Re: [glob2-devel] The future!

[Leo Wandersleb]

Bradley Arsenault wrote:
> Giszmo said that we should release 1.0
> imeddiettly following the completion of the new map editor.

Can't remember having said that. In my opinion glob2 is highly playable and 
thus should not be called alpha. If we release 1.0 we must change our release 
policy to something with release-candidates, feature-freezes, ... It must not 
happen that we have desyncs that can not be restored via auto-save. desyncs 
that force the players into 1 minute pauses every 15 minutes are no blocker for 
a 1.0 I'd say.

Having said that, I vote to have all feature changes post 1.0, take out the 
market that doesn't work, take out nicowar that causes desyncs (see 
bug-tracker) and release 1.0 after testing for at least 50 games including lan, 
windows, linux, macintosh, yog, campaigns, editor for any crashes that we don't 
want the player to be confronted with (system freezes, crashes while editing 
(work of an hour lost?), crashes in game that can't be restored (it's so 
frustrationg to not be able to see how the game ends), saved campaigns that 
can't be restored, ...

Let's modify this desync-"Waiting for ... -> You have won" to something like 
"Connection lost due to network problems. Autosavegame has been written. To 
continue please host the Autosavegame.".

For the feature changes that I would postpone:

> 1) We need a new, advanced campaign system.

I don't. In my eyes it would be perfectly enough to have training missions with 
no locks at all.
All that is more than that puts too much focus on single player mode for my 
taste.

> 2) Custom games need to be improved. Alliances need to be editable
> before the game starts.

Yes! Including a switch that the alliances can not be changed in game. I like 
it the way it is done in TA for example. The first to choose a team symbol gets 
it displayed broken. As soon as 2 players are on the same team the symbol is 
not broken any more.

> *6) No offense to Giszmo, but we should have another map generation
> system, perhaps using a different approach. I love Giszmos maps mainly
> because they are realistic and customizable. On the other hand, they
> are generally unballanced, making them better for scenarios when a
> realistic, unballanced terrain is desirable. The different approach
> I'm suggesting is using a somewhat constraint based system. For
> example - having the map generator select a position for an island of
> wheat based on the distance to each of the four players islands
> multiplied by a small random modifier between 0.8 and 1.2. This is
> certainly not a required feature by any means, just an idea.

I have plans to improve the balance of my maps by allowing symmetries. so 
before you generate a map u decide to have 1, 2, 4, 8, 16 or 32 equal regions. 
then the map gets generated for only 1/n-th of the whole map and this patch 
gets repeated. while on plane maps this would still result in unfair maps, on 
torodial maps you get 100% balanced maps what would be desirable for 
tournaments.
this change should be super easy to implement but unfortunately i lack of 
motivation :/

Just my 2 cents

Leo Wandersleb