Crazy Eddie's GUI

[daniel.santos]

I'm investigating this toolkit as a possible replacement for current Glest GUI wigidts and as a mechanism to make GAE completely skinnable, so that mods can mod the game more completely.  Thus, GAE could even be launch with a certain parameter and a "game framework" mod could take control, use the Glest engine for everything it does, but appear completely different than it does currently.  Some aspects of this may be a lot of work, like the way that the intro screen is really cool and stuff.  This is a proposal for implementation way down the line btw.

http://www.cegui.org.uk

[omi]

I thnik that this is a very bad idea.
We will need to change everything. From the processing of events, change the game state class, write some a new CE GUI
controls ( like a icon control from world screen which increases size when you click on it ). Rewriting all game screen
which uses GUI, next... design change, writing code and bug fixing.
And many other things ( problems ) about which I do not know...

Other reasons:
- CE GUI is too big and too complex.
- CE GUI is like a blackbox for us.
- We will need also write a renderer system for CE GUI which will use GEA rendering functions.

- GEA has own GUI system which works perfect for GEA needs.
- the GEA GUI system is very simple.

I think that this is a task for the two developers for 2-3 weeks... maybe 4 weeks.

if you want to have a skins... just change a GUI render system in the GEA engine.
GUI Render system controls how draws the GUI control. If you changne this you will
have skins.

[titi]

I also think, that this is really nothing we really need.

[ArKabZol]

Well, it's certainly a good idea if you really want to lay down a bunch of extra time to make GAE even bigger and also make even more different than normal Glest. I see no problems with it myself, I like things to be very customizable.

On the other hand, can't you change the GUI already by replacing all buttons?

[wciow]

Personally I would put GUI way down the list of changes. There are many more important things that need changing first.

Of course its totally up to the coders to decide since they are in a far more knowledgeable position regarding what can be done. IMHO the current GUI seems to serve rather well ATM and I agree with Omi that spending up to a month coding just so you can reskin the GUI is not top priority.

[hailstone]

I think it is not so important at the moment but will need to be implemented before other features are possible, such as using a scrollable list box in a lobby. Glest is missing a lot of GUI features, not just for skinning. It only has label, simple window dialog, progress bar, message box, button and list control. Also some game specific GUI like map, selection box and hard coded mouse pointer. It will free up some of the render file which is currently a bit less than 3000 lines long. I think it is worth adding in a GUI library once the stable release is complete.

Here are some other GUI libraries for comparison:
http://guichan.sourceforge.net/wiki/index.php/Main_Page
http://www.bramstein.com/projects/gui/ (more down the bottom)
http://sjgui.sourceforge.net/

[omi]

http://www.bramstein.com/projects/gui/ (more down the bottom)

I looked at this gui system and I think it is nice.

[titi]

Oh No! Pleeeease, don't put your affords in a new GUI system
If you want more eye-candy, improve the 3d-engine
For example:
- nice watereffects( dynamic water reflections ...)
- burning units ( movable ones / you can define them but ...... )
- black/dark shots ( black particels )
- working transparency in moving objects ( does not really work yet )
- animated trees? ( or does it need too much hardware??  )
...

[daniel.santos]

I think it is not so important at the moment but will need to be implemented before other features are possible...

This is exactly where I'm coming from.  The Glest/GAE engine will have to better accommodate modders who want to use the engine for an entirely un-Glest-like game.  The idea of bringing in a GUI library is that you don't replicate the work that others have already done and we can focus on the features that we want instead of re-inventing wheels.  Aside from being able to apply new "skins" as in textures, there are also a lot of look & feel features that could be supported as well, borders, alignment, padding, layout, etc.

But don't worry guys, I'm not thinking of this as something in the very near future, there are lots of other things that are needed right now.

- black/dark shots ( black particels )

Like this one .

- animated trees? ( or does it need too much hardware??  )

Yea, I'm working on this right now actually

[wciow]

 Does Blender and/or the g3d exporter let you decide how many meshes (groups of polygons/triangles) there are in a model and their order?

Yep this can be done, creating separate meshes in blender allows you to export them as one model. The amount of meshes are afaik unlimited in number.  I'll have to investigate how exactly the export script orders them. I think its by the order in which they are selected before exporting but I'm not sure.

[daniel.santos]

Yep this can be done, creating separate meshes in blender allows you to export them as one model. The amount of meshes are afaik unlimited in number.  I'll have to investigate how exactly the export script orders them. I think its by the order in which they are selected before exporting but I'm not sure.

Cool!  Well maybe you can chop of the Lich and put the cape in the last mesh in Blender, re-export it and see if it fixes that transparency problem.  I'm not certain that it will, but I noticed that it *might*

[gameboy]

have you'll try removing doubles?

[daniel.santos]

have you'll try removing doubles?

What's "doubles"?  Double-sided polygons (i.e., drawn on both sides)?

[titi]

no no, what he meant is the strange behaviour of the blender exporter for glest xmls.
you can switch transparency or teamcolor with the doublesided feature in blender.
This is only a hack, because in glest all faces are displayed doublesided ( visible from both sides ).