PROJECT: What a patch should contain. Where to draw a line.

[Pedrunn]

I believe the first patch should contain:
--> Fix and changes that ALL CTP2ers will aprove or dont care. Therefore adding new game start optons, bug in general and minor additions like new const.txt options, new slic functions, etc...
--> The questionable acceptance changes should be included as add-on

[Locutus]

Making the player colors optional would require a lot of work probably and only lead to a lot of work, so I doubt that will be done soon (though it'd be nice to have eventually). Other than that, moving some of the options in userprofile.txt to the game setup screen is something that certainly wouldn't hurt anyone and could be included in an early patch, providing it doesn't involve too much work.


I think my view roughly coincides with Keygen (I wrote this earlier today, it roughly comes down to what Keygen already said):

I think there people's ideas on what changes to make to the game with the source code can be divided in roughly 3 philosophies:

1) Keep the game (nearly) identical to how it is now, only fixing known bugs and maybe extending SLIC and mod functionality and similar stuff, but not (or hardly) touching gameplay at all (not dissimilar from the Apolyton Pack philosophy), leaving that to modmakers.

2) Making minor improvements to gameplay (i.e. rebalancing some things, improving the AI, streamlining some awkward SLIC workarounds in C code (e.g. unit updater), etc), but not completely rewriting the existing engine. Most if not all of this could be made optional.

3) Developing a full fledged sequel (or at least an expansion pack), with maybe new combat and economic models, maybe religion, strategic resources, new victory types and Mr Ogre knows what else... These features would require changes in all parts of the game and would be a pain in the @ss to make optional, and also not make much sense since their impact on gameplay is just too huge.

To cater to everyone's wishes, we could develop 3 versions of the game: a v1.2 patch, maybe eventually v1.3 to cater to philosophy #1. A v1.5 upgrade (and later maybe v1.6, v1.7, etc) that caters to philosophy #2 and a v2.0 version (CtP3, if you like) that caters to philosophy #3.

We can work on v1.2 and v1.5 concurrently, adding some improvements just to v1.5 and others to both v1.5 and v1.2. This is basically what's already happening (except that we're not really releasing stuff yet). v2.0 is currently in the early design/brainstorming phase and still a long way off.

[Locutus]

I forgot to note (but already implied) that we may develop more than 1 patch as we encounter more bugs or stuff to fix/add over time, where first patch that just fixes bugs (philosophy #1) would be v1.2, the next v1.21 (or v1.25 or v1.3), etc. Just as with the mods...

[Martin Gühmann]

Originally posted by Locutus
Making the player colors optional would require a lot of work probably and only lead to a lot of work, so I doubt that will be done soon (though it'd be nice to have eventually). Other than that, moving some of the options in userprofile.txt to the game setup screen is something that certainly wouldn't hurt anyone and could be included in an early patch, providing it doesn't involve too much work.

The player colors are currently associated to the players player 1 has his color also player 2 and so on, so we just need to make sure that the game uses the PlayerNumber from userprofile to start the game and to assign a player to the human. So far you can already download the interface option for this. Making sure that the game uses this index will also fix a bug from the patch 1.1 to do list.

But in fact to associate the color with the civilisation shouldn't be too difficuilt, but then we need to rewrite the colors database, with the dbgen tool it is not a problem. But then the whole copy and paste works comes.

-Martin

[Keygen]

BTW, in case there is any improvement or addition to the code from now on and there is a dispute or uncertainty whether it should be included in the first patch post it here for discussion.

[drulius]

I think it would be entirely possible to first release all three (2.1.2, 2.1.5, 2.2.0) patches at the same time, in the sense that some changes that are already done might go in all three and some others might only go in one, but whenever one of them is "done" the others could get released at their most recent playable state for playtesting. Of course the 2.2 mod might easily drift far away from the other two, but probably sometime after the first patch release when all the smaller fixes are already made.

Then the debate can center around what kind of changes belong in which patch, instead of which changes deserve priority to be in the first patch release.

