Be on the G4

adraken writes "Jean-Louis Gassée of Be Inc. talks about the recent release of IBM's PPC motherboard specs, the G4, and what BeOS will do about it in Be Inc.'s latest newsletter."

Comments

reverse engineering linux / BSD

[mcc]

"Some have suggested that we look into the Linux sources for such data. Perhaps, but I see little reason to open ourselves to possible accusations of reverse-engineering. We're welcome on x-86 hardware, we're not welcome on Apple G3/G4. We respect the logic and that settles it for us. "

"Accusations of reverse-engineering?"
Ignoring for a second whether it is logically _possible_ to reverse-engineer linux, who would accuse them of anything, and what bad could possibly come out of it? As long as you aren't actually reusing code, i don't see any way they could violate the rights of a GPLed program.

What if they just had one engineer read the linuxppc kernel source, write down everything you have to do to work on a G3 chipset, pass that information to another couple of engineers and have them put that into Be? They wouldn't be copying any code, so it would be perfectly legal. Does it even have to be that complicated?
I could understand if Be just wasn't comfortable getting information on the G3 chipset from a secondhand source, but that's not what they say the problem is.

Hell- forget linux, what about NetBSD? or Darwin? They could just take that code and it would be legal, wouldn't it? Is there anything in the ASPL that would at all limit Be's usage of code from Darwin?

I just wish it was possible to get more than one side of this story. All that we have to go on is what Be says, and while it sounds like it's probably true, i'm not certain how difficult it would really be for Be to work out the G3 if they wanted to. And maybe apple just has better things to do than provide tech support to Be?
-mcc-baka

INTELLECTUAL PROPERTY IS THEFT

if chipset it standard, effort to port is minimal

[Clith]

During the "Genki" beta [was that for R4.5?] there was a kernel produced [genki7] that ran on Apple G3's. I think it was done by an intern who then left, so the code wasn't supported and got axed.

So, given the following qualifications:

• the IBM platform uses standard, well-known chipsets
• it's easy to port the Be kernel to PPC variants

then it stands to reason, that it *would* be in Be's best interest to support the IBM PPC platform [does it have a name? Is it CHRP? PREP? NuCHRP? :-)], since their cost to do so is minimal and can only grow revenue.

Now one thing Be really has to do to be a real presence on PPC is to move to the same development environment [compilers] as the x86 side -- GNU cc. Otherwise, they have to do tons of software support, which ups the cost of supporting the platform. This bullet should have been bitten in R4 at the same time as it was for x86.

BeOS running on a quad-G4, each of which has a dual CPU core -- hey, I can dream, can't I? :-)