| *** tpb has joined #tp | 00:00 | |
| *** ChanServ sets mode: +o tpb | 00:00 | |
| *** andrei has joined #tp | 00:25 | |
| andrei | Hi, anyone related to SoC around? | 00:25 |
|---|---|---|
| andrei | I'm interested in asking about the AI project | 00:25 |
| andrei | In case I go to sleep before anyone answers; is the AI considered high priority? | 00:27 |
| andrei | I'm really interested in doing it; I'm a PhD student and my research area is AI | 00:28 |
| JLafont | andrei: Thats awesome! If mine doesn't go in I might still bug you for info on AI stuff :) | 00:31 |
| JLafont | andrei: I'm working on the AI thing too. I've been mostly talking with mithro about AI. You should probably start by checking out tpsai-py to kinda get a feeling for it. Mithro also suggested that you try to work on an AI for an existing Ruleset. | 00:32 |
| JLafont | andrei: As to how high of a priority it is; I am not entirely sure. | 00:33 |
| JLafont | hopefully someone who knows answers soon | 00:33 |
| andrei | Hmm, if I can find someone and talk to them soon I might be able to start working on it early | 00:35 |
| mithro | hello andrei | 00:35 |
| mithro | I'm a little busy at the moment so I can't really talk | 00:35 |
| mithro | but feel free to ask questions and I can answer them in a bit | 00:35 |
| JLafont | Yay! he's back | 00:35 |
| andrei | mithro, Ah, sure :) | 00:35 |
| andrei | mithro, My first question would be, which ruleset do you prefer? | 00:36 |
| mithro | andrei: we have an AI for MiniSec | 00:36 |
| mithro | so it would have to be RFTS or MTSec | 00:37 |
| mithro | The RFTS is better developed so would be less of a moving target | 00:37 |
| mithro | AI is a fairly high priority - as it gives players something to play again | 00:38 |
| andrei | mithro, Awesome, (yeah, the weak AI bugs me in freeciv) | 00:39 |
| [1]ApacheChief | Just taking a look at the libtpproto-php I really think I could do it :) | 00:39 |
| andrei | mithro, Do you have any preferences as to how things should be implemented? | 00:39 |
| [1]ApacheChief | Seems like all the hard work is done, and pretty similar to server/client stuff I've done before | 00:39 |
| andrei | mithro, In terms of preferred algorithms, or classes of algorithms? (more probabilistic, more ML, more tweak-your-own) | 00:40 |
| mithro | [1]ApacheChief: I would recommend prototyping stuff, if you could show something which logs in and shows a list of objects it would be a killer application | 00:40 |
| mithro | andrei: the AI must be a standard tp client, it can not cheat | 00:41 |
| [1]ApacheChief | Mithro: Great! I think I could do that pretty easily -- just need to learn more about TP :) | 00:41 |
| *** bddebian has quit IRC | 00:41 | |
| andrei | mithro, I wasn't thinking cheating :) (I really don't like it when they cheat, and no self respecting AI person would do that) | 00:41 |
| [1]ApacheChief | As for web design skill, this is the last site I made a long time ago for my buddy's band: http://kidkilgore.com/examples/Reverend%20Slush/ | 00:42 |
| tpb | <http://ln-s.net/1hPu> (at kidkilgore.com) | 00:42 |
| [1]ApacheChief | Wouldn't be surprised if it's broken in IE or a different version of FireFox, could barely get it to work in IE6/FF =\ | 00:42 |
| mithro | andrei: some people want to embedded the AI in one of the servers - this is a no-no | 00:42 |
| andrei | mithro, Ah, yeah.. I've seen that in a lot of places (that's how most commercial games do it) | 00:43 |
| andrei | mithro, Also, the language that the AI is written in. I'm fine with writing it in C, C++, python, etc. (or that class of languages); Might I suggest scheme? (it fits very well with writing AI-related code) | 00:45 |
| mithro | using a less common language is going to be a problem | 00:46 |
| mithro | as it means if you run off, nobody is going to maintain/improve/etc it | 00:46 |
| andrei | mithro, Ah, okies. Is C++ ok? (I don't write OO code, I just prefer a C-like subset that has some templates and a few other nice features) | 00:46 |
| andrei | mithro, I ask because the two sample AIs were C and Python | 00:51 |
| mithro | C++ is fine | 00:51 |
| mithro | so is ruby | 00:51 |
| andrei | mithro, I've only read ruby code, and read the occasional paper about it, so I'd much rather use C++. | 00:52 |
| andrei | mithro, What's your library policy? I tend to make heavy use of boost and gsl | 00:53 |
| mithro | anything which is apt-get able in ubuntu is fine | 00:54 |
| mithro | anything which is small enough to ship with your code is fine too | 00:55 |
| mithro | other things are probably fine too | 00:55 |
| mithro | personally I don't know why anyone would want to write in C++ - but I won't hold it against you | 00:55 |
| mithro | it's not like you are short on time | 00:56 |
| andrei | mithro, Heh, I like C++, that along with scheme and haskell are my favourite languages | 00:56 |
| andrei | mithro, I write high level code (using maps, lambdas, etc) in C++ which cuts development time by quite a bit | 00:57 |
| mithro | I find my efficency in C++ is like 50% less then in python | 00:57 |
| andrei | mithro, I'm decently close to my scheme output (though not close to my haskell one) | 00:59 |
| andrei | mithro, What kind of AI are you looking for? | 00:59 |
| mithro | andrei: one which can play the game to a level which is challenging to a majority of players | 00:59 |
| andrei | mithro, I mean, in terms of algorithms :) One can go the route of probabilistic algorithms vs machine learning algorithms that learn how to play better (each has its upsides); handcrafted rules vs rule databases and inference; you can go the more usual route which is hack up handcrafted rules with lots of variables (which is what civ games usually do; although I think it's a bad idea, especially for the long term) | 01:02 |
| mithro | I perfer deterministic AI's :) | 01:02 |
| mithro | extendable AI's are preferred - but it's a complicated problem | 01:03 |
| mithro | Thousand Parsec supports a wide varitey of games | 01:03 |
| andrei | mithro, Heh, that's my research area :P (the grant I'm on is getting a robot to learn how to play complicated games like chess from scratch) | 01:03 |
| mithro | chess is simple compare to most thousand parsec games :) | 01:04 |
| andrei | mithro, In the research it also has to figure out the rules to the games, which is the really hard bit | 01:04 |
| greywhind | mithro: so you say... but have any of us actually _played_ a Thousand Parsec game? :P | 01:05 |
| JLafont | I have | 01:05 |
| mithro | greywhind: I've played Minisec :) | 01:05 |
| mithro | greywhind: and I'm slowly figuring out TP RFTS | 01:05 |
| mithro | (as you should know from the commits :) | 01:05 |
| greywhind | mithro: i'd actually like to try to play Minisec or RFTS at some point | 01:05 |
| JLafont | I was playing myself on minisec the other day | 01:05 |
| mithro | greywhind: I'm happy to give you a game sometime | 01:06 |
| mithro | we will both be crushed by tpsai-py however ;) | 01:06 |
| greywhind | mithro: we should get my advisors in on it too | 01:06 |
| mithro | I think we need to setup Minisec so you get a bunch of ships on the game start rather then a single one | 01:06 |
| greywhind | they want to try the game out as well. | 01:07 |
| mithro | would make the first phase of the game much faster | 01:07 |
| JLafont | mithro: Yes please! Minisec starts reeeaaaallllyyy slow | 01:07 |
| mithro | JLafont: it should be less slow with the Ctrl-T now | 01:07 |
| JLafont | mithro: Took me like 10 turns to colonize my first planet | 01:08 |
| JLafont | mithro: What does Ctrl-T do now? | 01:10 |
| mithro | causes a turn to end on a correctly configured server | 01:10 |
| greywhind | mithro: when you say "causes a turn to end" do you mean "if one player presses Ctrl+T, the turn ends for everyone"? | 01:11 |
