@billt

There is developer documentation online at Apple:

General:
http://developer.apple.com

Direct link to Hardware & Drivers:
http://developer.apple.com/samplecode/HardwareDrivers/index.html


Anyone any idea if this information is enough to do some porting?


I will aready order a faast iBOOK. I think AOS 4 will burn the rubber!
Can you already smell it?

_________________
AmigaScene.nl

@T_Bone

Quote:

I keep hearing people say Macs are undocumented, yet it seems everything people writing OS's for the Macs comes up with, seems to work. Is there some specific thing people can't find documentation for?

Register and field names and descriptions, value names and descriptions for north/soughbridge stuff, wireless networking, and other chips, timing and buffer etc. descriptions for DMA things, preferred timing for laptop LCD panels, what Radeon BIOS data looks like compared to the Wintel target documentation we already have (Is it the same or different, if different how so?) as the driver does get some timing and memory information from the graphics card BIOS, basically ANY and ALL bits of information that would be needed by the kernel/HAL or any drivers to make the thing actually work.

I keep seeing people go on and on about Linux and BSD and whatever works on Mac so obviously documetnation is out there, but I don't see anyone posting links to this mythical documentation. Quid pro quo.

Hows about people stop assuming its out there and actually go find it, and prove the naysayers wrong, instead of us all arguing about the theoretical existence of it.

_________________
All glory to the Hypnotoad!

@sananaman

Quote:

Anyone any idea if this information is enough to do some porting?

Thank you, our first "documentation is out there" guy that actually found somethign to consider.

I've looked at a few things in there, some ATA stuff and iBook stuff, but as far as I can tell it decsribes what features an iBook has, and the MacOS API for making an IDE/ATA driver that runs in MacOS allowing it to use your new ATA hardware design. Kindof like having documentation on how to write a Picasso96 driver to the OS API, but which does not describe any particular hardware.

Yes, surely Apple wants it to be easy for people to find information allowing their new 3rd party hardware thing to be useable for MacOS users. The problem is, we want to use Apple hardware, not our own, and we want to use our own OS, not MacOS. So what is provided there appears to be backwards to what is needed to make OS4 run on Apple hardware. Kindof like if someone had a copy of the Picasso96 driver API but no chip-specific register or BIOS definitions cannot use the P96 API to create an Nvidia graphics driver for ANY operating system, since the Nvidia chip registers and BIOS datatypes are still completely unknown.

We don't need OS hardware access API descriptions, datatypes, or function names. We need chip register and value descriptions, which is not present there because Apple already wrote those drivers for MacOS themselves, there's not much need for a 3rd party to make new drivers for the same Apple-branded hardware.

I appreciate the links, and I encourage more. Perhaps if someone spends mroe time searching there they will find something I missed (I only looked for about 20 minutes) that can be used to enable Apple hardware in OS4.

First, we need to know how to talk to the Mac BIOS so we can begin collecting some of the same information we get from uboot, such as what IDE/ATA chips are present and how do we attempt to boot from a hard drive or CDROM, as well as what other chips are present (Voodoo or Radeon or Nvidia, which particular network/sound/etc chips are present, their addresses, how much system memory is present, etc...

http://developer.apple.com/documentation/Hardware/DeviceManagers/pci_srvcs/pci_cards_drivers/PCI_BOOK.3a.html#27578
http://developer.apple.com/documentation/Hardware/DeviceManagers/pci_srvcs/pci_cards_drivers/PCI_BOOK.59.html
http://developer.apple.com/technotes/tn/tn1061.html
http://developer.apple.com/technotes/tn/tn2001.html
http://developer.apple.com/technotes/tn/tn2004.html

There's a very basic beginning. Perhaps there's a list of exactly what chips are in each particular Mac model out there, or you could use one of them spiffy Linux distros an do an lspci to find out. I don't have any idea what might be off-the-shelf chips (sound, ethernet, usb, ATA/IDE) and what might be designed by Apple (northbridge, southbridge?, wifi), but if we had a list of off-the-shelf standard chips we might get the requried documentation from their respective vendors, sich as Radeon graphics documentation from ATI (AMD?)

