gmane.comp.games.tp.devel

Re: tpserver-cpp branches merged!

Ryan Neufeld — 2008-09-02T01:04:24

Great to hear (and a little bit scary might I say :P) As you mentioned in an email, could I get a branch called "risk++"; I'll probably start fiddling around on my own WIP branch until then. Ryan 'jphr' Neufeld --------------------------- visit me at hammerofcode.com On 1-Sep-08, at 12:53 AM, Tim Ansell wrote:

tpserver-cpp branches merged!

Tim Ansell — 2008-09-01T05:53:46

Hey guys, Lee has successfully merged all your branches (config, tae, and risk) into the tpserver-cpp trunk. This means that everyone now gets your stuff by default, yay! You should now continue your development on main branch (if you have any questions about switching, feel free to ask). I have removed the branches from the server. If you want a branch to continue doing some restructures (I know jphr wants to modify how some of the risk stuff) we are happy to provide one. Tim 'mithro' Ansell

Re: Starmapper and wormholes

Tim Ansell — 2008-09-01T00:14:14

A real simple solution would be to divide all your distances by a constant which is something like 1/1e6 of the universe size. I doubt you need more accuracy then that. Some ruleset might only use really small numbers too (IE around the 10s or 100s), so using "absolute" sizes is definitely not the best choice. 9634084866000 / 9634084 = 1000000 4773979192500 / 9634084 = 495530 Which are much more manageable. 3402000000 / 3402 = 1000000 2835000000 / 3402 = 833333 All these numbers definitely fit inside long ints. Tim

Re: Starmapper and wormholes

Vincent Verhoeven — 2008-08-30T13:18:29

2008/8/30 Krzysztof Sobolewski interia.pl>: You could probably use an approximate iterative method. I'm not entirely sure of a R^2 -> R iterative method though, someone good at numerical analysis here? Would Newton-Raphson with the gradient (instead of the derivative) work? Hmm. There's also the alpha-max + beta-min algorithm to quickly approximate the length of a vector. However, perhaps it's safer to just start using BigInteger or, you know, floating point numbers. BigInteger doesn't have a sqrt(), so you'll need to convert to float/double or use iterative methods anyway. Unless you just use the squared distance... So many options. :) Iwanowitch

Re: Starmapper and wormholes

Krzysztof Sobolewski — 2008-08-30T12:06:49

Re: Starmapper and wormholes

Vincent Verhoeven — 2008-08-29T21:18:39

2008/8/29 Krzysztof Sobolewski interia.pl>: Let's see... Correct me if I'm wrong. Some outside planet locations: -2399968326000, 1119431722500, 0 7234116540000, 889698663000, 0 6608869260000, -3654547470000, 0 Thus 9 634 084 866 000 x 4 773 979 192 500. For RFTS, I have 210000000, 2992500000, 0 3612000000, 157500000, 0 which yields 3 402 000 000 x 2 835 000 000, or essentially a factor 3000x2000 smaller. Hmm. :) Iwanowitch

Re: Starmapper and wormholes

Krzysztof Sobolewski — 2008-08-29T20:35:05

Re: Starmapper and wormholes

Krzysztof Sobolewski — 2008-08-29T20:36:45

Re: Starmapper and wormholes

Vincent Verhoeven — 2008-08-29T20:18:46

2008/8/29 Krzysztof Sobolewski interia.pl>: Yep. Test with 2 bots shows the 2 names anyway. (slight note: the text gets cropped off, but that's nothing really bad) Yeah... I think so. Each star has exactly one planet, so I'm not sure whether it's planets or stars I see. There are little diamonds for all the stars, which seem to grow with the turns - I presume the size represents the resources, thus the amount of troops on the planet. However, no regions at all, only white-on-black diamonds and star names. I want my pretty colors. :( Screenshot at http://img379.imageshack.us/img379/1654/tparseclocalhost0000sr3.png Iwanowitch

