a simple job -- recording all channel traffic. |
[12:36] Reply: Lines prefixed by [off] won't get recorded |
[12:36] Reply: I respond to the following: 'logfile', 'infofile', 'help', 'date', 'time', 'dance', 'poke', 'slap', 'ecky', 'boo', and 'reload {modulename}'. |
[12:36] < MS-> kamaeliabot: dance |
[12:36] Reply: does the macarena |
[12:37] *** vmlemon__ has joined #kamaelia |
[12:38] < mhrd> kamaeliabot: help reload |
[12:38] Reply: Name: kamaeliabot Channel: #kamaelia |
[12:38] Reply: I do a simple job -- recording all channel traffic. |
[12:38] Reply: Lines prefixed by [off] won't get recorded |
[12:38] Reply: I respond to the following: 'logfile', 'infofile', 'help', 'date', 'time', 'dance', 'poke', 'slap', 'ecky', 'boo', and 'reload {modulename}'. |
[12:38] < mhrd> kamaeliabot: reload |
[12:39] < MS-> It's just Jinna's code from last summer, but repackaged up for easier install/maintainence |
[12:39] < vmlemon__> Hi |
[12:40] < MS-> vmlemon__: Hi |
[12:40] < MS-> mhrd: https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/branches/private_MPS_Scratch/Apps/Kamaelia-Logger |
[12:41] < MS-> It expects to reload changes to this https://kamaelia.svn.sourceforge.net/svnroot/kamaelia/branches/private_MPS_Scratch/Kamaelia/Kamaelia/Apps/IRCLogger/Support.py (i think :) ) |
[12:49] < vmlemon__> Hmm, do they still sell the old (green, I think) BT Phonecards? |
[12:50] < MS-> No idea |
[12:50] < MS-> I don't think so |
[12:58] *** vmlemon__ wonders why phoneboxes still have slots for them |
[13:10] < MS-> You can put creditcards in |
[13:10] < MS-> So if they're magstripe readers.. |
[13:11] < MS-> maybe it's just a matter of "we didn't say this would go away and its too expensive to change" |
[13:11] < MS-> mayb |
[13:11] < MS-> e |
[13:18] < vmlemon__> I know that some boxes have both a credit card reader, and a legacy green card reader as two slots |
[13:19] < vmlemon__> Probably to avoid confusion with boxes that don't accept credit cards |
[13:22] < vmlemon__> They also did an incompatible "new-style" card |
[13:56] < vmlemon__> Shame some people tar all of the BBC with the same brush |
[13:56] < MS-> Of course they d |
[13:56] < MS-> o |
[13:56] < MS-> It's human nature, sadly |
[14:26] *** Davbo has joined #kamaelia |
[15:18] < Davbo> I've got a question about MiniAxon |
[15:18] < Davbo> Is there any difference between |
[15:18] < Davbo> result = self.boxes[inboxname][0] |
[15:18] < Davbo> del self.boxes[inboxname][0] |
[15:18] < Davbo> return result |
[15:18] < Davbo> and |
[15:19] < Davbo> return self.boxes[boxname].pop() |
[15:20] < MS-> >>> X=[1,2,3,4] |
[15:20] < MS-> >>> X.pop() |
[15:20] < MS-> 4 |
[15:20] < MS-> >>> X |
[15:20] < MS-> [1, 2, 3] |
[15:20] < MS-> >>> X=[1,2,3,4] |
[15:20] < MS-> >>> X[0] |
[15:20] < MS-> 1 |
[15:20] < MS-> >>> del X[0] |
[15:20] < MS-> >>> X |
[15:20] < MS-> [2, 3, 4] |
[15:20] < MS-> >>> X=[1,2,3,4] |
[15:20] < MS-> >>> X.pop(0) |
[15:20] < MS-> 1 |
[15:20] < MS-> >>> X |
[15:20] < MS-> [2, 3, 4] |
[15:20] < MS-> is probably what you're thinking |
[15:21] < Davbo> ah |
[15:21] < Davbo> so i need to change it to pop(0) so it takes one from the front of the list |
[15:21] < MS-> Well, needs to be the opposite end from where you append :) |
[15:21] < Davbo> otherwise it's working like a stack and not a list right? |
[15:21] < MS-> stack rather than queue |
[15:22] < Davbo> yeah, thanks MS- |
[15:23] < Davbo> kinda forgot you could just run "python" and easily test it there. |
[15:24] < Davbo> getting used to an interpreted language again |
[15:29] < MS-> :) |
[15:47] < Davbo> finished the MiniAxon |
[15:49] < Davbo> dictionaries seem so useful |
[15:50] < Davbo> i may have to hassle my lecturers to let me implement my project in Python :p |
[16:02] < MS-> COol |
[16:03] < MS-> Dictionaries are highly optimised in python btw |
[16:03] < MS-> since they use the same implementation as the byte code interpreter uses for looking up names |
