2.10 version fixes
Ugly version this time (=exactly what I mailed the beta testers).
Autoconf added for faster setup. If IP address (default start package) = "X" then it will be replaced by the local IP address. This should work well for non-LAN playing.
Debug host options now moved to "host options" (no more delay until activated).
Added "End current game" to host options (useful for the AI's playing against each others when the last human player dies).
"Use same options as last" now moved to host options as a toggle. Was too annoying to have as a question.
Host options - guarantee production added. Random levels now have a minimum production
Low = 1
Medium = 2
High = 3
Without this option
Low = 0-2 (in earlier version this was 0-1 which some people thought was too harsh)
Medium = 0-4 (0-3)
High = 0-8 (0-7)
Added
ResourceVictoryImportance=0.3
to Aggr.AI
Earlier it was 1.0 which meant the AI considered each turn to be worth 2%. In a long resource game the 50 turn cap became too important and thus had the AI throw away increase production cards.
Changed the usage of
ProductionImportance=1.0
so that productions of 0 and 1 temporarily boosts this parameter automatically.
2.09 version fixes
Certain random starting selections could generate autowins.
Unknown what other fixes I've done in the past years. Could be that autoregister doesn't work as well now as I seem to use hard coded ip values
again. Probably just an issue with Norton back in 2006.
2.00 version fixes
Host and Client Ports added to the startup window.
Aggr.AI has an added parameter (LowResourceImportance) for AI usage of resources. "1.0" might not be good enough and it should be more like "2.0" but it's too early to tell.
"Instant kill allowed?" option added. All (playable) cards that result in a player victory or loss during the first round will get discarded before the game starts. Note! Using a "two damage to tower" followed by a "pegasus lancer" is perfectly allowed, only single kill cards get discarded when this option is on. [known bug, if a lot of cards get discarded the "discard" texts don't get cleared properly]
1.99 version fixes
Total played time statistics were wrong for non-host human players (also enabled time statistics for AI's now).
AI's now realize they win when their tower equals exactly the victory tower height (oh sigh, what a bug :-).
AI's don't think killing a player twice is the best thing on earth (although it might not be a bad move but sometimes it is).
Card deck is now a few millimeters to the left and the border is not as thick (&/&%/% nitpicks!).
1.95 version fixes
First AI release. A few bugs remain but this plays well enough.
1.51 version fixes
See the readme in the package. Setup screen added.
1.50 version fixes
See the readme in the package. This is a mainly a TCP/IP release.
1.06 version fixes
Pretty much a bit of this and that release. Really minor things like random starting towers now required to be 10 points less than winning tower heights instead of the possible 5-6 points in previous versions.
Note. Next update will probably be a TCP update (throwing out UDP support to get better ADSL/ISDN performance). Scheduled for Summer 2003.
1.05 version fixes
Some funny things sometimes happened when starting a second game. Maybe it got introduced in 1.04b, maybe it was in 1.04 already. Uploaded a new version just in case. Also fixes one card which had a wrong cost in arco.crd.
1.04 version fixes
After a resouce win the next player still gets some resources (which can cause confusion). In some circumstances where several players had the same amount of winning resources the wrong winner was shown.
Dead/surrendered players were also required to acknowledge a double.
Added some statistics for the last game played.
AnimationSpeed added to ini because of Windows ME. Set to 0 for no card animation. 100 is the current default.
PacketTimeout added to ini (in 1/10 of a second, default 50, i.e. 5 seconds). If you have difficulties playing over a slow connection like an ISDN line you can increase the timeout and see if it helps.
Protocol version check added. If playing with different versions you will get a warning.
Card back now shown while card is sliding into place. Earlier it sometimes showed the played card which could be confusing as you could think you got another one.
1.03 version fixes
Average resource victory didn't work when resources where exactly the required amount.
Doubling and clicking a card before everyone had acknowledged "hung" the game.
Minimizing the windows now also hides the info window.
Registering players are now shown by name using the info window (host only).
Debug menu available for host only (if no play for 30 seconds - hung possibility).
Moneygame double and surrender both require confirming.
Messing around with screen resolution can force window positioning to fail (window shown off screen). More off screen checks added.
Random victory settings could produce a tower that was very close to the winning height. Random values fixed somewhat. Note! It is still possible to give stupid enough input values to produce bad games but they should not be frequent with decent boundaries given. I'm not making things impossible as someone might think a starting tower of 12 and a victory height of 13 is fun. With max random values a tower victory height of 70 will not produce starting towers higher than 56.
1.02 version fixes
Dropped player text shown for the wrong player. Client player is always shown at the bottom but this swap was missing from the dropped code.
Double discard fixed (actually double play was possible too).
Next two fixes are BETA. No known bugs but extensive four player testing not done as I'm away for the next 8 days.
Improved player handling for players quitting the game and rejoining. Now it is
possible to start with three players, one quits the game and goes for a smoke,
the two remaining players continue, later when the third player rejoins he will
automatically enter at the start of the next game. The host will get a "player
already exists" message - left in on purpose as it is informational.
Note! There is a slight delay when starting each game if one or more players
are missing. Currently the timeout is between 5 and 10 seconds.
Improved TCP emulation over UDP. Should work better over low bandwidth connections. Need someone to test these.