I understand, BUT the older discusions are about porting the engine to a diferent language, like Java. Or try to run on Windows mobile or Iphone based OS.
Android is a diferent matter, first off, it does support OpenGL(in fact you can indeed run OpenGL games for x86 using Dosbox), and units with embedded GPU has enoght power to run the FSO graphics engine. Second, its Linux, and people have had some success in compiling C/C++ projects on it.
So its not imposible like trying to port the game to Java...
As for input discussions, this is not the time to discuss that.
Anyway, i think the first step will be trying to compile the bins for Linux ARM, maybe a Ubuntu running in parallel with Android on the tablet.
No, you do not seem to be understanding what we are saying to you. Those previous discussions covered many of the same things, you seem to be focused exclusively on the apparent ability of the hardware and/or OS to run unmanaged C/C++ code, that is just one of many issues related to porting.
I will give you the different language bit, though if was the only problem it wouldn't be a deal breaker (though it will still be more than 4000 hours of work (that is, two years full time for one person)).
The iPhone uses OpenGL ES, in fact there is very little that uses 3D accelerated graphics but doesn't support OpenGL at all. With some quick googling the only platform that I can find that doesn't support OpenGL at all is the XBox (which really doesn't come a much of a surprise considering they are DirectX boxes
); the PS3 supports ES, some of the Windows Mobile Phones actually use ES to implement the DirectX API, the Wii supports ES, so do the the DS and PSP.
OpenGL is never the problem with these ports. Having to rewrite everything to use the ES dialect of OpenGL is the first problem (which if I am remembering my OpenGL feature sheets correctly means you need to lose the shaders for one, which is basically the source of all of the "modern" graphical features of the engine). Next would be finding a GPU with enough grunt to render and enough bandwidth to push through the amount of data that we throw at the GPU is a different story. Third would be finding something that has storage that is fast enough to stream the data that we do. Fourth would be finding a CPU that could handle a mission full of physics/collision calcs. Also, finding enough RAM to run the game could be a bit of challenge as well.
Yes, Tegra can do some pretty nice things, but that is when you compare them to other embedded graphics cards. Even the lowly Intel GMA's are able to soundly trounce a Tegra 3.
This is the perfect time to discuss input because there is
no point in porting anything to another platform if you can't use it. Remember the discussions about how to port FSO to a game console? Or even the, how do you map the 30+ controls of FSO onto a game console's controller? In this case you are likely going to a device that has no buttons just a touch screen, which means you have what? 1 (maybe 2 or 3 if you have an expensive multi-touch capable device) simultaneous "button" press (which btw would have be on a "keyboard" that covers half or more of the screen).
You would be much further ahead to do something that is Freespace themed like
Polpolion's BASIC "port" or
Nuke's top down mod for FSO. If you built/ported the engine that you use for Freespace correctly you could get the other TCs to help and/or do their own version.
Compiling for ARM will probably be least of your problems because the last confirmation that I heard that FSO still compiles and runs on PPC was about a year ago, which covers the endian issues. FSO also compiles for 64-bit on SCP_UNIX systems already which has also helped find quite a few CPU related portability issues.
Don't forget the tired old maxim of the computing world: high performance, low price, low power consumption; pick two.
Keep in mind this is the 40 minute high level overview (ie. I spent 35 minutes too much to explain this). This is not an exhaustive list because I don't know everything about how the engine works on all platforms.
Now don't get me wrong, I would love to see FSO on more platforms, but just porting the engine as is... it will be incredibly disappointing.
Its great that you have an plan of action now (even if it is only one step long). Please keep us posted on your progress.