What the documentation isn't available people are concerned with is the designed-by-Apple chips. Apple designed the chips, and they also wrote the MacOS drivers. Nobody else needs to write MacOS drivers for these chips because Apple already did that, so no reason to post the register defintions publically for people like us.

I've started some Mac BIOS links, and the challenge continues, can we find all required chip register documenation to make one _entire_ Mac actually work? And by actually work, they all and especially laptops must also support available power managment in OS4...

_________________
All glory to the Hypnotoad!

@billt

First of all I like your enthousiasm! That's great.

Well I think if we gather all the information and do research to what is needed, we've probably done 20%-25% of the work.

I also remember a hacked version of AOS 4.0 that works with PearPC. There might be some interesting parts in there. I've got this PearPC configuration + hardfile somewhere in my "applications" folder ;-P. If someone needs it give me a yell.

Another thing is the UBOOT. I've looked into the sourcecode of the "generic" UBOOT from sourceforge (http://uboot.sourceforge.net). But there's not much interesting to see there. It's very basic. There's a lot more into the Amiga UBOOT version if you aks me.
Is the AmigaUBOOT version also open source?

I see the following steps:
- Gather all information about Mac's hardware (from any source).
- Figure out why AmigaOS 4 is running that easy on PearPC.
- Get a UBOOT like system working (just the basics).
- Get the Kickstart loaded from UBOOT.
- Graphics driver (we need to see something ).
- Load parts of the OS and show it through graphics driver.

I think a Phase 1 is then finished.
Then the Finetuning starts; with things like Networking, PCI, USB, Audio, finetuning graphics supprt, etc.


Yess, I would vote for such a project!
New Amiga Work Group?

"AmigaInAMac"....

_________________
AmigaScene.nl

It seems a port of OS4 to Mac hardware is very unlikely but how hard would it be to produce an emulator for the Mac able to run OS4? Of course the problem is that since OS4 is unavailable without hardware people would be forced to use pirated copies of the OS. BUT would the threat of a release of an emulator such as this change the mind of Hyperion/Amiga Inc about producing a port for the Mac?

Another thought, perhaps the emulator would run on x86 Mac's too?

Skunkfish

_________________
Currently planning to upgrade my Amstrad CPC

@sananaman

Quote:

I also remember a hacked version of AOS 4.0 that works with PearPC. There might be some interesting parts in there.

I'm not looking for a hack. I'm asking for legit documentation that would allow a legit and fully licensed port of OS4 supporting all hardware features without having to reverse-engineer open-source stuff from Linus or whatever. No hacks, no grey-area legality. Full-stop.

I'm sitting here making a fool of myself because of all these guys saying that the "lack of hardware documentation" is not an acceptable excuse to say no legitimate OS4 on PPC Macs. My proposal is for these people to point to where this missing documentation can be found, and thus prove these "lack of documentation" arguments are bunk, or if this challenge cannot be completed then stop whining about whatever means Linux and other OSes came to work on Macs and shut up about it. Hackers, emulators and pirates belong in different "projects", not this one.


Quote:

There's a lot more into the Amiga UBOOT version if you aks me.

I didn't realize there was any special AmigaOne addtions, but yes it should still be open-source based on the original uboot's license. As I remember thre had been some new additions like command caching and up-arrow history stuff that the "real" uboot team refused to apply to the main sources.

Last edited by billt on 03-Oct-2006 at 04:52 AM.
Last edited by billt on 03-Oct-2006 at 04:06 AM.

_________________
All glory to the Hypnotoad!

@billt

I would try to find info from Agere systems and Broadcom as they make the G4 and G5 northbridges.

