NGlayout for CEGUI

[kiolakin]

I am really interested in porting NGLayout to work in CEGUI. In order to stop from hijacking my own CEGUI Lua thread, I started a new one.

I broke into mozilla/gfx today and I can already see it is going to take a great deal of work before even considering mozilla/widget.

I was right about DeviceContext, but thankfully enough it is a fairly abstract nsDeviceContext. It appears that the gfx engine was already moved to Directx (although now they are using a complete rewrite of even that engine). so in some respects it will be easier than I thought, but on the other hand it showed me just how much I didn't understand about the inner workings of CEGUI.

I wonder if it is even worth trying to port the stable nglayout widget, or should I try to port their trunk cairo widget. The only problem I see with cairo is that it is supposed to be a cross platform opengl/postscript implementation all by itself. Although for the purposes of most it may be more portable to other OSes and platforms, it could make porting to CEGUI more difficult all together.

[kiolakin]

After careful consideration I am going to shelf this until Cairo has stabilized and is more widely supported.

1.Creating a mozilla based html widget with the outgoing engine will do nothing for games in early enough development stage to consider using it.

2.Porting mozilla/gfx is a super project in itself. I am not sure it is in the realm of a 1 man show, atleast not a man with my abilities.

3.I just found out that the few changes that I wanted to make to PLSM2 are more widescope than I originally anticipated. I am going to focus my time there, because I can already see that I am capable of helping a great deal. PLSM2 is currently much more crucial to my overall project than the convenience of an in-game browser.

Sorry to let you guys down so early in an undertaking. At first glance it looked easy because mozilla is already setup to be ported and the layout widget is already setup to be extracted for other use. Getting into the code I realized that all that portability came with the overhead of needing to create 1000's inherited implementations of mozilla classes. The overall bulk of the project greatly outweighs the level of difficulty, but it is a real world fact that enough brute force will conquer any ingenuity.

Hopefully as my game progresses and I get more of the content in (part of which is a UI entirely scripted in Lua, I will be able to add more to CEGUI. Although I will admit it is very tough to help develop a project when you can't reasonably fix yours on an unstable tree. Either way, hopefully Ogre will migrate to CEGUI 0.5 as soon as it becomes a release and atleast if I have a patch it will be closer to what the CEGUI team is looking at.