[16:03] < MS-> and values |
[16:07] *** LightningMonkey has joined #kamaelia |
[16:07] < LightningMonkey> hello |
[16:09] < MS-> Hi |
[16:10] < LightningMonkey> I am interested in some of the google summer of code projects |
[16:10] < MS-> ok |
[16:10] < MS-> which one(s) ? |
[16:10] < LightningMonkey> specifically working on the topology viewer to make it 3D |
[16:11] < MS-> Cool. |
[16:11] < LightningMonkey> it sounds like it would be a fun project |
[16:11] < MS-> Indeed. |
[16:11] < MS-> The existing topology visualiser is based on a simple physics model |
[16:11] < LightningMonkey> I have experience with OpenGL |
[16:11] < LightningMonkey> mhm |
[16:11] < LightningMonkey> I am also a fan of physics |
[16:11] < MS-> which whilst 2D based, works on 2D vectors |
[16:11] < LightningMonkey> mhm |
[16:12] < MS-> extending it should be a matter of changing that to 3D vectors, and an appropriate open GL visualisation component |
[16:12] < LightningMonkey> so is it as simple as particles having vector directions, then add forces to slow/speed them up |
[16:12] < MS-> No,particles are simply points |
[16:13] < MS-> there's linkages which provide tensions |
[16:13] < MS-> (attract/repulse) |
[16:13] < LightningMonkey> ah |
[16:13] < MS-> It's a very simple model :) |
[16:14] < MS-> http://kamaelia.svn.sourceforge.net/viewvc/kamaelia/trunk/Code/Python/Kamaelia/Kamaelia/Visualisation/PhysicsGraph/ has the code for the 2D components |
[16:14] < MS-> http://kamaelia.svn.sourceforge.net/viewvc/kamaelia/trunk/Code/Python/Kamaelia/Kamaelia/Visualisation/PhysicsGraph/TopologyViewer.py?revision=2735&view=markup |
[16:14] < MS-> is the main entry point |
[16:15] < MS-> http://kamaelia.svn.sourceforge.net/viewvc/kamaelia/trunk/Code/Python/Kamaelia/Kamaelia/Support/Particles/ contains the particles themselves |
[16:15] < LightningMonkey> By the way, I am current at work so if I don't respond quickly it is because I have to help a suer |
[16:16] < LightningMonkey> 'user |
[16:16] < LightningMonkey> I work in a computer lab |
[16:16] < MS-> along with the laws that get applied at each iteration (esentially its a simulation) |
[16:17] < MS-> http://kamaelia.svn.sourceforge.net/viewvc/kamaelia/trunk/Code/Python/Kamaelia/Kamaelia/Support/Particles/ParticleSystem.py?revision=2356&view=markup |
[16:17] < MS-> is the core of the particle system code |
[16:17] < MS-> The key part of that piece of code is this: |
[16:17] < MS-> 73 ParticleSystem will work for particles in space with any number of dimensions - |
[16:17] < MS-> 74 so long as all particles use the same! |
[16:18] < LightningMonkey> very nice |
[16:18] < MS-> Open GL components are here: |
[16:18] < MS-> http://kamaelia.svn.sourceforge.net/viewvc/kamaelia/trunk/Code/Python/Kamaelia/Kamaelia/UI/OpenGL/ |
[16:18] < MS-> And you'd essentially be rendering using the same approach there |
[16:19] < MS-> Or expected to |
[16:19] < LightningMonkey> right |
[16:19] < LightningMonkey> so the project would be just to make it into 3D, or would there be tinkering with the physics model as well? |
[16:19] < MS-> http://kamaelia.svn.sourceforge.net/viewvc/kamaelia/trunk/Code/Python/Kamaelia/Kamaelia/UI/OpenGL/Button.py?revision=2735&view=markup |
[16:19] < MS-> Gives an example of a simple 3D "button" that can sit in space |
[16:20] < MS-> The project's goal would be to make the topology visualiser work in 3D |
[16:20] < MS-> If that involves changes to the physics model, that's fine |
[16:20] < MS-> If it doesn't need it, that's fine too |
[16:20] < LightningMonkey> alright |
[16:20] < MS-> The key use that we'd be interested in (due to someone asking for it :) ) |
[16:21] < MS-> is the ability to dump a social network into the visualiser and explore it |
[16:21] < MS-> and do so in a 3D environment |
[16:22] < MS-> (not necessarily a huge one, but that's what the aim would be) |
[16:22] < LightningMonkey> That is a very cool idea I will admit |