_________________
The only spanish amiga news web page/club: CUAZ

@afxgroup

Quote:

afxgroup wrote: @Mark Quote: At the end of the day OS4 is a closed source OS ( as I mentioned previously) this means that the devs cannot "look at" the source to Linux which is GPL'd to develop OS4.x. 3rd parties can write drivers based on Linux drivers as long as they are GPL'd drivers but thats as far as it goes. this is not true since you can try to undestand how they initialize the chipset for example. you cannot use it's code but you can understand how the code works

Add to that there are also the *BSD varieties of OS's available, and from those you can take as much as you want.

Also, hear hear, you could use an "hosted" approach, just like AROS does. Indeed, AROS already runs on PPC Macs! Now, just put MOS compatibility in AROS, add OS4Emu to the mix, and you've got an Amiga-like OS running on PPC Macs and able to run MOS and AOS4 programs. Who'd need AOS4 anymore?

_________________
“It is well enough that people of the nation do not understand our banking and monetary
system, for if they did, I believe there would be a revolution before tomorrow morning.”
~~ Henry Ford

Um, one thing I haven't seen covered by the thread so far... who's going to pay for the development effort up front?

_________________

@falemagn

Quote:

Also, hear hear, you could use an "hosted" approach, just like AROS does. Indeed, AROS already runs on PPC Macs! Now, just put MOS compatibility in AROS, add OS4Emu to the mix, and you've got an Amiga-like OS running on PPC Macs and able to run MOS and AOS4 programs.

I am afraid there are not many developers with the required knowledge to add MOS compatibility to AROS.

If a lead coder like you showed how to start maybe it would be easier.

ATM the only thing I know is that structures need to be aligned and the aim is keeping the public structures between:
#include AROS_COMPATPACK_START
&
#include AROS_COMPATPACK_END
so it uses the same alignment as MOS/OS4/68k

But what about the rest?

When you worked in the Ariadna Box you probably thought about it... did you made some tests with simple MOS commands/apps? Did you plan to add a 68k emu?

_________________
The only spanish amiga news web page/club: CUAZ

@gnarly

The most important point is overlooked - we do not have a license to produce a Mac version of OS 4.

ALL other points (and yeah I mean ALL of them), hence, are moot.

_________________
Seriously, if you want to contact me do not bother sending me a PM here. Write me a mail

@gnarly

Quote:

Um, one thing I haven't seen covered by the thread so far... who's going to pay for the development effort up front?

The point is, this would (or should?) be an investment. The assumption, here, is that porting AOS4 to the Mac would boost its sales (as would porting it to x86, but that's another story), therefore the development would pay itself in the mid run.

Of course, whether or not that assumption stands valid, is an issue open for debate.

_________________
“It is well enough that people of the nation do not understand our banking and monetary
system, for if they did, I believe there would be a revolution before tomorrow morning.”
~~ Henry Ford

@Rogue

How much would it cost to get such a license?
I'm not a programmer, but since the concept is ppc, i reckon it wouldn´t be too hard to port os4, right?

Regards,

Ricardo

Last edited by estik on 03-Oct-2006 at 10:52 AM.

_________________
Power Mac DP 1.8 GHZ, 4 Gb Ram and still trying to get EUAE in full speed

@Rogue

Quote:

ALL other points (and yeah I mean ALL of them), hence, are moot.

No, they are not, since if they were, you'd not have spent that much time in the past trying to counterargument them.

And then, licenses can be obtained, hm?

_________________
“It is well enough that people of the nation do not understand our banking and monetary
system, for if they did, I believe there would be a revolution before tomorrow morning.”
~~ Henry Ford

@Crumb

Quote:

Crumb wrote: @falemagn Quote: Also, hear hear, you could use an "hosted" approach, just like AROS does. Indeed, AROS already runs on PPC Macs! Now, just put MOS compatibility in AROS, add OS4Emu to the mix, and you've got an Amiga-like OS running on PPC Macs and able to run MOS and AOS4 programs. I am afraid there are not many developers with the required knowledge to add MOS compatibility to AROS. If a lead coder like you showed how to start maybe it would be easier. ATM the only thing I know is that structures need to be aligned and the aim is keeping the public structures between: #include AROS_COMPATPACK_START & #include AROS_COMPATPACK_END so it uses the same alignment as MOS/OS4/68k But what about the rest?

