About 1: It’s an absolute amazing idea.
My considerations regards item 2 follow bellow.
I don’t know why people don’t use cocos2d-x for Marmalade. Some guesses:
* maybe because they dislike cocos2d.
* maybe because they are not aware of it.
* maybe because it has it’s own graphical functions and they prefer to stay with them instead of relaying on a third party maintainance. I ’m almost sure it’s it because I would thing: Marmalade always will support there own graphical functions, but what about this cocos2d-x port? They are not Marmalade. They may decide to give up this port and I will have a lot of trouble to port from cocos2d-x into Marmalade natives APIs.
I feel that if cocos2d-x had the possibility to be multi-platform as Marmalade is, it would make life easier for both cocos2d-x developers (you guys) and cocos2d-x users (like me).
Let me put some points so we can try to figure out what would be less work / more productive:
- at the moment, there are many ports of cocos2d-x (5):
For each new platform you guys have a lot of work to keep it, right? Now imagine cocos2d-x running on Bada and Symbian. It would mean 7 different ports to maintain.
Let me examine Cocos2d-x on QT:
QT currently running on
BogDan Vatra, the guy which ported QT for Android, is working in a port to Bada (you can see it here). It means a new supported platform and more opportunities.
Today you guys expend many time adapting the C++ code under cocos2d-x to deal with idiosyncracies of each platform. With QT, on the other hand, it won’t be your business. There are many great guys developing QT in order to make it as transparent as possible for QT users, no matter what platform it will run on.
You won’t need to reinvent the wheel because QT framework is so powerful and has too many auxiliar classes for almost anything we can need. You will use you precious time with what really matter: improving cocos2d-x.
Moreover, for each new platform QT were ported, cocos2d-x will be ported with almost no extra effort (if any) of you guys.
I know decision is up to you guys, but if it were up to me, I would do the following:
- create two groups: a group to port cocos2d-x to QT and the other one to keep all platforms
- port cocos2d-x for QT and after that I would discontinue other ports.
It seems to me that in the long term using QT as the framework for cocos2d-x development will have much more advantages than supporting multiple platforms.
Oh, and I’m pretty sure that there wouldn’t be any problem for users to use cocos2d-x for QT because:
* in this case you’re not the third party guys (as it happens on marmalade port)
* conversion from any actual cocos2d-x project into cocos2d-x for QT would be almost a matter of import resources into the QT Creator IDE and compile it. A conversion tool could do the job.