| 07-08-2005, 11:37 PM | #1 |
14/07/05 Next progress update is early as I will not be here this weekend to do anything. So what we have done this week
So after my 4 day holiday, i am going to find a way to stop humans tower rushing with every blimmin peasant. I will see about adding thenonhackers tower rush idea for later game with humans. I am going to turn the millitia expansion code into a job which all the other special abilities now are. I'll also have alook at the zepplin code to see why its not working correctly all the time. I am also going to check if the running out of rescources bug on reaching the tavern is fixed. It might be in my last update already. 08/07/05 The 2.45 Beta1 seems to be extremly welcomed and we have already received many replays and bug reports. But keep providing more. You testing is most greatly appreciated. So all testers keep up the great work and a great big thank you from the AMAI team. Now some updates of progress this week
So it was quite a nice accomplishment. No more friends map fountain exploit at last. Unforunatly you guys won't see till a new version is released. This next week I will possibly add the mana fountain usage and start re-looking at the different strategic jobs e.g. Tower Rushing, Millitia expanding, Ancient Expanding. They need a nice touch up as no longer seem to be working great anymore. Expect another progress update in a weeks time. |
| 07-09-2005, 06:07 AM | #2 |
Glad to hear it. I wasn't able to complete two best out of 10 tournaments, AMAI vs. Blizzard AI, because 1) Blizzard AI would run out of money, unable to expand, then sit idle in base and 2) AMAI wouldn't finish 'em off because it was stuck in a loop trying to climb over trees or jump across streams to unreachable mines. Hopefully next release, AMAI will make a clean sweep ^_^ Team1 = AMAI Team2 = Blizzard |
| 07-11-2005, 06:49 AM | #3 |
Great work there! Optimizations, elegant algorithms, I know those are hard work and entailed intensive analysis. Especially the Friends map bug, it's fixed at last! ![]() BTW, I'm studying Game Programming in Flash, which makes me ask: Did you implement in Perl the A* pathing algorithm? It's a popular topic in Flash Game Programming books, which I think can be applied to AMAI. |
| 07-11-2005, 08:16 AM | #4 |
Aiandy's pathing is a sort of modified A*. The System is based on A* with some few differences. This caused problems though as on large maps, grids used for algorithm became extremly large causing map to be recorded inaccuratly (reason why is because theres only a certain number of spaces available in a Jass array). Its still perfect for certain features e.g. front tower computation and so this won't be removed, but it did not work correctly in detecting that you could path to a point as we all know. I solved this by basically cheating. I used blizzards inbuilt pathing system. Sending hidden 0 food units at movement speeds of 522 to shops, works perferctly. If unit reaches target you know mine/shop is pathable with the minimal of effort. well thats the gist of it. It slighty more complicated then just sending units but it is compared to A* very simple. |
| 07-11-2005, 11:19 AM | #5 |
@Strategy Master: Hey, I'd say that's a very smart alternative, compare CPU needed to move 0-food unit to intensive pathing algo. :) |
| 07-13-2005, 06:11 PM | #6 |
@thenonhacker CPU to move 0 food units is much less than using pathing algorithm due to pathing algorithm makes more overall loop cycles. 0 food cycle only does about 3 :) @All Anyway i need some info about tower rushes and ancient expansions. They at the moment can only be chosen within the first 5 mins. I just wondering if whether people feel they should be able to do any of these techniques throughout the whole game instead of being limited to the start. |
| 07-14-2005, 09:33 AM | #7 |
IMHO: – Tower rushes in the very beginning only. – Ancient expansions may be possible through the hole game. But at some maps, like the Market Square with 8 players, there shall be troubles. |
| 07-14-2005, 03:46 PM | #8 |
I have this idea mid-game where AMAI Human buys 3 Ivory Towers, attacks your base with 100-food units, plants the 3 Ivory Towers, upgrades them to Guard Towers or Cannon Towers. That would be cool for the next next version of AMAI. |
| 07-14-2005, 05:04 PM | #9 |
@Tez Done! @thenonhacker Its a nice idea. Could be a sort of heavy tower attack. Usage of cannon towers there might be more useful as great to damage buildings. EDIT @All I have updated the progress list for this week (see first post). A few days earlier than its supposed to be but i am away for 4 days. |
| 07-15-2005, 05:23 AM | #10 |
Its a nice idea. Could be a sort of heavy tower attack. Usage of cannon towers there might be more useful as great to damage buildings. 100 food army will destroy base faster than these canon towers build up. ^_^ I have updated the progress list for this week (see first post). A few days earlier than its supposed to be but i am away for 4 days. If you will be absent, can we have the second beta? Please! |
| 07-18-2005, 05:12 AM | #11 | ||
Quote:
Makes sense. AMAI should build Guard Towers instead. Quote:
Yeah! I agree! 2nd beta please! There are many fixes worth testing now. ![]() |
| 07-19-2005, 06:37 AM | #12 |
100 food army will destroy base faster than these canon towers build up. Makes sense. AMAI should build Guard Towers instead. Really? IMHO after the second minute of game any beaten AMAI can slaughter enemy workers at his base. Ñertainly, if AMAI count such nonsense as an attack. |
| 07-19-2005, 11:58 AM | #13 | ||
Quote:
I want to check a few other issues first then its possible. |