[16:23] < LightningMonkey> So if it was a social network like say, facebook, would each node be a person and connection be who they are friends with? |
[16:23] < LightningMonkey> at least something of that nature? |
[16:23] < MS-> The idea is actually based around organising & exploring information around contacts between organisations |
[16:24] < MS-> "who knows X who also knows Y in organisation A, B, C" |
[16:24] < MS-> But you can also view it in terms of facebook |
[16:24] < MS-> though with a much lower number of friends per node :) |
[16:24] < LightningMonkey> very true |
[16:24] < MS-> and more colleagues than friends :) |
[16:24] < LightningMonkey> lol |
[16:25] < MS-> not that the two are necessarily disjoint :) |
[16:25] < LightningMonkey> yes i was about to say that one would hope some of the colleagues were friends |
[16:25] < LightningMonkey> lol |
[16:26] < MS-> So there may be some interesting possible extensions to the physics model to say "these things should be in the same locale, even though there isn't a link between them" |
[16:26] < MS-> and "these others are from a different group" |
[16:26] < MS-> the 2D topology visualiser has turned out to be very useful in the past and its customisability has turned out useful |
[16:27] < LightningMonkey> so like nodes that represent people in the same group should remain close to each other even if they dont have a link? |
[16:27] < MS-> yes |
[16:27] < MS-> For example I needed to remodel a large datamodel and didn't particularly want to draw it, even though visualising it was the best way - so I created an ER modelling tool - which is documented roughly here: |
[16:27] < MS-> http://yeoldeclue.com/cgi-bin/blog/blog.cgi?rm=viewpost&nodeid=1195955570 |
[16:28] < MS-> The separation of rendering from physics makes it really nice to work with |
[16:28] < MS-> (I didn't write the original engine :) ) |
[16:28] < LightningMonkey> nice |
[16:29] < LightningMonkey> it would seem you could just use the existing attraction/repelent model to keep people together, just give them lower attraction constants |
[16:29] < LightningMonkey> or something of that nature |
[16:29] < MS-> Quite possibly. You could also think of it like a weak force & a strong force |
[16:29] < LightningMonkey> yup |
[16:30] < MS-> Weak attract/repulse for the overall type and strong at the nodel level |
[16:30] < LightningMonkey> exactly |
[16:31] < MS-> The nice thing about that is that you'd be able to see if there are individuals in groups that are highly connected beyond groups |
[16:31] < LightningMonkey> this music model is pretty cool by the way |
[16:31] < MS-> you'd expect "yes" to that, and it'd be nice |
[16:31] < MS-> music model? |
[16:31] < MS-> Oh that |
[16:31] < LightningMonkey> the link you jsut sent |
[16:31] < MS-> That's from looking at my old universities website and looking for current examples |
[16:32] < MS-> and then using that to re-render :) |
[16:32] < LightningMonkey> mhm |
[16:33] < LightningMonkey> So as I said before, I have worked with opengl before, but not through python. I have always done it though C |
[16:33] < MS-> The PyOpenGL interface is pretty much the same |
[16:34] < LightningMonkey> ya I noticed |
[16:34] < MS-> It calls out to the same libraries |
[16:34] < MS-> So as long as you use display lists, it's not a huge deal |
[16:34] < LightningMonkey> mhm |
[16:34] < MS-> But then that's not exactly unusual |
[16:35] < MS-> The Open GL support was added BTW by a GSOC'er 2 years ago |
[16:35] < LightningMonkey> Well I may hopefully continue in that grand tradition |
[16:36] < LightningMonkey> hehe |
[16:36] < MS-> His blog from GSOC was here: http://thfsoc.blogspot.com/ |
[16:38] < MS-> OK, I'm going AFK for a couple of hours. kamaeliabot is logging the channel - logs are http://yeoldeclue.com/logs/ - if you have questions, please feel free to ask away and I'll see them when I come back |
[16:38] < MS-> cya |
