Idea to get (more) games to the Pyra by supporting Cocos2d-x

[Drammurt]

[SIZE=11pt]Hi![/SIZE]

I don't know anything about Pandora's/Pyra's software ecosystem and supported frameworks and hardware design, so pardon me if this has already been discussed or is simply impossible.  To bring more people to the Pyra, I think it would be great if the Cocos2d-x framework would be supported by the Pyra.

What is the Cocos2d-x framework?

It is a free multi-platform open source game development framework written in C++. It supports iOS, Android, Linux, OS X, Windows RT, Windows 32-/64-bit, Windows Phone, Blackberry, Tizen and others. Developers only need to write their code once, in one language (C++) and can then compile the code on/for the different platforms without any extra effort. [SIZE=11pt]Cocos2d-x is maintained by Chukong (a huge Chinese mobile game developer) in cooperation with engineers from Google, Microsoft, ARM, Intel and Blackberry.[/SIZE]

The framework is used by a lot of individuals and small companies, but also large companies are using it, like Zynga, Wooga, Gamevil, Glu, GREE, Konami, Handygames and  many others. [SIZE=11pt]With the framework it's relatively easy to create 2D games, but also applications can be created if you don't need the OS-gui-elements. Also, it's fast and robust[/SIZE][SIZE=11pt].[/SIZE]

How would the Pyra benefit from Cocos2d-x support?

It would bring more native games to the Pyra. Developers could simply compile their game(s) for the Pyra and make it available to everyone through a repository without any extra effort. I for one would make the game that I'm currently developing available to the Pyra, why not? Having support for Cocos2d-x would not only attract more developers, it would also give some free exposure for the device in the Cocos2d-x developers community. Also, developers that are currently developing for Pandora/Pyra, have the chance to use the Cocos2d-x framework which is a great open source framework for game and app development on mobile and desktop environments.

How would developers benefit from Cocos2d-x support on the Pyra?

I don't know which frameworks (if any) are currently being used by developers, but if the Pyra would support Cocos2d-x it would give Pandora-developers the chance to also release their games in the AppStore and Google Play and even make (some) money from a game that otherwise would only be available on the Pyra. For any developer the Pyra is a great mobile testing/development device with touch support, keyboard, mouse(emulation), gaming controls, a high dpi 1080p screen, wi-fi, bluetooth, 3g, GPS, HDMI out and huge (though invisible) sign saying "GODLIKE HANDHELD" on the lid. Developers could use (misuse?) a release of their game on Pyra as a free beta-test before releasing to the huge iOS- and Android-market. Feedback from the Pyra-community is more useful for developers than the ratings and comments on the various appstores, especially in beta-stage.

How can support for the Pyra be added to Cocos2d-x?

A fork of the current source can be made and adapted for the Pyra. Afterwards those changes could be pushed to the main codebase so that there's native support for the Pyra. I think it won't be a lot of work to get a basic working version. Cocos2d-x already supports both Linux and Android on ARM so only Pyra-specific implementations would require changes. Sure, there will be work to get everything bugfree and working, but it's not like porting code of a  framework from one platform to another. There's a Cocos2d-x test-set that will cover a large part of the framework. Compiling this testapplication will give a good estimate on the initial compatibility.

I think there is a win-win-win-win-win situation: more Pyra sales - more games - more developers - more users. And I would be more than happy to help with adding support for Cocos2d-x to run on the Pyra.

So, what do you think? Is this viable?

Drammurt

 

[Wally]

I think we should wait until we actually have a proper running Pyra board

See you in Two Months™

 

[FBnil]

Its seems to have no DRM. So it fits quite well, and maybe just recompiling works, so probably little work to port it. Monetizing could be seen as a problem, what we have is the repository, and that is on a try before you buy basis. So maybe that puts some developers off?

Here is some experience from the rambling programmer, I think its worth the effort.

http://ramblingprogrammer.com/?cat=14

 

[Caine]

Developers could simply compile their game(s) for the Pyra and make it available to everyone through a repository without any extra effort.

And this is how all those crappy ports are born. Also sounds more like marketing than reality to me.
That said, Pyra is an ARM GNU/Linux platform, since you state that Cocos2d-x is an open source platform with support for Linux as well as ARM cpus it sounds viable. At the very least it might decrease the threshold for bringing games to the Pyra.

Keep us informed about when you have added Pyra support to Cocos2d-x.

 

[_jr_]

AFAICT it needs GL, otherwise looks like it could work out of the box. The value of the available games is difficult to evaluate for me, since most of them seem to be in some asian language.

 

[Drammurt]

Its seems to have no DRM. So it fits quite well, and maybe just recompiling works, so probably little work to port it. Monetizing could be seen as a problem, what we have is the repository, and that is on a try before you buy basis. So maybe that puts some developers off?

There is indeed no DRM. Cocos2d-x has no binairies for download, just the source code on Git under the MIT-license. As far as monetizing goes: I don't think many developers would see the Pyra-platform as a target market to make money from. It's a niche market and developers wouldn't gain much from it anyway. A "try before you buy"-possibility might even attract some developers. Every developer that releases his games for Pyra is a plus.

Developers could simply compile their game(s) for the Pyra and make it available to everyone through a repository without any extra effort.

And this is how all those crappy ports are born. Also sounds more like marketing than reality to me.

I think the opposite is true about the birth of crappy ports: most of the time the actual porting of code and resources from platform X to Y is the cause of different implementations and new bugs. Having only one source for a game prevents those porting issues. Cocos2d-x has been available for more than three years and is based on the very popular Cocos2d-platform for iOS that has been around for 6 years already. It's pretty real. I develop and debug in Microsoft Visual Studio for Desktop. When I start Eclipse I can compile the code and run the exact same game on my android device. Then I push my code to Git, switch to my Mac Mini, pull the code from Git with Sourcetree and compile the code in XCode and run the application on my iPhone or iPad. Then I can run Visual Studio for Windows 8 and compile the code to run it on my Windows RT tablet or Windows 8 tablet.

AFAICT it needs GL, otherwise looks like it could work out of the box. The value of the available games is difficult to evaluate for me, since most of them seem to be in some asian language.

Basically, it needs OpenGL ES 2.0. On Windows RT initially they used DirectX 11 because RT does not support OpenGL ES. Last year however Microsoft Engineers rewrote this code and made use of the ANGLE-project that basically translates OpenGL ES 2.0-calls to DirectX 9 or 11 calls with hardware acceleration.

There are many games created by Cocos2d-x. To give you an impression: these are just a few of the popular games: Zenonia 5 (trailer, Google Play), Contra Evolution (trailer, Google Play), Clouds & Sheeps (trailer, Google Play), Hill Climb Racing (trailer, Google Play), Dragon City (trailer, Google Play) and Banana Kong (trailer, Google Play). And here are some statistics of the games created with Cocos2d-X and published on AppBrain (an alternative Android application store).

Marketshare by category


Casual - 7.54%

Arcade & Action - 6.08%

Cards & Casino - 5.17%

Racing - 4.93%

Sports Games - 4.69%

Brain & Puzzle - 3.65%

 


Even by attracting only a small group of developers, the Pyra would get a nice boost in sales, popularity and software.

 

[FBnil]

... wow...

 

[ptitSeb]

Ok. Nice showcase. Now if I wanted to start working on porting the engine, what would be a good opensource game that I can port (porting just an engine is no fun)

 

[sebt3]

(porting just an engine is no fun)

It's not about fun but about being able to : how do you test the engine without games ?

 

[Granitehead]

(porting just an engine is no fun)

It's not about fun but about being able to : how do you test the engine without games ?

With the test app mentioned in the first post.

 

[ptitSeb]

(porting just an engine is no fun)

It's not about fun but about being able to : how do you test the engine without games ?

With the test app mentioned in the first post.

Yes. Where is the fun ;-)