Hi DFergATL,

i think it will not be hard to write those requirements, but atm, after the dbus happening, i dont think its currently a priority, so for the time being i will refrain from work on it.

we have some nice things going on, like fat32 and tcp/ip, but otoh our dev-env got stalled...and after some thinks to myself, i will focus on how make it happens!

sorry, but after dbus, been against my feelings will be like i am negligent myself. (and im sure someone else can write those for you)

Mr Cat. Could you write up the requirements?

Quote:

See how much I know. Does that mean that it could boot from an AFFS partition?

OK, who is going to write the requirements? My only personal wish it that it be as seamless to the user as possiable. I don't want to have to mess with the linux stuff. We should still be able to use the Aros Partition app and install just like we do for Native install now. Change screen res..etc... all from within Aros. We should only use as little of the Linux kernel as necessary. As much as possiable should be on the Aros side. Can things like the TCP/IP stack being developed be used like this? Or does this mean that the TCP/IP stack in the Linux kernel must be used? Just curious.

AFFS would only allow you acess your aros partition from inside linux, but once we have fat32 support we could use it instead!

regards,

If we had AFS support in the kernal, that would prevent us from having to have one partition for booting, and an other for Aros. Also, couldn't some things that aren't needed be removed from the kernal? There are a lot of minimal kernals out there that could be used and might just need AFS support added. I would prefer to not use X, in the case of Aros it would just be adding extra weight and using up ram. The idea, I think, would be to use as little of the Linux kernal as necessary.

I am not sure if the only thing that we would be using the Linux kernal for would be drivers. Are there other parts of the kernal that could be accesed for added functionality?

I still agree, Native all he way. Howerver, that is a mighty tall order for the number of devs. My thinking is this...use linux for now. Hopefully attract more devs, users and migrate over to Native. It would take a lot of pressure off the current devs, get it to a point where porting apps would be pratical...get more devs....get more users.... and so on and so on... There is a lot of pressue on the devs at this time and just not very many of them. My opinion is that it would give Aros a jump start, releave some pressue on the current dev team and maybe, just maybe attract more devs as they see that Aros can actually be useful.

Really guys. If just one software vender ported an app to Aros, it would give Aros more than any other Hobby os out there. I am not a huge fan of Open Source apps. They are ususally not finished, not well tested, poorly documented...and amost impossiable to support because of that. Trust me on that I work a support desk for living.

And on that note. Think of this. Amiga was always known for it's media capabilities. Now if some of those apps could get ported to Aros....imagine that. FYI - Babalon 5 had all its CGI done on an Amiga, an old Amiga. Can you imagine Cinema 4D on Aros and modern hardware? I know this is all speculation. But does SkyOS even have that type of thing as an option? And Linux for all it's hype has very little desktop software being written for it by actual software vendors.... Peopel love Amiga, but don't like the price and hardware of the new one. I really think Aros could be big...well not small anyway...

I dont't think that Aros devs need to spend a whole lot of time making apps.. There are plenty of apps out there. They just need an OS....

My, .002 cents.

Native : Use a wrapper for the linux X drivers (that would be a wide shot, but not the best solution)

Hosted : Having aros hosted in a minimal linux enviroment wouldnt upgrade nothing from dev pov, but maybe from the user pov, as it will be customized for AROS (we can have for example the real cross compiler, that is not that easy of build).
otoh, for use the linux side of the dev env could break the AROS env imersion, that why i asked if it would be possible acess the linux dev env from inside AROS.

If we use the Fbdriver would allow us to avoid the Xserver.

and still we could have a sdl hidd but dunno if it could avoid X, but would make port aros for other archs easier.

Im not sure how a custom kernel can help (besides enable affs support)

i hope clarify my points.

ps: i think Native is the priority (personal opinion), the only advantage for hosted is have a complete dev env that is missing under native, atm.

:vik1:

***EDITED***
we allready have a fbdev driver, read the topic "some output"...just search aros-exec for e-uae.

***EDITED2***
i can make, in a couple of days, a morphix based dual boot iso for knoppix and aros...it will not be customized only packed together, but imo is pointless...i thik staff had make some work in this way using a HD image!

Holley, thanks.

The focus part is why I asked for help on defining the actual bounty. I don't want it to be vague, but I am not qualifed to actually write it up.

So, it would seem at this point the Linux + Aros on top is the way people are headed. Unless I am completly lost, which is entirely possiable.

The smarter people on this board can decided which way and define what it is. If it needs to be broken up, that would be fine. I can split what I have between them. Basicly I am saying I would start off the bounty by donating the $100.00, split across multiple bounties if that is what is necessary.

At some point someone will write it up, let me know and away it will go..

It is all I can do.

Dang it, want Aros!!

DFerg - What I was trying to say was that for either bounty to work they would need to be more focused on what they were trying to accomplish.

CoolCat is talking about being able to more-tightly develop in Linux hosted. While either solution would result in better hardware support a lower level of hosting in Linux would help with development, and the other would help the Native version.

we allready have a framebufer driver...but its not complete/stable, thought!

One thing I would like to have cleared up for myself. Do both of these options provide the same thing? That is more driver support? Or does intragating Aros and Linux give us that plus other things? If so, what???????????????