[16:38] *** MS- has parted #kamaelia |
[16:38] *** LightningMonkey has parted #kamaelia |
[17:27] *** mhrd has parted #kamaelia |
[18:12] *** MS- has joined #kamaelia |
[19:37] < Davbo> MS-, now i've got my MiniAxon is there anything you'd recommend i read up on before applying? |
[19:38] < MS-> I'd suggest, given you've got time, |
[19:38] < MS-> playing with Matt's code in /Sketches/MH |
[19:39] < MS-> Largely because Matt's done all sorts of things which could be interesting from a GSOC perspective |
[19:39] < Davbo> Got 3 weeks off Uni |
[19:39] < MS-> but he didn't move over to /Code for whatever reason |
[19:39] < MS-> There's interesting stuff in there like gesture recognition, open gl based "page turning" and so on |
[19:40] < Davbo> It's not necessarily about gsoc i've just got myself interested now - especially with Python |
[19:40] < MS-> in my area /Sketches/MPS theres some interesting stuff there as well, but I'm trying to package up |
[19:40] < MS-> Ah , cool |
[19:41] < MS-> I think one book (which is available online) that I'd read if I was in your position would be "how to think like a computer scientist using python", simply because its a well written book (ie very skimmable) |
[19:41] < Davbo> Just read the cookbook pages for Pipelines Graphlines and Carousels |
[19:42] < MS-> I think also I'd read through the /Cookbook pages as well and have a ponder |
[19:42] < Davbo> lol |
[19:42] < Davbo> :) |
[19:42] < MS-> The Open GL examples are very interesting/fun as well |
[19:43] < Davbo> We've not started 3d graphics in my course yet so i'm curious with Open GL |
[19:43] < MS-> I mean, I know I've got some clear ideas of what I would like to see, (multi-window image editor/manipulation/video annotation/simple editor), but that's the interesting thing |
[19:43] < MS-> I see |
[19:43] < MS-> Have you ever heard of Logo? |
[19:43] < Davbo> nope |
[19:43] < MS-> no problem |
[19:44] < MS-> Stack based languages? |
[19:44] < MS-> in the style of forth or postscript? |
[19:44] < Davbo> umm no :( |
[19:44] < MS-> no problem - just looking for a point of reference really |
[19:44] < Davbo> i've heard of Stack based languages |
[19:45] < MS-> I was just going to suggest that, in a way, open GL is a bit like that. |
[19:45] < MS-> Essentially open GL has a concept of "the current model transform" |
[19:45] < MS-> which is a combination of this: |
[19:45] < MS-> [translate]*[scale]*[rotate]*[transform]*[scale] |
[19:46] < MS-> which since they're internally modelled as matrices boils down to a single matrix |
[19:46] < Davbo> Ah right |
[19:46] < MS-> When you add a new transform (with a "begin"), it extends that transform |
[19:46] < Davbo> I wonder if MATLAB can do such things |
[19:46] < MS-> [translate]*[scale]*[rotate]*[transform]*[scale]*[.....] |
[19:46] < Davbo> sounds similar |
[19:46] < MS-> and when it "ends" it pops one |
[19:47] < MS-> yep, matlab will do that sort of thing (I've not used matlab) |
[19:47] < Davbo> it's horrible. |
[19:47] < MS-> So in opengl you end up defining everything in terms of local co-ordinate systems, and then let it handle transformation into the 3D space |
[19:47] < MS-> which is fairly friendly |
[19:48] < Davbo> oh okay |
[19:48] < MS-> as long as you know what's going on :-) |
[19:48] < Davbo> Maybe a look at an old version of the Red Book? |
[19:48] < Davbo> or is that too advanced you think? |
[19:49] < Davbo> I heard it was the only book you need for Open GL |
[19:51] < Davbo> http://openbookproject.net//thinkCSpy/ |
[19:54] < MS-> The Red Book is fine IMO |
[20:02] *** vmlemon__ has joined #kamaelia |
[20:02] *** Uraeus has joined #kamaelia |
[20:02] < vmlemon__> Hi |
[20:04] < Davbo> Hi vmlemon__ and bye! i'm off to do the pub quiz |
[20:04] < Davbo> bye all |
[22:35] *** vmlemon__ is now known as vmlemon_ |
[23:12] *** robertofaga has joined #kamaelia |
[23:28] *** Davbo has joined #kamaelia |
[23:55] < vmlemon_> Hi |