| JLafont | mithro: Awesome! Does every player have to do it before it will end the turn? Or is it more for playing alone with AI | 01:11 |
| mithro | greywhind: depends on the server configuration | 01:11 |
| mithro | normally once a "threshold" of players to a Ctrl-T the EOT will be shorten to something like 10 seconds | 01:12 |
| andrei | mithro, Would a proposal for something based around inference in rules databases (which would lay the groundwork for getting it to work with other rulesets) + graphs to do more inference on (possibly use some nice machine learning algorithms, lay the groundwork for more general future AIs that would learn how to play better (they'd be probabilistic but would be great for showing off new games and trying out different rulesets)), with a plugin ar | 01:12 |
| andrei | chitecture to put in a dag of planners? (so that you can have tiny planning modules each responsible for simple things; I find this keeps things modular and clean) | 01:12 |
| andrei | erm, that should end in ", go over well?" | 01:13 |
| mithro | andrei: don't assume we have any idea about real AI :) | 01:13 |
| mithro | andrei: take a look at tpsai-py and then run away screaming ;) | 01:13 |
| andrei | mithro, Heh, I did :) | 01:13 |
| mithro | tpsai-py is something I wrote in about 3 days | 01:14 |
| mithro | it's pretty good at playing minisec | 01:14 |
| JLafont | I hope to dethrone tpsai-py as the current minisec champion! You just wait! | 01:14 |
| andrei | mithro, Nice; it will be amusing to watch the various AIs play each other | 01:15 |
| mithro | tpsai-py tends to win because it doesn't sleep however ;) | 01:15 |
| greywhind | mithro: could set the turn length to a day, or something like that, and make a long-term game | 01:16 |
| greywhind | then the AI wouldn't have the advantage of playing turns where nobody else is awake | 01:16 |
| andrei | mithro, I'll remove the AI-heavy bits from the above then. Would an architecture that has rules encoded in a database (prolog-like), where you get to write tiny planners that each contribute to the overall goals (so you can write something that takes care of getting a loaded ship to one place, now some other planner can invoke that previous one and you don't have to duplicate code), be ok? | 01:17 |
| JLafont | greywhind: That would be a very loooooong game | 01:17 |
| greywhind | JLafont: true... | 01:17 |
| JLafont | PROLOG is great. | 01:17 |
| JLafont | greywhind: Takes like 2 turns to build 1 ship | 01:17 |
| andrei | JLafont, Yeah, but they weren't happy with scheme, they'll be way less happy with prolog. So I'll just use an inference engine in C++ :) | 01:18 |
| JLafont | andrei: that sounds painful. heh. Be sure to remember that the AI should also be moderately fast. | 01:18 |
| andrei | JLafont, prolog is actually really fast :) | 01:18 |
| mithro | andrei: we are not anti-scheme, we have TPCL which is scheme based | 01:19 |
| mithro | it's just a maintance nightmare | 01:19 |
| JLafont | andrei: Not compared to say... lisp/python/scheme/c++ | 01:19 |
| andrei | JLafont, There are even variants that run in realtime (that run robots, it's quite cool) | 01:19 |
| JLafont | andrei: did not know that! Thats sounds great | 01:19 |
| andrei | JLafont, Well, by real time I mean wallclock time (not realtime as in the actual definition) | 01:20 |
| andrei | mithro, Heh, it's ok; I don't mind either. My C++ code tends to look scheme-like anyway (wish this was post-C++0x though, I look forward to getting lambdas in the language) | 01:20 |
| JLafont | andrei: Well yeah, declarative languages will always have that speed constraint haunting them | 01:21 |
| andrei | JLafont, Actually; I think it's quite the opposite. I suspect in ~10-15 years you'll find declarative languages that are faster than imperative ones | 01:21 |
| andrei | JLafont, If only because compilers are getting better at doing whole program analysis and humans aren't | 01:22 |
| andrei | JLafont, (and you can play really fun tricks with parallelism in declarative languages) | 01:22 |
| JLafont | andrei: Most declarative languages are interpreted though. But yeah, I can see them getting faster, although I dunno about them being faster than C | 01:26 |
| andrei | JLafont, It really depends what interpreted means. VMs these days have done a good job of blurring the line between interpreters and compilers | 01:28 |
| andrei | JLafont, I think the big issue is that the programming models are getting way more complicated. Imperative code is fine when your execution model is simple, but parallelism is really hard. So compilers for declarative languages are going to get to do a lot of nice things for parallelizing code that just can't be done in imperative languages | 01:30 |
| mithro | I really wish I had learnt prolog and haskell | 01:33 |
| JLafont | andrei: Yeah, I see what you are getting at. | 01:33 |
| JLafont | mithro: I wanna learn haskell too :(. prolog isn't that hard. Its a very simple language | 01:34 |
| andrei | JLafont, Similar to what happened to assembly a while back. Writing hand crafted x86 assembly while trying to beat icc/gcc is really hard (I used to write real time raytracers). The cpu got really complicated so almost everyone isn't going to gain anything by writing assembly these days | 01:34 |
| JLafont | Anyways, I'll discuss languages some more with you later andrei. I'm gonna go sleep | 01:34 |
| andrei | JLafont, Heh, night :) | 01:35 |
| JLafont | andrei: They still use assembly... its just... painful | 01:35 |
| JLafont | but yeah, I'll talk to you later | 01:35 |
| * mithro does simulation at work, in most processors 90% of the instructions are NEVER used :) | 01:35 | |
| *** JLafont is now known as JLafont-ZzZ | 01:35 | |
| andrei | mithro, You should learn haskell. It's a great language :) | 01:35 |
| andrei | mithro, Nice, what type of simulation? | 01:36 |
| andrei | I should go to sleep as well | 01:47 |
| andrei | I'm really glad there are AI-related projects in SoC | 01:47 |
| andrei | Night! | 01:47 |
| *** tuna has quit IRC | 02:16 | |
| *** tuna-fish has joined #tp | 02:19 | |
| *** tuna-fish is now known as tuna | 02:22 | |
| mithro | well, I'm heading of home now | 03:07 |
| mithro | see you all later | 03:07 |
| *** mithro has quit IRC | 03:15 | |
| *** mithro has joined #tp | 03:57 | |
| mithro | greywhind: it works, but you need to resize the pop-up to fix it's contents | 04:08 |
| *** Epyon has quit IRC | 04:19 | |
| *** [1]ApacheChief has quit IRC | 04:20 | |
| *** llnz has joined #tp | 04:23 | |
| llnz | hi all | 04:26 |
| *** JLafont_ has joined #tp | 05:57 | |
| *** JLafont-ZzZ has quit IRC | 06:13 | |
| JLP | morning all | 06:30 |
| llnz | hi JLP | 06:35 |
| llnz | tp foo | 06:40 |
| llnz | humm... | 06:40 |
| mithro | tp foo? | 06:56 |
| llnz | trying to see if a highlighter was working in my irc client | 06:57 |
| llnz | mainly to know when someone is interested in tp on #gsoc | 06:57 |
| llnz | try again, please? | 06:58 |
| llnz | my first attempt highlighted every url | 07:06 |
| *** remaxim has joined #tp | 07:08 | |
| remaxim | hi | 07:09 |
| remaxim | ping mithro | 07:09 |
| *** remaxim has left #tp | 07:12 | |
| *** remaxim has joined #tp | 07:12 | |
| llnz | hi remaxim | 07:26 |
| * llnz pokes mithro | 07:26 | |
| remaxim | thx llnz | 07:26 |
| llnz | he was here 30 minutes ago | 07:27 |
| *** ApacheChief has joined #tp | 07:35 | |
| llnz | hi ApacheChief | 07:36 |
| remaxim | this name remembers me of http://www.youtube.com/watch?v=hYpXNG9WwI0 | 07:52 |
| tpb | Title: YouTube - Apache Chief (at www.youtube.com) | 07:52 |
| llnz | humm... | 07:54 |
| remaxim | I couldn't find the family guy snippet ;) | 07:59 |
| * llnz pokes mithro | 08:05 | |
| llnz | again | 08:05 |
| mithro | kind of busy.... will be back later | 08:05 |
