[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Adonthell-devel] weather fx

From: James Nash
Subject: Re: [Adonthell-devel] weather fx
Date: Thu, 23 Sep 2004 21:14:31 +0100

Ok, as you know, I'm no expert on this stuff, but if we choose to use someone else's front-end library it should support the features we were going to implement, such as:

- the 16-bit style isometric 2D graphics we use :D
- having 3D properties internally to allow for different height levels on maps, jumping around etc.. as we discussed back at the 2002 summer meeting - cross platform support (wanna play Adonthell on my PowerBook as well as my Linux box :P )

From an artists point of view the more fancy graphics effects (realtime transparency, hue/saturation effects etc - which could be used for weather fx for example) the better. Since 3D gfx cards are pretty standard these days some hardware acceleration would be cool (dunno if that's possible)

Just my 2¢, hope it helps!


On Sep 23, 2004, at 08:22, address@hidden wrote:

Joel Vennin wrote:

I think we should use a more abstract library (for graphics and network) which is already done in order to save time. Because as you can see it, we have not lot's of time. Alexandre seems to be die under paper and under its
thesis (it's like me).

Mh. I was thinking about that too, but a little afraid to raise the topic. There are a few things to consider, though:

[1] do we want to keep the backend code? I personally am in favour of that, because it ensures a long lifetime of the engine. Who knows what gfx library will be in use in 10 years time? With the current approach, we just have to replace a thin layer to make use of it. And that without locking out people who want to stay with SDL 1.x.

That means, whatever library we'd chose, we'd need to put it on top of our backend architecture. So whatever lib we pick, it should be as backend-independent as possible to begin with.

[2] Technical concerns aside, we need to specify what we want to be able to do. No use in picking a library that needs to be rewritten to a large extend, because that would probably be as costly as writing our own.

So yeah, we have to compile a list of what we want, in order to make a good choice.

[3] Regarding 'yet another dependency': I imagine that if we have to make some changes to any code we chose, we can as well fork it and copy into our source tree. So there wouldn't be any more dependencies to be concerned about. But, as you said, Joel, we had to do it nicely, with approval of the owner. (As it would be GPL'd, there shouldn't be any problem with this anyway)

To conclude: I wouldn't see much problem with picking an existing GUI lib right now (although even that means a little bit of porting work).

Picking a map engine seems a little harder, but if we could find a suitable one, I'd be happy to try it out. We do need some advancement.

Any comments, suggestions?

Le Mercredi 22 Septembre 2004 08:24, address@hidden a écrit :
Driving through some nice landscape last weekend with cool weather fx gave me the following idea: Each map should have a 'weather schedule' attached. I.e. a Python script that controls special fx like clouds, fog, rain, snow,
sunrise, sunset, etc ...

These scripts need to be per-map for the following reasons:
- each map is different, so global weather fx won't work. There would be no
snow in the desert, only sand storms. Fog would be in forests or near
water. - Tuning the weather to fit to each map will vastly improve the
atmosphere of the game. - for performance reasons, the python scripts
execution could be simply disabled, thus stopping costly graphic
calculations (the scripts itself would be event based, so little overhead

One particular idea for an effect I had was to have water reflect the stars at night. Should be possible with an additional, transparent layer on top
of water tiles.

Really good idea, it will be more realistic.
I've some suggestion about the adonthell core. In fact, we have (or kay has :p) a good background for adonthell. All is scriptable, it's possible to create scenario, (we have all ready some good scenario, story, etc) . But on the foreground, what we have ? We have some primitive functions that allows to change between graphics library, to draw some pixels on the screen with
nice color. My suggestion or my think is :
I think we should use a more abstract library (for graphics and network) which is already done in order to save time. Because as you can see it, we have not lot's of time. Alexandre seems to be die under paper and under its
thesis (it's like me).
I would continue to contribute to adonthell, but if we can save good time using more higher library it will more acceptable for myself. We can find a higher library and why not hack a little this library with is owner in order
to support python script.

I'm waiting yours responses. I don't want a response that tell me: "Oh no, i
don't want another libraries depends".


Well, so much for that idea.


Adonthell-devel mailing list

Practically perfect people never permit sentiment to muddle their thinking.
                -- Mary Poppins
Adonthell-devel mailing list
Adonthell-devel mailing list

Personal site:

reply via email to

[Prev in Thread] Current Thread [Next in Thread]