[MrBaggins]

An issue with the current strategy we've been considering here is that the AI needs to be able to deal with anything we include in a patch...

thus if we...

bug fix -> improve the AI -> make significant game balance/improvement alterations

.. then the AI will probably be badly equiped to deal with those changes, and thus we'll need to repeat at least some of the effort, after those changes.

So, in my view, maybe, after we fix the bugs that make sense to fix, we should get together the ideas that are achievable to code, and then vote on those... then code... then do some structured playtesting.. probably with MP.

Thoughts?

[kaan]

Originally posted by MrBaggins


An issue with the current strategy we've been considering here is that the AI needs to be able to deal with anything we include in a patch...

thus if we...

bug fix -> improve the AI -> make significant game balance/improvement alterations

.. then the AI will probably be badly equiped to deal with those changes, and thus we'll need to repeat at least some of the effort, after those changes.

So, in my view, maybe, after we fix the bugs that make sense to fix, we should get together the ideas that are achievable to code, and then vote on those... then code... then do some structured playtesting.. probably with MP.

Thoughts?

I cant see the problem in workin in iterations.
We analyse the problems, design the solution, implement the design in the code, test and bugfix the whole thing and then make the AI cope with it.
Thats the way i see it anyway, also we are in the test and bugfix fase right now as we got the code delivered at that stage

[MrBaggins]

Well.. we will be working in interations... absolutely... but I'm thinking that dealing with the AI is theoretically the largest job, and thus the one we wish to repeat as little as possible...

Just my 2 pennies.

[Pedrunn]

Why dont we publish a patch with what we have right now?

[Tamerlin]

Originally posted by Locutus
I think there people's ideas on what changes to make to the game with the source code can be divided in roughly 3 philosophies:

1) Keep the game (nearly) identical to how it is now...

2) Making minor improvements to gameplay (i.e. rebalancing some things, improving the AI, streamlining some awkward SLIC workarounds in C code (e.g. unit updater), etc), but not completely rewriting the existing engine. Most if not all of this could be made optional.

Thanks Locutus, this is exactly what I am waiting for at this moment, a bug free CtP2 (philosophy #1), a better AI and why not some optional changes (philosophy #2)...

3) Developing a full fledged sequel (or at least an expansion pack), with maybe new combat and economic models, maybe religion, strategic resources, new victory types and Mr Ogre knows what else...

Now we are talking about another game, though this is not what I am waiting for from the source code I would nonetheless test such a game.

But I am convinced that some changes should still be optional (like strategic resources)... if it is possible.

[Tamerlin]

Originally posted by Pedrunn
Why dont we publish a patch with what we have right now?

IMHO there is no point in publishing a patch when the patch that would be released is not at least on par with the 1.1(1) patch released by Activision.

As far as I know the source code's version is that of the original game and though many bugs have been solved many of those that were corrected in Activision's patch are yet to be solved by the current programmers. Am I wrong?

[MrBaggins]

The work up until now is partially pre 1.10, partially 1.10 and partially post 1.10.

We've fixed some glaring bugs and omissions, done some SLIC 2.1 functions.

The main area that the patch dealt with, IIRC, was MP stuff, thats not been touched.

[Solver]

We still don't know the exact changes that the Activision 1.1 patch made about MP.

However, unless all SLIC functions from 1.11 are reimplemented in our patch, the game patched with it wouldn't support the popular mods. Other than that, our fixlist already looks great.

[Martin Gühmann]

We need also the Database access through slic otherwise you can't play mods like CTC.

But it wouldn't be wrong if someone with a clean version of the latest changes could compile a *.exe for testing. Unfortunatly I have some changes in my code that are experimental I don't want remove, and in addition my VC++ is not a professional version rather something like a demo with full functionallity, but with a disclaimer telling everyone that this program was made by MS VC++ Autorenedition and this *.exe is not indended for commercial use of course the disclaimer is in German.

-Martin