| llnz | well remaxim, you can either wait, or write it now for mithro to read later | 08:07 |
| llnz | me should be asleep | 08:07 |
| remaxim | thx | 08:07 |
| remaxim | mithro, just wanted to ask if you searched for the source files and deleted them | 08:08 |
| * llnz wanders off | 08:10 | |
| llnz | later all | 08:10 |
| *** llnz has quit IRC | 08:10 | |
| *** niphree has quit IRC | 08:30 | |
| *** flyankur has joined #tp | 08:49 | |
| JLP | flyankur: ahoy | 08:50 |
| flyankur | hey | 08:53 |
| flyankur | what is the scope of the devlopment.. | 09:00 |
| JLP | flyankur: i'm not sure i understand what exactly you mean by scope | 09:01 |
| flyankur | what would be best for me... rulesset( game) devlopment.. or TP client devlopment... COz i was also intreseted in it | 09:01 |
| remaxim | bye | 09:02 |
| *** remaxim has left #tp | 09:02 | |
| JLP | flyankur: hm i think it is hard for me to judge this, you have to think what would be more interesting to you | 09:02 |
| flyankur | ok .. if you could describe what task would be there in client dev. and what are requirements on my side... | 09:04 |
| JLP | flyankur: it depends on which language you use, for C++ and Python there are already quite mature protocol libraries so you could already use them to write the GUI and use protocol libraries for TP specific stuff | 09:05 |
| JLP | flyankur: if there is no protocol library you would have to also write this | 09:06 |
| flyankur | i would like to work.. on MTsec .. | 09:36 |
| flyankur | plz guide me | 09:36 |
| JLP | ruleset then, in this case it is best to check out an existing MiniSec ruleset in one of the servers | 09:38 |
| flyankur | ok. | 09:39 |
| *** mithro has quit IRC | 09:40 | |
| *** flyankur has quit IRC | 09:44 | |
| *** bddebian has joined #tp | 09:54 | |
| bddebian | Howdy | 09:55 |
| JLP | bddebian: ahoy | 09:55 |
| bddebian | Hi JLP | 09:56 |
| JLafont_ | I have arisen! | 10:53 |
| JLafont_ | bddebian: Hi! | 10:54 |
| *** JLafont_ is now known as JLafont | 10:54 | |
| bddebian | Hello JLafont | 10:57 |
| JLafont | I managed to find a way to run Linux on this laptop until I can get it fixed! mwuahahaha. Going to run TPSAI-PY so I can evaluate how to best annihilate him! | 11:01 |
| * JLafont is a little too excited | 11:04 | |
| JLP | JLafont: that's the way! teach this tpsai-py some respect! :) | 11:04 |
| JLafont | JLP: And hopefully I don't go crazy doing it | 11:06 |
| * JLP will be back in a few minutes, lunch time | 11:07 | |
| *** JLafont has quit IRC | 11:13 | |
| *** jm_ has joined #tp | 11:17 | |
| *** jm_ is now known as jmt | 11:20 | |
| *** jmt is now known as jmtan | 11:20 | |
| *** jmtan has quit IRC | 11:24 | |
| *** Lukstr has joined #tp | 11:27 | |
| *** JLafont has joined #tp | 11:32 | |
| *** Lukstr has quit IRC | 11:33 | |
| *** Lukstr has joined #tp | 11:35 | |
| *** jphr has joined #tp | 11:36 | |
| *** jphr has quit IRC | 11:38 | |
| *** Lukstr has quit IRC | 12:02 | |
| *** Lukstr has joined #tp | 12:05 | |
| *** JLafont has quit IRC | 12:13 | |
| * JLP is back | 12:24 | |
| *** MihailBalan has joined #tp | 12:31 | |
| *** Epyon has joined #tp | 12:31 | |
| JLP | MihailBalan, Epyon: ahoy | 12:32 |
| *** Lukstr has quit IRC | 12:33 | |
| JLP | MihailBalan: great to see you again | 12:33 |
| MihailBalan | hello JLP :) | 12:34 |
| *** MihailBalan has quit IRC | 12:35 | |
| *** jmtan has joined #tp | 12:42 | |
| jmtan | hi, anyone? | 12:42 |
| Epyon | anyone here :P | 12:42 |
| jmtan | oh, hey :) | 12:43 |
| jmtan | i'm interested in helping out with tp | 12:43 |
| Epyon | generaly or as part of GSoC? ;> | 12:43 |
| jmtan | but have not done gsoc before | 12:43 |
| jmtan | err yeah gsoc | 12:43 |
| jmtan | i was hoping to do a 3d client | 12:43 |
| Epyon | Well, the first thing would be to take a look at the ideas page | 12:44 |
| *** flyankur has joined #tp | 12:44 | |
| Epyon | Oh, exciting! | 12:44 |
| *** flyankur has left #tp | 12:44 | |
| jmtan | i have some experience with python-ogre | 12:44 |
| Epyon | Do you know of thousand parsecs workings? | 12:44 |
| jmtan | having done a few of my own mini-projects | 12:44 |
| jmtan | i'm still reading through the docs | 12:45 |
| JLP | jmtan: ahoy | 12:45 |
| jmtan | hi jlp | 12:45 |
| Epyon | A pyOgre client has been started a long time ago, have you taken a look at it? | 12:45 |
| Epyon | Oh, the bigguys are back, so my work is over ^^ | 12:45 |
| jmtan | nope, to be honest i have a lot of catching up to do :) | 12:45 |
| jmtan | i hadn't heard of this project until i saw the gsoc page, but it is certainly exciting | 12:46 |
| Epyon | jmtan, last advice -- start catching up before writing the app, so you can show some knowledge of the task ahead :) | 12:46 |
| JLP | Epyon: you can continue even when the "bih guys" are here :) | 12:47 |
| JLP | Epyon: you're giving great advices to jmtan | 12:48 |
| Epyon | JLP, well, I walked that path before :D | 12:48 |
| jmtan | will there be any 3d assets needed for the client? from what i can see there are some static planet images only | 12:49 |
| Epyon | Although I'm starting to be afraid that I wont make it in this year :/ | 12:49 |
| Epyon | jmtan, try the media repository | 12:49 |
| Epyon | there are some models there, and if you'll need more, just ask :> | 12:50 |
| JLP | jmtan: yeah, Epyon is our 3D master | 12:50 |
| jmtan | haha ok | 12:50 |
| Epyon | modeler* :P | 12:50 |
| jmtan | may i know if the pyogre client is abandoned? | 12:51 |
| Epyon | Placeholder models can be done really quickly. And if the client will start working I'm sure that new and nicer models will start to appear | 12:51 |
| JLP | jmtan: in need of much love | 12:51 |
| JLP | jmtan: the python guy, mithro, is working on a 2d client, and noone else had the time for pyogre client | 12:52 |
| JLP | jmtan: but then agin pyogre client shares the same thousand parsec protocol libraries with 2d client | 12:53 |
| jmtan | great, my plan is to look through the pyogre client and port it over to python-ogre if possible | 12:53 |
| JLP | jmtan: seams like a great plan | 12:54 |
| JLP | jmtan: and also look at libtpproto-py | 12:54 |
| JLP | and maybe even libtpclient-py (the python client library) | 12:54 |
| Epyon | It'd be nice if a need for models would appear while doing that too :). | 12:54 |
| Epyon | There's a world of difference in doing models that will be sitting in a media repo as opposed to doing models that will be actually be used in a working program :P | 12:55 |
| jmtan | what kind of models do you think will be needed? | 12:55 |
| Epyon | jmtan, that much depends on how the 3d client will be executed | 12:56 |
| JLP | jmtan: and it also depends on the actual game ruleset | 12:56 |
| Epyon | however, don't worry about the models -- you do the excellent code, and we'll try to provide cra... err... excellent models for it ^_^ | 12:57 |
| JLP | exactly | 12:57 |
| jmtan | alright, i guess i might be thinking too far ahead :) | 12:57 |
| Epyon | A well documented and researched 3d client proposal would definitively be on the list of the most welcome apps | 12:58 |
| jmtan | i'm not too sure about the structure of the proposal, i guess it should be as detailed as possible? | 12:58 |
| JLP | yup it would be great to have some sexy looking client | 12:58 |
| Epyon | jmtan, not exactly | 12:59 |
| JLP | jmtan: just send it in as soon as possible | 12:59 |
| JLP | jmtan: we will help you make it as good as possible, but the more time the better | 12:59 |
| Epyon | jmtan, it should show that you understand on what you are undertaking, and that you have a sound plan for doing it -- that maximizes a positive result chance :) | 12:59 |
| jmtan | so i will be able to modify it after sending | 13:00 |
| jmtan | ? | 13:00 |
| JLP | jmtan: yes | 13:00 |
| Epyon | jmtan, only the second (private) part afaik. The public part stays frozen. But I maybe wrong | 13:00 |