The simple answer to your question is: the ABI and API have to be the same. Therefore, any effort in making AROS compatible with MOS has to be leaned towards that goal. The goal is not hard to accomplish, as AROS and MOS share a lot of the core source base.

The detailed steps to achieve that goal, I'm afraid, are out of the topic of this discussion.

Quote:

When you worked in the Ariadna Box you probably thought about it... did you made some tests with simple MOS commands/apps? Did you plan to add a 68k emu?

No and no. The port was an almost straight recompile (except for a few issues related to the lack of a floating point math unit in the Ariana processor). Any other improvement would have come later.

_________________
“It is well enough that people of the nation do not understand our banking and monetary
system, for if they did, I believe there would be a revolution before tomorrow morning.”
~~ Henry Ford

@Rogue

Quote:

Rogue wrote: @gnarly The most important point is overlooked - we do not have a license to produce a Mac version of OS 4. ALL other points (and yeah I mean ALL of them), hence, are moot.

i hate to get on your bad side rogue but i really have to say this. a license is no excuse. if a project needed to be done and you wanted to do it you would either find a way around the license or (re)negotiate the license. hiding behind a current license makes no sense. if you can present a buisness plan that ainc wants to meet and say to meet this requires targeting pre-existing systems if the benefit is huge they might bite and give additional leeway.

now this discussion is to determine (in the communities eyes) if it is even possible. even though this and other topics have already been covered before, talking about these things does not hurt any company and actually shows the community is interested. i would be more worried if nobody discussed anything about os4.

personally i believe the mac hardware pursuit is useless since there is no longer any new macs available. but do you really blame people for wanting to use faster hardware than what has been made available? do you really blame them for wanting a laptop? you can't expect to create a product in any field and any targeted group without getting some criticism or ideas on where to take that product. i have been involved on several forums for several programs and no customer is completely satisfied. this is healthy since this helps to push the evolution of the product.

anyway i will wrap this up. catch ya later.

_________________
Win•dows: n. A thirty-two bit extension and graphical shell to a sixteen-bit patch to an eight-bit operating system originally coded for a four-bit microprocessor which was written by a two-bit company that can't stand one bit of competition.

@afxgroup

Quote:

this is not true since you can try to undestand how they initialize the chipset for example. you cannot use it's code but you can understand how the code works

Ahh but then you get into shaky legal ground, just look at the MOS arguments that abounded a while back, in which it was argued that because the programmers had, had access to the source of OS3.x, that MOS was an illegal copy of OS3.x (Note to trolls or other interested parties Im not making any statements as to the accuracy of the comments, simply using the case as an example). Using this you could make the same arguments about using GPL'd code to understand how something works, would mean that you may have to make OS4 GPL. Im not saying it necessarily is true, however it would definately leave it on shakey legal grounds, which no professional company would want.

Mark

_________________
IceStar Media Ltd.

@SoundSquare

Quote:

Sonic, MOSnet dev on MOS almost made possible to boot OSX natively (without mol) on Pegasos, his project is called bootx.

No, Apple's project is called bootx (Apple: The boot process)

Sonic's intention is to be able to boot OpenDarwin - booting MacOS would, of course, be a total violation of the license...

@nine

It's an EULA thing, and as such moot... you had payed for it, you have the right over what you buy, it's that 'simple'

_________________
Indigo 3D Lounge, my second home.
The Illusion of Choice | Am*ga

@Mark

Quote:

Using this you could make the same arguments about using GPL'd code to understand how something works, would mean that you may have to make OS4 GPL. Im not saying it necessarily is true, however it would definately leave it on shakey legal grounds, which no professional company would want.

Following source code is distributed under GPL:

int main(void)
{
printf("This program is distributed under General Public License (GPL)\n");
return 0;
}

Oops, what I have done.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook