Mobile platform support? (was: Can I haz iPad?)

[james876]

 I think if so megaglest run in ipad on, it would be very pleasant thing, I would like to know if you know the idea, thank you!

[tomreyn]

Not for the foreseeable future. This device is based on an ARM arcitecture, but MegaGlest only works on x86 style architectures. Also, the operating system of this device is a very closed environment. MegaGlest also doesn't run on toasters.

[james876]

ok, it does feel weird, I wonder whether you can run it in the android system, should be a lot of Tablet PC use the android system. thank you!

[will]

It should be possible to port glest to both android and ios and even chrome nacl.

The challenges are moving to opengl es2 and the networking.  Of these, networking is perhaps most code churn, and there is a risk that despite streflop that games between arm and x86 don't work.  But there is a good chance they do, and certainly they ought to be able to play other players on the same arch.

A touch-optimised ui (no rigght-click
etc) would be needed.

The big barrier is not code - its time and money.  Developing for ios is not free, and all tablets need access to hardware

[titi]

ios will never come as long as apple only distributes software via its appstore! This is a no go for all GPL based things, because you cannot distribute the source with it.

[ultifd]

Besides that, it's kinda hard to play a RTS with a touchscreen...

[will]

Just to note that a version of Chrome NaCL is now available (in beta but working) that does full-screen opengles 2 and local storage.

It would be possible to port Glest to be distributed by and run in the chrome browser.

[Dritominous]

Wow, now that would be cool!

Heck, if we could play MG in a browser, just think of how much easier development would be!!

[Omega]

Wow, now that would be cool!

Heck, if we could play MG in a browser, just think of how much easier development would be!!

While it could be cool, the development would likely be harder, as developers would need more experience, without decreasing the previously required experience. And I'd imagine the performance trade off would suck, chrome is fast, but it's really heavy on my RAM.

[will]

While it could be cool, the development would likely be harder, as developers would need more experience, without decreasing the previously required experience. And I'd imagine the performance trade off would suck, chrome is fast, but it's really heavy on my RAM.

This is not true.  This is not how computer programming works.  A programmer would not think this way.

Its also not how chrome nacl works performance-wise.

I am not volunteering to do the port, however.

[titi]

Up to now all mobile devices are too slow to handle megaglest, so all thoughts about this are useless. I think the day they will be powerful enough is the day where they will have full OpenGL drivers and so on too.
Then its time to think again....

[will]

Everywhere I look OpenGLES is *replacing* OpenGL.  OpenGLES is substantially more modern than the small subset of OpenGL that MG currently uses.

The GPUs in mobile devices are typically non-shoddy and often faster than my laptop on which I comfortably play glest.

There's no lack of power on the PC and no limitations in its OpenGLES support from a glest perspective.

The big problem for mobile as I see it is the HCI problems with touch.   That's where most of the porting effort would go.

Of course, this conversation was resurrected by discussion about Google Chrome's NACL.  That is a way of distributing applications to desktops predominately, although later there may well be chromium tablets about.

[petevine]

I've built the original Glest 3.2 on my Odroid C1 (armv7, Mali-450 MP) and it runs using glshim from ptitSeb's fork (opengl 1.x to gles v1 wrapper).

The rendering is completely ok until an actual scenario is started though (crash due to lack of texture 3D support). Other devices should work too (Rpi, etc) so as soon as glshim compatibility improves this game could experience a revival on arm SoC's.

[tomreyn]

That's great!

It would likely still need a different input concept - MG is not exactly touch friendly currently. But other than that it could surely work with glshim. Also note that Titi + Softcoder are currently working on an SDL2 branch which may improve flexibility in this area.

[petevine]

Also note that Titi + Softcoder are currently working on an SDL2 branch which may improve flexibility in this area.

If you mean native GLES2 rendering that would be nice. However, from what I've managed to learn so far, porting to GLES2 is far from trivial (unlike GLES 1.x) which is the reason there's no GLES2 support in glshim and not many projects actually porting to it.

OpenGL 1.x support in glshim is being actively improved and has enabled many new games over the last few months. As long as no shader stuff is mandatory in MG it should be possible to handle it in the future. (probably)

I have no idea how useful it is at the moment but Android support has recently appeared in glshim.

[tomreyn]

Porting to GLES 2 would be a whole different project which is not likely to happen unless anyone will volunteer to help in this area.

[softcoder]

I started looking into this, we will see how it progresses.

[kagu]

@softcoder porting to GLES ?

[softcoder]

Just looking into it, i have a raspberry pi2 and will see if its reasonable to achieve.