| JLP | jmtan: the basic thing to keep in mind is to write what you willl do, why, how and put in a good time plan | 13:01 |
| jmtan | great to hear. the sending will be done via the gsoc page? or should i email the devs | 13:01 |
| Epyon | It wouldn't hurt to show the app to the mentors for suggestions before submitting it to the google app though :) | 13:01 |
| JLP | jmtan: it ends up on gsoc page | 13:01 |
| JLP | jmtan: but you can e-mail us a draft (mithro would probably be the best person) even before you put the application into gsoc system | 13:02 |
| *** Erroneous has joined #tp | 13:03 | |
| JLP | Erroneous: hi | 13:03 |
| jmtan | ok, thanks for the info | 13:03 |
| Epyon | Erroneous, are you mentoring this year again? xP | 13:03 |
| Erroneous | JLP: hi | 13:04 |
| Erroneous | Epyon: for bzflag, yes | 13:04 |
| JLP | jmtan: no problem, if you have any more questions feel free to ask | 13:05 |
| Epyon | Erroneous, seen it? : http://bzflag.chaosforge.org/images/cowtank-promo.png | 13:05 |
| tpb | <http://ln-s.net/1haO> (at bzflag.chaosforge.org) | 13:05 |
| Epyon | xP | 13:05 |
| Erroneous | yeah, saw it... like it | 13:06 |
| JLP | nice | 13:06 |
| * JLP think about CowSpaceShip :) | 13:07 | |
| Epyon | JLP, I definitively want to put that one on a coffee mug :P | 13:07 |
| JLP | there could be a pink one MilkaCowSpaceShip | 13:07 |
| Epyon | Shouldn't be much of a problem, just changing black to madcow-violet :P | 13:08 |
| *** JLafont has joined #tp | 13:38 | |
| *** Skuz has joined #tp | 13:49 | |
| Skuz | hi everyone | 13:49 |
| JLP | Skuz: ahoy | 13:55 |
| JLP | Skuz: interested into TP for Google Summer of Code? | 13:56 |
| Skuz | yep | 13:59 |
| JLP | Skuz: cool, then you have come to the right place :) | 14:00 |
| JLP | Skuz: already checked out the ideas page? | 14:00 |
| Skuz | I am currently "browsing" all the ideas | 14:01 |
| JLP | Skuz: ok, if you have any questions feel free to ask | 14:03 |
| *** Skuz is now known as skuz] | 14:30 | |
| *** Lukstr has joined #tp | 14:32 | |
| *** JLafont has quit IRC | 14:34 | |
| JLP | skuz]: MTsec is runing on demo2 - http://metaserver.thousandparsec.net/ | 14:36 |
| tpb | Title: Thousand Parsec : Metaserver Server Listing (at metaserver.thousandparsec.net) | 14:36 |
| JLP | skuz]: but keep in mind that MTsec is not finished and still needs a lot of love :) | 14:37 |
| skuz] | I ran it,it seems fun | 14:42 |
| JLP | skuz]: it should be even more fun when it gets finished :) | 14:44 |
| skuz] | if you can get graphics as in haegemonia it could be very nice | 14:46 |
| JLP | skuz]: yeah, it would be great to home Google Summer of Graphics and Sound :) | 14:47 |
| JLP | skuz]: but first we must finish the gameplay | 14:48 |
| skuz] | what can you do so far ingame? | 14:48 |
| *** jphr has joined #tp | 14:49 | |
| JLP | if jmtan gets selected we might get an opengl client and with help from Epyon who will make models it could even be looking nice in a few months :) | 14:49 |
| JLP | jphr: ahoy | 14:49 |
| jphr | how many students can TP have? | 14:49 |
| jphr | JLP: hey | 14:49 |
| JLP | jphr: don't know, last year we got 3 spots, google allocates them | 14:50 |
| JLP | skuz]: protocol defines what is possible in games | 14:50 |
| jphr | JLP: oh ok, does TP order their selections and google picks x of the top ones to be in? | 14:50 |
| JLP | jphr: yup, that's basicly how it goes | 14:51 |
| jphr | JLP: how many students applied last year? | 14:52 |
| JLP | skuz]: http://www.thousandparsec.net/tp/dev/documents/protocol3.php <- here is the version 3 of the protocol | 14:53 |
| tpb | <http://ln-s.net/Ix2> (at www.thousandparsec.net) | 14:53 |
| JLP | skuz]: version 4 will be out soon: http://www.thousandparsec.net/wiki/TP04 | 14:54 |
| tpb | Title: TP04 - Thousand Parsec Wiki (at www.thousandparsec.net) | 14:54 |
| skuz] | JLP: ok i will yake a look | 14:54 |
| JLP | jphr: hm, i forgot the number, too bad google doesn't keep the page from last year | 14:54 |
| jphr | JLP: yeah | 14:55 |
| jphr | JLP: do you know of any minor projects I could wet my teeth on b4 applications? | 14:56 |
| JLP | jphr: aha found it, there were 46 applications | 14:57 |
| jphr | JLP: woah, thats a lot of competition | 14:57 |
| JLP | jphr: yeah, quite a lot of effor needs to be put into application to be at the top of the list | 14:58 |
| JLP | jphr: but no fear, so the sooner it gets into our hands the more time there is to give suggestions to improve it | 14:59 |
| jphr | JLP: Well i'd like to throw together a wiki page on the diplomacy ruleset, similar in detail to xdtox's from last GSoC | 14:59 |
| jphr | JLP:I can probably get that up mid friday, with the schoolwork I have going now | 15:00 |
| JLP | jphr: yup that would be a great start, and we can then make it even better then xdotx's :) | 15:00 |
| jphr | JLP: I've never actually played diplomacy, but I can see it working really well, it could be played on TP over a long time or short time, there just need to be a few improvements to the server to allow for chat and whispers | 15:01 |
| Epyon | skuz], I'll be sharing graphics that I do for my own game with TP, they may not be commercial level, but they hold their ground on an amateur level :) | 15:01 |
| JLP | jphr: yes i've seen your post on the forums | 15:03 |
| skuz] | that's cool | 15:03 |
| *** niphree has joined #tp | 15:03 | |
| jphr | JLP: any initial comments? | 15:03 |
| niphree | hello | 15:03 |
| JLP | jphr: i've just read quicky over it, will read it more carefully later and see i there is anything to add | 15:04 |
| JLP | jphr: but probably mithro and llnz will have more to say | 15:04 |
| JLP | niphree: ahoy | 15:04 |
| skuz] | JLP:what is the next required feature to add to MTSec? | 15:14 |
| JLP | skuz]: i'm not sure how far it is implemented, i think llnz was working on it so he would know what remains to be done | 15:15 |
| JLP | skuz]: or you can inspect the code in tpserver-cpp and compare it to MTsec documentation | 15:16 |
| cherez | Are last years SoC applications saved anywhere? | 15:18 |
| Epyon | cherez, the accepted ones, or all? | 15:18 |
| cherez | All. | 15:18 |
| cherez | Well, I'm only interested in mine. :P | 15:19 |
| Epyon | You'd have to ask the mentors :P | 15:19 |
| Epyon | I have to find mine too ^^ | 15:19 |
| Epyon | I have them somewhere on my HD or server I think | 15:19 |
| cherez | I can't find my offline copies, and it would be nice to have last year's as a basis. | 15:19 |
| Epyon | Ha! Mine are still on the server :D | 15:20 |
| JLP | cherez: mithro printed them out on paper | 15:21 |
| JLP | cherez: probably he stil keps them around, at least i would | 15:21 |
| Epyon | Wow! Paper! Was that the wierd scruffy thing used before screens were invented? D: | 15:22 |
| JLP | Epyon: yeah that's the thing, a real tree killer this mithro ccc | 15:22 |
| cherez | I think that's what people used to word process when there were too many trees in the world and they had to use the mup. | 15:23 |
| Epyon | cherez, I'll be resubmitting a heavily modified proposal from last year :) | 15:23 |
| cherez | Epyon: That sounds like a plan. :) | 15:25 |
| *** JLafont has joined #tp | 15:25 | |
| cherez | Epyon: Which project were you applying for last year? | 15:25 |
| Epyon | Well, both were original :D | 15:26 |
| Epyon | Oh, you mean orgs? | 15:26 |
| cherez | No, for TP. | 15:26 |
| Epyon | EliteSec and starship model generator | 15:26 |
| cherez | Ahh, right. | 15:27 |
| cherez | A model generator would be awesome. | 15:27 |
| cherez | Would be really nice if someone makes the 3D client too. | 15:27 |
| Epyon | I wont apply with that one again, it's madness :P | 15:27 |
| JLP | Epyon: you can still submit it and we can again hope that more applications would translate into more slots :) | 15:29 |
| JLafont | I was thinking of maybe a second proposal for the 3D client. | 15:29 |
| Epyon | JLP, okay, but I'll kill you if it gets selected :P | 15:29 |
| Epyon | But you may be right -- the more proposals the better for the org. | 15:29 |
| JLP | yup i guess it is good, google see oh this one got a lot of good applications let's give them a few slots more | 15:30 |
| cherez | JLP: What project ideas are the devs excited for? | 15:34 |
| andrei | Hopefully AI is one of them :) | 15:35 |
| JLP | cherez: rulesets are very high on priority list, AI too, and nice clients, so that the game finally gets as interesting and fun to play and attrackt players | 15:36 |
| cherez | Sweet, those are the projects I thought sounded fun to do. | 15:37 |
| Epyon | I want to see a client making use of 3d objects -_- | 15:39 |
| Epyon | Or ship graphics at all :P | 15:39 |
| JLP | my personal 3 most wanted things are 3D client, AI and a fun ruleset | 15:39 |
| cherez | I am sort of inclined to apply for 3d client, but I'd feel obliged to get good models too. | 15:40 |
| Epyon | JLP, you'd consider EliteSec fun? :P | 15:40 |
| Epyon | cherez, don't worry about that | 15:40 |
| Epyon | cherez, better think about how to solve the UI nicely :) | 15:40 |
| cherez | Epyon: I know, but it just feels like I should do it. :P | 15:41 |
| Epyon | cherez, http://chaosforge.org/sd_halberd.png | 15:41 |
| Epyon | cherez, http://chaosforge.org/latern-promo.png | 15:41 |
| Epyon | And more will come :P | 15:41 |
| Epyon | (untextured as yet) | 15:41 |
| JLP | Epyon: is elite like Privater: The Darkening? | 15:41 |
| JLafont | cherez: If I do a second proposal for a 3D client, I'd probably have the models be boxes or spheres. Someone else with artistic abilities can do the models. | 15:41 |
| cherez | Epyon: are those yours? | 15:42 |
| Epyon | JLP, actually it will be inspired by all Privateer, Elite, Frontier and Traveller | 15:42 |
| Epyon | cherez, yes. | 15:42 |
| JLP | Epyon: aha, then it sure sounds like a lot of fun | 15:42 |
| andrei | Out of curiosity; can't we use the work done by various science fiction fans? (there's an insane amount of say star trek or B5 ship models) | 15:43 |
| Epyon | andrei, nope | 15:43 |
| Epyon | andrei, licence issues :/ | 15:43 |
| JLP | andrei: probably, depends on the license of these models, if they are under something like GPL then it is fine to use them | 15:43 |
| * Epyon 'd be ripping of star wars ship designs otherwise xP | 15:44 | |
| JLP | :) | 15:44 |
| JLafont | I'm sure there are some public domain models out there | 15:44 |
| Epyon | JLP, you'd quickly get a C&D if you started to make a StarCraft ruleset :P | 15:44 |
| andrei | Epyon, JLP The usual way around that is not to bundle the ships with the game. Write a small downloader that the person has to click ok in before starting the game the first time | 15:44 |
| Epyon | JLafont, there are, but their quality varies. | 15:45 |
| Epyon | andrei, using Star Wars models for a non-starwars game would seem extremely silly and unproffesional for me | 15:45 |
| JLP | andrei: basicly how graphics are now downloaded for 2d client | 15:45 |
| JLP | except for the naging dialog about non-free content | 15:45 |
| JLafont | Epyon: I'm not surprised | 15:46 |
| andrei | Epyon, A lot of those fan sites make ships that are similar to the ones in the show but still different enough | 15:47 |
| andrei | JLP, Epyon There's also google 3d warehouse; they have a lot of sci-fi ship models | 15:48 |
| * Epyon itches to make a fleet for a 3d client -_- | 15:48 | |
| andrei | JLP, Epyon And the license is GPL-compatible | 15:48 |
| Epyon | andrei, the beauty of a good space game is that the ship design can tell you it's origin. That the designs share some common feel. That'd be lost if we'd use ships from all round the web. Unless we're wanting to make NetHackSec :P | 15:49 |
| *** skuz] has left #tp | 15:50 | |
| andrei | Epyon, That is true, but this provides a quick and easy to way to get lots of models and attract players as soon as the 3d client is up and running; before going on to make nicer ships :) | 15:51 |
| Epyon | andrei, well that's what I'm doing -- doing placeholder ships until we attract a real artist :P | 15:52 |
| JLP | well i guess ship models would not be such a problem, first we need to get a 3d client | 15:52 |
| Epyon | JLP, yep that's why I tell all interested in a 3d client that models are not a problem :) | 15:53 |
| JLP | so to all that are interested in applying with a 3d client: no need to worry to much about the models | 15:53 |
| Epyon | heh xP | 15:54 |
| JLP | Epyon: looks like we are on the same wavelength :) | 15:54 |
| andrei | Epyon, Heh, finding artists for open source projects seems to be hard (I'm really curious why universities don't slightly change their courses in order to help out open source projects) | 15:54 |
| JLafont | I'm mostly concerned with relearning Ogre >.< | 15:54 |
| Epyon | andrei, I believe that what if you're doing is good, good people will come to help you :) | 15:55 |
| Epyon | If it wouldn't work, then there'd be no OS games anyway. | 15:56 |
| Epyon | However, to be frank, I'm quite happy that true 3d artists aren't found that easy ^^ | 15:56 |
| andrei | Epyon, Oh? | 15:56 |
| JLP | more work foe Epyon | 15:57 |
| Epyon | I love modelling, and the lack of professional 3d artists allows my work to have a chance to be used :P | 15:57 |
| Epyon | How egoistical one my part D: | 15:57 |
| Epyon | on* | 15:58 |
| andrei | Ah, I have no artistic talent at all; so I stay away from anything related to it :) | 15:58 |
| JLP | andrei: same here | 15:59 |
| JLP | but my sister is in graphic design and i can't convince her at all to do anything for TP or KDE or something | 16:00 |
| JLafont | JLP: I hear blackmail works | 16:00 |
| andrei | JLP, Does she use Linux? | 16:00 |
| JLP | andrei: yup | 16:00 |
| JLP | JLafont: i've tried many things, but nothing worked, or maybe i'm to soft on her | 16:01 |
| JLafont | JLP: Sisters can be difficult like that >.< | 16:01 |
| andrei | Most programmers that I know just sit and bitch that some applet has broken python code and noone is fixing it (when it would take them 2 minutes to fix it and a few more to submit a patch). So it's nothing special :) | 16:02 |
| JLafont | 6 days until I get back to my computer! | 16:03 |
| andrei | JLP, I usually go the guilt route; that they're complaining about something that people are doing out of love. It usually works :) (I got one friend a few days ago to submit a few patches to ubuntu by doing this) | 16:05 |
| *** JLafont has quit IRC | 16:05 | |
| Ohm | heh | 16:05 |
| JLP | andrei: yeah i tried this tactic, didn't work so far, but i think it is the best one that has the highest chance of working on my sister | 16:06 |
| * Epyon looks in horror at his personal bugtracker. | 16:07 | |
| Epyon | 420 open tasks D: | 16:07 |
| * Epyon should definitively stop working solo -_- | 16:08 | |
| Ohm | I enjoy jumping in and trying to understand code when you want to just change some small thing | 16:09 |
| *** jphr has quit IRC | 16:09 | |
| JLP | Ohm: yeah, lots of things get learned this way :) | 16:11 |
| andrei | JLP, Epyon I don't think I mentioned this, but I'm interested in doing the AI (I was talking to mithro yesterday about it) | 16:11 |
| Ohm | mm, gotta go fix some food | 16:12 |
| andrei | I'm fine with working in a few languages, but he said that scheme wasn't the greatest because he feared there wouldn't be people around to maintain the code if I leave (which I don't plan on, I'm a PhD student in AI and I like hacking on these things :)) | 16:12 |
| andrei | I was wondering what other people thought about me using scheme (I'm ok with not using it, but the code will be much simpler to understand, change and improve if it's in scheme as it's such a good fit for wiriting AI-related code) | 16:14 |
