Sorry, Kolibri is for x86, not for ARM.
You can launch it in QEMU, but I assume you'll be worried.

Why not create KolibriOS under ARM? Linux in fact create.

Because you'll have to rewrite it from scratch. Only some ideas and about 0% code can be used from KolibriOS on non-x86 architecture

It is not true. Programs written in HLL can be ported.

Erm. When I'm talking about OS (e.g. KolibriOS), I mean only OS, i.e. kernel itself.
Sorry, if it wasn't obvious.

How do the case with Linux (eg Android)? Rewrite all the code or is the compiler?

Nable wrote:Because you'll have to rewrite it from scratch. Only some ideas and about 0% code can be used from KolibriOS on non-x86 architecture

It is already done to some extent.

Kolibri box in Phantom OS is not a complete Kolibri, but quite close to, and is being ported to Arm right now. So it will be possible soon to have Kolibri on ARM. We just need to redefine API to map Intel registers to ARM registers, which is simple. And port some highlevel Kolibri code to ARm as well.

Here is Kolibri emulation layer in C:

http://code.google.com/p/phantomuserlan ... /kolibri.c

redefining registers quite simple, because of 15 registers on ARM

Am I the only one missing something here or are you going to map Intel instructions as well? Because I don't see the purpose of having Kolibri emulation layer on ARM without any Kolibri programs for that platform. Regarding HLL-based ones, they could be ported to other architectures even easier *without* Kolibri layer. So if there's nothing Kolibri could do on ARM, why are we having this conversation in the first place? And moreover, what dzavalishin is now doing (apart from trying to prove something to someone) is just an API wrapper for not that good API, so when he's done (if ever) it'll have nothing in common with Kolibri itself except for pretty name.

mike.dld wrote:Am I the only one missing something here or are you going to map Intel instructions as well? Because I don't see the purpose of having Kolibri emulation layer on ARM without any Kolibri programs for that platform. Regarding HLL-based ones, they could be ported to other architectures even easier *without* Kolibri layer. So if there's nothing Kolibri could do on ARM, why are we having this conversation in the first place? And moreover, what dzavalishin is now doing (apart from trying to prove something to someone) is just an API wrapper for not that good API, so when he's done (if ever) it'll have nothing in common with Kolibri itself except for pretty name.

Has anybody ported it or made any progress?

Thanks

Jeff

jeff1909 wrote:I'm a little late to the game here but I was wondering what the update on this was?

Has anybody ported it or made any progress?

Thanks

Nope.

That's all for today, afaik. But author is ver proud of this though