Quote:

CoolCat, I am very sorry, but I can't follow half of what you are saying. Can I get a CoolCat to DFerg translation?

Quote:

Holley, my intention is either 1 or 2. Not 1 and 2. But from what you say then DFB is possiable. If that route is taken I don't think I have the money to start both bounties. Someone would have to to start the DFB one. However that would not prevent one person form taking on both.

This is good. Talk is good. I don't quite follow everything. But, if a concensus can be reached and the specs decided on. the bouty will be posted.

i, personally would prefer native investments allways, as in the long term, native should be prior.

but the ideia is not bad, and maybe it can be very simple (like morphix mods)...i thought on build something that would be similar to this ... a minimal linux enviroment with aros in it...like the uae/knxlight of cloanto for aros!

i wonder if it would be possible to call the host dev enviroment from inside aros...cause the only real good from this kind of task would be have a complete dev env setuped....under hosted, as atm native miss it...and thats why i make the posix bountie...instead of using a workaround i did go to the root of the problem.

in a second view, if aros had a native dev end (we just miss a nix shell atm afaik, we would still need some way out...can be fat32, cd write or tcp/ip, thought)

and just for remember, there was once a talk about using a wrapper for X drivers so native could use it.

best regards,

Maybe need to split it down a bit from that, ie. for 1 ask for a framebuffer driver for Hosted, for 2 ask for the modular kernel that would allow porting of GPL drivers.

GPL drivers could be pulled from wherever, and yes I agree would be better in the long term. Taking Hosted down to a lower level could make for a more quick n' easy solution to get more things to 'just work'.

OK, going to start a abounty. Now what will it be. That depends. I need some input.

Option 1:

custom linux kernal with Aros laryer on top. Linux to be "hidden" as much as possiable. Someone suggested the Amilthon kernal.

Option 2:

I have seen it tossed around before, but not in a long time. Making Aros be able to use Linux drivers or FreeBSD drivers.



Personally I think that option 2 is better. If I understand correctly this would be more "natvie" then option one. But, I don't really know. I know that it was talked about, but I dont' know if it is really possiable. If it is, which one Linux or FreeBSD?

On the other hand. What are the pros and cons of each one?
What do we gain/loose with either option. That is of course that option 2 is even doable.

What about X, if we go with option 1, do we need X? Can we use the Direct Frame Buffer? I know not as many drives, but more than we have now. Is this an option? Is it desireable? Is this idea, POOP?

These are things that I would like to know before I start a bounty that may or may not go completely against what the devs wish. yes, I want aros to be done. No, I don't want to go too far away from where the devs are trying to go.

I can and will start this bounty with $100.00 as it is all I can do for the moment. (I say this to show that I am serious here, not just talk).

So, lets have some real, NICE, PRODUCTIVE discussion.

Also, which ever way it goes. I would like to have the aid of a dev to write up the specs of the bounty. That way it would clear, and not "over the top"


Maybe a "pole" if the two options are close to doing the same thing. I would prefer a Native Aros, so for me the option that gives Aros a boost while being able to propel the Natvie version would be good.

However, I am not "set in stone" either. If the benifits of going Linux kernal + Aros on top outweigh the other. Then that is cool.