| Epyon | andrei, sorry, can't help you on that one :). I can at most read your applications and give some hints on how to make it better :P | 16:14 |
| *** JLafont has joined #tp | 16:14 | |
| andrei | Also, which ruleset would be preferred for the AI (I'm planning on something rather general as that's closest to my research area, but I won't have time to port it to multiple rulesets)? | 16:16 |
| *** bddebian has quit IRC | 16:17 | |
| andrei | mithro suggested RFTS as it's less likely to change, but it doesn't bother me if the rulesets change as I go (indeed that's a good thing) | 16:17 |
| JLP | andrei: i don't know much about scheme, i only know that a small subset of it is used already in thousand parsec, for components and skip designs and things like that | 16:19 |
| Epyon | andrei, ideally it would be a framework for AI related methods that could be then extended for any game | 16:20 |
| andrei | JLP, I'd be willing to write a scheme tutorial, and one on how to extend the AI system if that helps scheme get accepted (I really think it'll be a better system if scheme is used) | 16:20 |
| JLP | andrei: but i understand mithro that it could be a problem if noone stayed at TP who would know it well enough | 16:20 |
| Epyon | Most of the TP rulesets share some similarities from the AI point of view | 16:20 |
| JLafont | Epyon: I thought the same thing too, but then mithro showed me TIM and I was put in my place :P | 16:21 |
| andrei | Epyon, Yup, it would use a rulebase and inference (as well as some graphs) (this is what I read papers on all day) but I just don't have time to write up the rule databases for multiple games and write inference engines and stategy engines to make them work nicely (eventually as more components are written new rulesets will just work out of the box) | 16:21 |
| JLP | andrei: tutorials like this would definitely help, and maybe even some assurance that you would remain with TP even after the GSoC is over | 16:22 |
| Epyon | andrei, that's one way to go about it :) | 16:22 |
| JLP | andrei: if you go the scheme route you definitely mention these plans in your application | 16:22 |
| JLP | andrei: and yes i would definitely do it for RFTS ruleset as it is currently the most advanced game we have that is close to finished | 16:24 |
| andrei | JLP, I would gladly promise to stay on afterwards and be an active developer (but such things are vacuous and I wouldn't take anyone's word on them). I can promise to always answer my email and answer questions about it regardless of what happens. I will most likely stay on. I see no reason why I wouldn't. But I've been involved in open source projects that went down badly | 16:25 |
| Epyon | BTW, RFTS is in python, right? | 16:25 |
| Epyon | the ruleset that is. | 16:25 |
| JLP | Epyon: c++ | 16:26 |
| andrei | Like alky (if anyone remembers that) where they decided to close source it (along with my work.. thankfully I hadn't released most of my threading work at that point, bastards) | 16:26 |
| JLafont | RFTS is in c++, but the AI can be in any language | 16:26 |
| JLafont | The AI is acts as a client | 16:26 |
| JLP | andrei: well all i can suggest is that you put these potential concernes into application, under Risk Management or something like that | 16:26 |
| JLP | andrei: like what happens if I get hit by the bus, what steps will i take that the impact is as minimal as possible if i'm gone | 16:27 |
| Ohm | cloning! | 16:27 |
| andrei | JLP, Sure, (it's not a major concern, just that I've been burned by open source projects changing licenses midstream and taking my code with them; I very much doubt that'll happen here though) | 16:27 |
| Ohm | projects can only do that if they have been granted ownership of the code | 16:28 |
| andrei | Ohm, Yeah.. well.. tell them that :p | 16:28 |
| Ohm | not if you have given them access to the code under <your choice of license> | 16:28 |
| Ohm | heh | 16:28 |
| Ohm | Call the FSF! | 16:29 |
| andrei | Ohm, It's ok; they already failed as a company :) | 16:29 |
| Ohm | hooray | 16:29 |
| Ohm | Eben Moglen to the rescue! | 16:29 |
| JLP | andrei: yeah i think we will stay for a long time, TP seams to be quite stable and stuborn, exists for quite some years and i think everyone here wants it to last forever | 16:30 |
| Ohm | he is one of the best speakers I've ever seen | 16:30 |
| andrei | Anyone have any preferences as to the AI? My preferred route is the whole ruleset database inference way + a very modular system to plug in different strategies in a directed graph. Although it seems a lot of civ-like games prefer much more hardcoded AIs with lots of hand-crafted values. | 16:31 |
| andrei | There's also the machine learning route | 16:31 |
| Ohm | I think the best thing to do would to make your AI implementation as portable as possible | 16:31 |
| Ohm | since I think everybody is hoping for lots of new rulesets as time goes by | 16:32 |
| andrei | Ohm, Yup, I actually want to make it so that I can eventually plug it into freeciv too :) | 16:32 |
| JLP | andrei: hm, can't say much about this, i know nothing about AI programming, as long as it feels like playing against human and has maybe the possbility to set how hard it plays, it is fine with me | 16:32 |
| andrei | JLP, Okies :) yup, what I'm proposing would easily be able to change difficulty levels; and it would be much more flexible than most game AIs. | 16:34 |
| JLP | andrei: cool, you're the pro when it comes to AI, so you probably know the most what is the best way to do it | 16:35 |
| andrei | For some reason games companies don't hire real AI people usually (or maybe by the time you know enough you've realized that working for a game company is soul destroying :P) | 16:35 |
| JLafont | andrei: Thats cause most real AI people don't take time into consideration | 16:35 |
| andrei | JLafont, Development time or running time? | 16:36 |
| JLafont | andrei: both | 16:36 |
| JLafont | They just run around scripting everything in game companies because it is "easier" | 16:37 |
| JLafont | blah! | 16:37 |
| JLafont | blah to them! I say | 16:37 |
| andrei | JLafont, I find most AI research is rather light on CPU usage. Almost every project runs on just one box (though there are a few that do insane things) | 16:37 |
| JLafont | andrei: Tell that to the game companies :). | 16:38 |
| JLafont | andrei: You have to remember that the rest of the game is also running at the same time | 16:38 |
| JLafont | And games nowadays are all about the pretty graphics | 16:38 |
| JLafont | No CPU time for having a fun game | 16:39 |
| andrei | JLafont, Yeah.. I'm bitter about the whole graphics thing | 16:39 |
| andrei | JLafont, It's really hard to find good games to play, they're all just shiny and boring | 16:39 |
| JLP | yeah me too, but that's what people want and it TP wants to be widely used it will have to have some sexy graphics | 16:39 |
| JLafont | I mean, pretty graphics help to catch your eye, but they only take the game so far. I'd rather have a friendly UI | 16:39 |
| JLafont | and fun gameplay | 16:40 |
| andrei | JLP, I'm not saying one shouldn't have nice graphics; but at least open source games start with the gameplay | 16:40 |
| andrei | Most games start by choosing a 3d engine.. | 16:40 |
| andrei | (commercial that is) | 16:40 |
| JLP | andrei: yeah i guess most of us here at TP are also more for gameplay, thats's why 3d cliet has been so neglected :) | 16:41 |
| andrei | JLP, That's why I'm interested in the AI. I used to write MUDs a long time ago; I used to use simple AIs to test out economy systems. With the ruledatabase system you can do a lot of nice things. Like use the AI to balance the game | 16:43 |
| andrei | (well, what I was doing back then with AIs was stupid; I now know I should have used some Markov chains and just proven that my economy system isn't crazy; but that was before university) | 16:44 |
| * JLP googles for Markov chanes, noting that he isn't even a CS student | 16:44 | |
| andrei | JLP, They're fairly simple. Make it really easy to see if a stochastic process will converge | 16:47 |
| *** JLafont has quit IRC | 16:47 | |
| andrei | And to what (you don't want your economy to baloon out of control or to shrink into nothingness) | 16:48 |
| JLP | ah i see, i think i get abit about it, something similar to statistics in physics | 16:50 |
| andrei | JLP, Yup, stochastic processes are used all over in physics | 16:51 |
| andrei | JLP, You mentioned you weren't a CS student, what field are you in? | 16:52 |
| JLP | andrei: physics, measurement | 16:52 |
| andrei | JLP, Nice; I was going to be a physics student (but opted for CS/Math instead) | 16:53 |
| *** JLafont has joined #tp | 16:54 | |
| andrei | (sometimes I still wish I'd gone into mathematical physics instead of CS :P) | 16:54 |
| JLP | andrei: i wish i decided for CS, over the years i have somehow lost much of the interest into physics and am gaining more and more interest in computers and computer science | 16:54 |
| andrei | JLP, You can still switch :); I know a lot of people that are doing that in both directions | 16:55 |
| JLP | andrei: i have no more option to change, i already did once (from theoretical physics to suposedely more practical measurement physics (not much difference at our une :()) | 16:56 |
| andrei | JLP, Ah, you can switch after you're done. Quite a few people in CS grad schools have physics degrees (in the workforce too) | 16:58 |
| JLP | so currently my plan is to continu to work at a software company here and save some money, and maybe get the company to pay me for some study of CS | 16:58 |
| andrei | JLP, Nice; I suggest going to a major university if you can. CS programs vary widely in quality | 16:59 |
| JLP | andrei: hehe not much choice here in little Slovenia, onl 2 universities here | 16:59 |
| JLP | andrei: but yeah maybe if i look EU-wide | 17:00 |
| andrei | JLP, EU-wide there are quite a few. The German ones are good (stay away from most UK ones; they're horrible) | 17:00 |
| andrei | JLP, One of my labmates is from the UK and his CS program is considered one of the best ones (it's ridiculously bad compared to a good N. American university) | 17:01 |
| Epyon | I was going to be an artist or musician, but I sucked so I became a CS student -_- | 17:01 |
| JLP | Epyon: :) | 17:01 |
| andrei | Epyon, Heh, computer generated art is a huge field these days :) | 17:02 |
| JLP | andrei: yeah i hear good things about german ones, my only concern is language, know almost no german | 17:02 |
| andrei | JLP, In the better ones most CS classes are in english | 17:03 |
| * JLafont is taking German this semester | 17:03 | |
| JLP | andrei: oh, realy, i didn't know this, then this would be perfect | 17:03 |
| * JLP notices that mithro should wake up soon 7:38 at his place | 17:05 | |
| JLafont | JLP: Make sure it gives you a good foundation of the fundamentals. Like data structures, discrete mathematics and automata theory. You can do just about anything once you know those. | 17:05 |
| * JLP notes down what JLafont said | 17:06 | |
| andrei | JLP, Yeah, avoid any place that doesn't have algorithms and theory of computation research groups | 17:08 |
| JLafont | Once I took those courses, I can say that I find most of my CS courses to be too easy. Even AI which is my favorite so far. | 17:08 |
| JLP | JLafont: btw, do you know any good online resources/books about these subjects? | 17:08 |
| Epyon | andrei, I generate cities xP | 17:08 |
| andrei | JLP, ocw is the greatest :) | 17:08 |
| JLafont | what he said | 17:09 |
| JLP | online courseware at mit right? | 17:09 |
| andrei | JLP, Yup | 17:10 |
| JLafont | http://ocw.mit.edu/OcwWeb/web/home/home/index.htm | 17:10 |
| tpb | <http://ln-s.net/1hfT> (at ocw.mit.edu) | 17:10 |
| JLP | i'm currently watching videos of Structure and Interpretation of Computer Programs | 17:10 |
| andrei | Nice, SICP is supposed to be one of the best resources out there (sadly by the time I heard of it I already knew scheme and what it covers so I've never read it) | 17:12 |
| JLP | i've seen the first part so far, but looks like i'm on the right track then | 17:13 |
| andrei | Definitely. If you have time get the book and work through some of the problems. Most CS graduates would benefit a lot from doing that :) | 17:14 |
| JLP | and looks like i will even get to learn scheme while watching, so this would make AI in scheme a little less risky | 17:14 |
| JLafont | Its all about the car and cdr | 17:15 |
| andrei | Heh, scheme is such a tiny language that learning it isn't much of an issue | 17:16 |
| andrei | JLafont, car and cdr rock. I laughed when I saw something like cadar for the first time. It's really neat. | 17:17 |
| JLafont | I think its amazing. I keep thinking of using it when I do C++ or Prolog | 17:18 |
| andrei | It should get much more doable in C++ after 0x | 17:18 |
| * JLafont wishes he could do caddr on PROLOG easily | 17:18 | |
| andrei | Heh, you've got pattern matching; that's close enough | 17:20 |
| andrei | I wish scheme had pattern matching | 17:20 |
| *** jphr has joined #tp | 17:46 | |
| *** jphr has quit IRC | 17:47 | |
| *** mithro has joined #tp | 17:55 | |
| JLP | mithro: morning | 17:57 |
| mithro | morning | 17:57 |
| mithro | anyway off to work | 18:05 |
| mithro | bblr | 18:05 |
| *** mithro has quit IRC | 18:28 | |
| *** mithro has joined #tp | 19:06 | |
| *** darrglud has joined #tp | 19:11 | |
| JLP | darrglud: ahoy | 19:16 |
| *** darrglud has left #tp | 19:17 | |
| *** darrglud has joined #tp | 19:17 | |
| mithro | greywhind: ping? | 19:21 |
| mithro | hey JLP | 19:24 |
| JLP | mithro: yo | 19:24 |
| mithro | JLP: how goes everything? | 19:25 |
| JLP | mithro: fine | 19:26 |
| mithro | JLP: any progress on the screencast front? | 19:26 |
| mithro | JLP: btw, you keeping track of the forums/ | 19:27 |
| JLP | mithro: i tried to play a bit with these xrc gui file to see if there is anything that can be done there to fix the dissapearing search | 19:27 |
| JLP | mithro: unfortunately i couldn't make it work | 19:28 |
| mithro | JLP: any idea how to get an RSS feed on the forums or something? | 19:28 |
| JLP | mithro: yup, i've seen the post from jphr, who is looking for thoughts about the diplomacy ruleset | 19:29 |
| mithro | there is a post from a person called Jabwock | 19:29 |
| mithro | ~seen xdotx | 19:29 |
| tpb | mithro: xdotx was last seen in #tp 1 day, 14 hours, 10 minutes, and 36 seconds ago: <xdotx> heya :) | 19:29 |
| JLP | mithro: u i didn't see this one yet | 19:29 |
| mithro | ~seen BZA | 19:30 |
| tpb | mithro: BZA was last seen in #tp 1 day, 12 hours, 34 minutes, and 35 seconds ago: <BZA> great, sent that off. One less thing to do for now =) g'night guys | 19:30 |
| * JLP switched to high latency time, star trek time :) | 19:31 | |
| mithro | JLP: I still can not reproduce that bug locally | 19:33 |
| JLP | mithro: i think i've found a workaround that should work, i moved the search bar above the tree widget | 19:49 |
| JLP | it started being almost one half of a panel high, but on resizing it resized to normal size | 19:50 |
| JLP | damn why didn't i think of this simple thing before, let's see what happens if i move to the bottom again | 19:52 |
| mithro | ? | 19:55 |
| mithro | are you working with or without the patch I posted? | 19:55 |
| JLP | mithro: currently without, with 0.3.0-inplace | 19:56 |
| mithro | so, I'm a little confused :/ | 19:56 |
| *** darrglud has quit IRC | 20:09 | |
| *** niphree has quit IRC | 20:17 | |
| *** bddebian has joined #tp | 20:28 | |
| bddebian | Howdy | 20:30 |
| cherez | mithro: Do you have my application from last year perchance? | 20:34 |
