This is how we do it in the firmware. Actually it's a bit of a hack, still. For 16 and 32bit modes, it is just a lookup table of stored values..

Does anyone need 8-bit anymore? MacOS X won't run in 8-bit and Windows XP hides those modes.. I think even 16-bit modes are having their day (since there is no alpha channel, those fancy window effects would be and are disabled in Vista in 16-bit)
Yep. 800x600x32bit is what XP boots into. If you ever bought a monitor in the last 5 years it HAS to support that mode to get a Windows logo.

All TV cards since 8 years support it.

And previously, if you had a monitor that didn't support that mode.. well.. SuperVGA resolution, if you don't support that, you're well and truly in the wrong business if you're a monitor manufacturer.

_________________
Matt Sealey

During playing with DirectHardware VGA driver (which doesn't work yet) i've found out an interesting thing...
CRT controller is programmed using two ports: register no and data. And if reg number port is N and data port is N+1 there. In original code (okd XFree86) an outw operation is used to program it. On PPC it is implemented as ONE word-wide store operation and eieio after it (i've copied the operation from Linux kernel, but it's the same anywhere else). My question is: is it okay? I assume it's not. This approach does not guarantee that N gets written before N+1 and the things go wrong way. Am i right here?

eieio guarantees that IO writes are done in memory address order (N, N+1, N+2 etc.) so it actually works.

A lazy performance trick

However if you wanted to be truly, totally correct about it, you could do two byte writes, with eieio after each. This is slow, though (as store reordering, write combining etc. gets disabled for the entire store queue), and for graphics operations you want to spend the least time waiting on the card as possible.

_________________
Matt Sealey

Dammit, seeing a reply in this thread gave me hope of some sort of progress update, and then it's just spam.

Sorry about that! We will reach out to one of the Darwin Developers that has an ODW. He is also on the Gentoo Power Team so he is still around. Apple announced the end in July and then in August Darwin 8.7 (Mac OS X 10.4.7) was released. We are not the people to ask where Darwin is headed, but we will check on the situation.

R&B

_________________
http://bbrv.blogspot.com

At least, the major project drive is dead;

http://www.opendarwin.org/en/news/shutdown.html

What point is there now? Apple moving off to Intel, they will drop Power in the end. Running Darwin on it's own with X and so on is no different to booting FreeBSD.

Time would be better spent getting one of the other operating systems off the ground. FreeBSD, NetBSD, both are good candidates and share similar system design (in fact Darwin's toolset is based on FreeBSD, as Linux's toolset is GNU)

_________________
Matt Sealey

Running Mac OS X native on the Pegasos is "point" enough for me, and I don't see how focusing FreeBSD or NetBSD would help accomplish that. Just my 5 cents.

My work's primary goal is MacOS X, not Darwin itself. To tell the truth i don't see any need for BSD since there is Linux and i see no task which BSD can do and Linux can't do. From the application point of view they are the same.
MacOS X is useful by itself because it provides a possibility to run many commercial apps which are helpful in everyday work.

Next year, and the year after, and when MacOS X stops being produced for Power Architecture?

Are you going to wait a year for this app to run the OS, and then just stop the next?

Do you still run MacOS 8.1 on your system under emulation, for example? How useful is it?

I think efforts are best placed where they are most needed. MacOS X on Pegasos is fast approaching becoming a toy.

_________________
Matt Sealey

Heh...
If we follow this way of thinking, Pegasos is a toy by itself. The only operating system in production is Microsoft Windows. And this is the only operating system with non-obsolete software.
From the marketing point of view Linux is nothing. It's simply a next replica of ~15-20 years old technologies for which noone is going to pay, this is why it's freeware and opensource - it's just a crap that noone needs.

So why do you want to use MacOS X?
You can say the same of MacOS X.

I don't see the point in spending a lot of time trying to support an OS which will be actively unsupported by the vendor in a matter of a year to 18 months, and then only on a life support machine for security updates.

Don't get me started on the developers and users; they have already started throwing away their PPC Macs for Core Dualies. Even Bill has one.

OpenDarwin is a dodo, too, so no other development of this technology is going to go on. I have no idea what happened to our machine but we got absolutely zip from these guys before they shut down. I expect it is in the trash or in a dusty closet.

Where is the marketing advantage in running MacOS X on Pegasos, that you so deride Linux for not having?

_________________
Matt Sealey

Sonic does it for himself and for Pegasos users. Not in a marketing perspective. That's hobby stuff and you should better encourage him in doing that instead of pointing out the useless aspects of it.
So OK, the amount of Pegasos users is not that much, OK the number of Pegasos users interested in having OSX running natively on Peg is even less, OK it's pointless as a marketing thingie for Genesi, but this is not the point here. It's about a guy that is engaged in a challenge and what he will learn from it is worth the effort i think. It can't be a waste of time.

Now what does OSX does that linux doesn't on PPC ? modern browser with flash ? Skype ? and access to some amazing OSX apps for music, video production ?
Yes, only some popular stuff for users that linux can't provide yet.

Oh yeah, i'd love to use cubase SX on my Peg, work on my own music with it and claim "i made it on Pegasos". something i can't do for now. And it's only an example.
You think in terms of business/marketing only here, i think in terms of FUN !

Would you reboot out of Linux or MorphOS just to run Skype?
I didn't think of business/marketing, but from a project management point of view (that is what I do!) - at the end of it all you have to make a feasibility study, feature study, cost evaluation, schedule your project time, and look at the end and benefit for the customer of everything, and looking at "porting OpenDarwin", that was a good project until it slowed to treacle, Apple didn't update for 18 months, everyone quit and then they shut it down.

Running MacOS X on Pegasos was fun, exciting, useful back in 2004. We had a demonstration system at SNDF Dallas in April 2004, with MacOS running under MacOnLinux (it booted directly into it). It was interesting and intriguing to some of the visitors to the stand. We are coming up to 2007 - Apple have not got a single Power product in their line-up anymore. Eventually it will stop being released for Power, and before that, some new features and certainly optimizations will stop coming.

If you want to make a business case out of it, why not think of one for running CuBase on Linux? Or getting the Skype guys to make a Power version of their Linux client?

_________________
Matt Sealey

@Neko

many people registered PegXMac to boot directly into macosX (macos9). All of them wants to use 3d acceleration in two(three, four, ....) years old games/ apps/systemsoftware than using uptodate linux crap soft.


@Sonic
any news about project?

Park of PowerPC Mac's is too huge for momentary abandon this supply. Coming OSX 10.5 "Leopard" still duo-architecture-supported, and tons of application has universal binary (tm) versions.

_________________
wintel free