Re: Starmapper and wormholes

Krzysztof Sobolewski — 2008-08-29T19:57:51

Re: Starmapper and wormholes

Vincent Verhoeven — 2008-08-29T13:45:19

2008/8/21 Krzysztof Sobolewski interia.pl>: So let's go one layer deeper and just use star systems instead. :) Since these are displayed in tpclient-pywx (in contrast with galaxies), it can perhaps be assumed that developers will place them at the right positions, and thus that you can use them to determine map bounds. I think this is the same thing tpclient-pywx uses. I attached a patch for this, however you should review it and probably design something that is a bit better, because this replaces one ad-hoc method with another. Also I only tested this with the Risk ruleset, it might break other rulesets. The current status is that it now manages to draw the maps. Only problem is that those are useless because they lack pretty colors. :) It's probably because of the peculiarities of the ruleset: it doesn't use ships and I think planet ownership is different or something. Also, strength is measured by resources because of the absence of ships. Would it be possible to create a configuration that dr

Re: AI battle moved

Victor Ivri — 2008-08-27T17:52:44

2008/8/27 Krzysztof Sobolewski interia.pl>: That's a great idea. I'll be sure to make tpclient-pywx screenshots each turn :).

Re: AI battle moved

Krzysztof Sobolewski — 2008-08-27T16:39:30

Re: AI battle moved

Tim Ansell — 2008-08-26T00:48:46

Hi guys, I won't be able to make this time. Good luck with the battle. Tim On Mon, 2008-08-25 at 18:52 +0200, Vincent Verhoeven wrote:

AI battle moved

Vincent Verhoeven — 2008-08-25T16:52:59

Hello all, vi1985 and me moved the battle between our AIs to 49 hours later, http://www.timeanddate.com/worldclock/fixedtime.html?month=8&day=28&year=2008&hour=14&min=0&sec=0&p1=0 http://tinyurl.com/58tltz because I had some family stuff interfering. See you then! Iwanowitch

Re: Scratchpad patch to setup daneel-ai

Tim Ansell — 2008-08-24T23:48:36

Applied - d003bdb74686a40e198c65f41d32722e8860bebc Tim On Mon, 2008-08-25 at 01:37 +0200, Vincent Verhoeven wrote:

Scratchpad patch to setup daneel-ai

Vincent Verhoeven — 2008-08-24T23:37:07

Mithro: please apply attached patch to the scratchpad repository. For those that cannot wait for mithro to do this (cough nash cough), feel free to patch locally. Git should be able to handle that. Iwanowitch

Tigris and Euphrates

Dustin White — 2008-08-22T20:11:21

Re: AI battle

Tim Ansell — 2008-08-22T06:01:43

It must be Noon somewhere in the world! :) Tim

Re: AI battle

Victor Ivri — 2008-08-22T03:49:10

never mind that :). It seems to be a mishap on my end: I set the whole thing up, such that each asynch "time" frame, which is > 0, acts as a catalyst to start a new turn routine. I didn't know this, but the server seems to also send one to indicate the turn will be "n" seconds longer, after it receives "finished turn" frames from enough clients. This makes my bot act like there are two turns, when there is only one :).

Re: AI battle

Victor Ivri — 2008-08-22T03:16:11

Below is an output of three turns in a game between two bots. The bots are "puk_1" and "puk_2", and it's puk_2's output-stream. It attempts to conquer the last planet of puk_1, which has 7 troops on it to begin with, and gets a zero reinforcements each turn. Despite the massive attack on it, the planet only succumbs on the third turn. This seems pretty counter-intuitive to me. I believe the output reflects what's being done, and there's nothing wrong with it, but if you say that the following is impossible, I'll have to take a closer look :) . Here it is: ______________________ puk_2's output: Turn number: <64> (since launch of client). Start of turn routine commencing... 4 seconds to end of turn. Initializing bot with new data..... - - - Universe data: Neutral stars: Stars owned by player puk_1 (1): <66> Total stars: 1; Total army strength: 7; Total reinforcements: 0 Stars owned by player puk_2 (2): <26> <6> <45> <73> <83> <64> <8> <89> <60> <10> <79> <55> <36> <57> <42> <77> <32> <71> <91> <40> <62> <2