| mithro | cherez: no, I'm afraid not | 20:34 |
| JLafont | bddebian: Hey! | 20:34 |
| greywhind | mithro: pong | 20:34 |
| mithro | we don't have access to last years applications | 20:34 |
| bddebian | Hello JLafont | 20:34 |
| cherez | Well, guess I start from scratch again. | 20:35 |
| mithro | greywhind: your button seems to work | 20:36 |
| mithro | but it need to resize to fit it's contents | 20:36 |
| greywhind | mithro: cool. what size is it now? | 20:36 |
| *** llnz has joined #tp | 20:45 | |
| *** Erroneous has quit IRC | 20:55 | |
| mithro | greywhind: it's barely fits the contents | 21:03 |
| mithro | I can send you a screenshot if you wish | 21:03 |
| greywhind | mithro: please do | 21:03 |
| *** andrei has quit IRC | 21:06 | |
| mithro | greywhind: http://www.thousandparsec.net/~tim/Screenshot-7.png | 21:07 |
| tpb | <http://ln-s.net/1hj6> (at www.thousandparsec.net) | 21:07 |
| greywhind | mithro: i see | 21:07 |
| greywhind | mithro: i'll check it out | 21:07 |
| mithro | it's also not poping up in quite the right place | 21:07 |
| *** andrei has joined #tp | 21:09 | |
| mithro | morning andrei | 21:12 |
| Epyon | We have a very interesting proposal on tp-devel :) | 21:13 |
| Epyon | Try to drag that guy down to here on IRC :P | 21:13 |
| mithro | Epyon: feel free to reply to the mailing list :) | 21:18 |
| andrei | mithro, Hi; it's actually 9pm where I am :) | 21:18 |
| andrei | Do people around here use gentoo? (I'm writing ebuilds for some tp packages as I install them and I'm wondering if people care for me to release them) | 21:19 |
| Epyon | mithro, I can only reply from the adress I registered from right? :P | 21:19 |
| Epyon | andrei, I use Windows :P | 21:19 |
| mithro | Epyon: yes | 21:19 |
| JLP | andrei: i use gentoo | 21:20 |
| * Epyon loves to say that and imagine the grin of disgust on the other end :) | 21:20 | |
| Epyon | mithro, then I can't do that -- I lost my auths on the account, and can't reconfigure it :P | 21:20 |
| mithro | Epyon: sign up with a different account then | 21:21 |
| mithro | or give me a new email and I change change the destiniation of your account | 21:21 |
| andrei | JLP, Nice, ok; so I'll put the ebuilds somewhere then | 21:22 |
| Epyon | mithro, the problem is that all my accounts are broked so much that none of them can send, except the google account -- and I'm too lazy to fix it. | 21:23 |
| mithro | Epyon: I can map to your google account if you wish | 21:23 |
| * Epyon just realized that fixing it would take less than the time it took for him to write those two sentences D: | 21:23 | |
| * Epyon is really hopeless sometimes -_- | 21:23 | |
| Epyon | One moment, I'll try to fix them :> | 21:24 |
| mithro | JLP: don't forget to reply to people on the forums too :) | 21:25 |
| Epyon | LOL | 21:25 |
| Epyon | I fixed it :D | 21:25 |
| mithro | Epyon: yay! | 21:26 |
| *** JLafont_ has joined #tp | 21:27 | |
| JLP | mithro: will do, was distracted by wxFlexGridLayout, wxSizer .... documentation, trying to find out what could go wrong here | 21:27 |
| Epyon | mithro, I replied, however I feel quite uneasy posting on tp-devel with advice while not being a dev O.o | 21:32 |
| JLP | Epyon: if it is any like the advices you were gibing here then it should be fine | 21:32 |
| Epyon | JLP, I just advised him to come here :P | 21:33 |
| JLP | Epyon: well it is still helpful | 21:35 |
| *** JLafont has quit IRC | 21:45 | |
| andrei | Is the installer for the pywx client supposed to work? | 22:04 |
| andrei | (It doesn't install all the other libraries; just the game itself) | 22:05 |
| *** JLafont_ has quit IRC | 22:13 | |
| *** JLafont has joined #tp | 22:15 | |
| andrei | What's your preferred way of getting patches? | 22:22 |
| andrei | doc/tpclient-pywx-installed doesn't stop once it starts up the game (so in Gentoo's case after it's installed it goes to all the other places that are symmlinked to lib64 and starts the game 3 times in a row) | 22:23 |
| *** JLafont_ has joined #tp | 22:23 | |
| *** JLafont has quit IRC | 22:23 | |
| andrei | And, is there anywhere you have set up to host things like ebuilds? | 22:23 |
| bddebian | Get a real distro like Ubuntu or Debian ;-P | 22:23 |
| * bddebian hides | 22:23 | |
| andrei | bddebian, Heh, I want my box to work :) | 22:24 |
| mithro | andrei: so you are already using Ubuntu then? | 22:52 |
| *** jphr has joined #tp | 22:53 | |
| jphr | llnz: ping | 22:53 |
| jphr | JLP or Mithro around? | 22:54 |
| mithro | jphr: I'm afraid I'm about to go into a meeting | 22:55 |
| jphr | mithro: thats ok, i'll toss llnz an email to llnz, or should I mail it to the list? I dont really know mailing list etiquet | 22:56 |
| jphr | -1 "llnz" from that, hehe | 22:56 |
| mithro | list is better | 22:56 |
| jphr | just the general one? | 22:57 |
| andrei | mithro, Do you have a place for me to store the ebuilds? | 22:57 |
| llnz | jphr: pong! | 22:57 |
| jphr | llnz: oh hey! | 22:58 |
| llnz | hi | 22:58 |
| jphr | llnz: wanted to ask your opinion about chat in diplomacy, would you like to see it implemented in server/client? or would that be a longer term goal? As well I wanted to know what your feelings were about risk over diplomacy, I am far more familiar with it and it involves mucho less chat | 22:58 |
| llnz | there are message boards in the client, server and protocol already which could be used | 23:03 |
| llnz | risk vs diplomacy: i'm not concerned either way | 23:04 |
| jphr | ok, with that aside, what do you think the community would prefer Diplomacy, or Risk? I see diplomacy as something with say, longer turns, a little bit more out of reach of a few people. I see risk as a pretty good common ground, most people who have played diplomacy have played risk, but not the other way around (atleast i believe so) | 23:04 |
| jphr | (sorry i was mid paragraph :P) | 23:04 |
| mithro | jphr: requiring to modify the protocol is probably a good way to not get very far | 23:05 |
| Ohm | Personally, I'd say the risk playing style is a subset of the diplomacy playing style | 23:05 |
| jphr | mithro: good point | 23:05 |
| Ohm | in terms of which is possible to play after implementing | 23:05 |
| jphr | Ohm: true, I think risk is a more common experience though | 23:06 |
| jphr | I hadn't even thought of risk as a ruleset idea though, and i am far more comfortable thinking about developing it, as unlike diplomacy, I could probably recite the rules to it | 23:06 |
| jphr | llnz/mithro: Would it be appropriate to create a wiki page for a Risk ruleset as part of my application | 23:07 |
| llnz | jphr: yes, i would be helpful | 23:08 |
| llnz | s/i/it/ | 23:08 |
| jphr | ? | 23:08 |
| llnz | jphr: yes, it would be helpful | 23:08 |
| jphr | llnz: thanks that got lost in translation there somehow | 23:08 |
| *** Lukstr has quit IRC | 23:09 | |
| jphr | I best be off for the evening, see you all soon | 23:10 |
| *** jphr has quit IRC | 23:10 | |
| Ohm | i made meringues | 23:21 |
| Ohm | at.. 2am | 23:21 |
| *** Epyon has quit IRC | 23:31 | |
| *** JLafont_ has quit IRC | 23:31 | |
| greywhind | mithro: ping? | 23:40 |
| mithro | greywhind: pong | 23:49 |
| greywhind | mithro: do you know why setting the size of the window to the best size might not be working? | 23:49 |
| mithro | greywhind: I think you should look into the pop-up code as that resizes the window correctly | 23:50 |
| greywhind | mithro: ok | 23:50 |
| mithro | JLP: ping? | 23:53 |
| mithro | ~seen JLP | 23:53 |
| tpb | mithro: JLP was last seen in #tp 2 hours, 18 minutes, and 0 seconds ago: <JLP> Epyon: well it is still helpful | 23:53 |
| mithro | ~seen xdotx | 23:54 |
| tpb | mithro: xdotx was last seen in #tp 1 day, 18 hours, 34 minutes, and 59 seconds ago: <xdotx> heya :) | 23:54 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
