Would you buy OS4 or MOS for x86?

[nyteschayde]

The people who own the 1% of all Macs ever made that are supported you mean?

For x86 you will need to spend 1000s of man hours going from PPC to x86.....then send 1000s more creating drivers for a fraction of the 1 million component combinations still in use that equal the x86 'standard'. That's the problem....x86 being a standard is an illusion and the Windows registry file he duct tape holding it together[badly].

You know it's funny. Everybody always brings up the thousands of drivers that would need to be written. I really don't think the situation is that dire. How is supporting one type of motherboard (or chipset) plus one or two types of graphics cards and same thing for sound and whatever else any different than buying the (very) few supported cards for a mediator or G-REX.

You don't have to support them all to make it work and we'd still be able to buy the hardware fairly cheaply given it's X86 hardware; much cheaper than buying the PPC hardware. Even if you only supported one video card, one sound card and one motherboard chipset it would be fine as long as you chose common variants to work with.

[koaftder]

I'd buy MorphOS and torrent Hyperion OS.

[nicholas]

I don't bealive in your words, do you spent 110 euros in MorphOS for x86 when for 90 you have a Windows or the best Linux free? OS4 and Morphos are for PPC or ARM systems, systems much better than x86.

I have no use for Windows but I use Linux, OSX, AROS, BeOS,Haiku and FreeDOS on x86 hardware.

OS4 is a joke compared to MorphOS IMNSHO and neither of them are available for ARM hardware.

[Karlos]

Guess what, all my MorphOS capable HW has exactly 1 OS installed.

NetBSD! :lol:

[nicholas]

I'd buy MorphOS and torrent Hyperion OS.

If OS4 was released for x86 but MorphOS wasn't, then I might be tempted to buy a copy if it was cheap enough.

[koaftder]

If OS4 was released for x86 but MorphOS wasn't, then I might be tempted to buy a copy if it was cheap enough.

I lost all interest in OS 4 when I bought a MorphOS license for my eMac. Of there had been a hw solution for OS 4 for a similar price as what I got that eMac for earlier I probably would have bought an OS 4 license. I can see going for one or the other, would there be any good reason to have both?

[Karlos]

Would people buy MorphOS/OS4 for x86 if they included no backwards support whatsoever for 68K/PPC applications?

Or are we just assuming that it's a given they'd both include some high performance JIT for such things?

[Pyromania]

Would people buy MorphOS/OS4 for x86 if they included no backwards support whatsoever for 68K/PPC applications?

Or are we just assuming that it's a given they'd both include some high performance JIT for such things?

I think it's assumed, after all WinUAE on x86 has a great JIT and backwards support for 68K. For PPC they would have to do something like Rossetta.

[Karlos]

For PPC they would have to do something like Rossetta.

A somewhat foolhardy assumption if you ask me. Firstly, despite being a JIT implementation, Rosetta's performance was certainly less than stellar on x86/x64 machines that were significantly more powerful than the class of PPC being emulated. Secondly, who is going to write it?

[nicholas]

I can't think of a single piece of MorphOS software that is both un-maintained and must have.

So PPC emulation is not absolutely necessary as a few endian tweaks to the source and a re-compile would be the most effort needed for most software.

[Karlos]

I can't think of a single piece of MorphOS software that is both un-maintained and must have.

Well, the entire classic version is unmaintained AFAIK, but as we're talking about MorphOS 2.x...

Isn't backwards compatibility with (OS friendly) 68K, PowerUP and WarpOS applications one of the key features of MorphOS? If all that were dropped, what tangible benefits would it have over AROS?

So PPC emulation is not absolutely necessary as a few endian tweaks to the source and a re-compile would be the most effort needed for most software.

I think you are over simplifying it somewhat. A lot of effort was put into optimising areas of the code for the PPC vector unit and while I'm sure there are generic versions of all these, porting the optimised versions to SSE+ is certainly no simple compiler switch change. Just look at the apple developer information pertinent to this.

[Templario]

Well, perhaps OS4 is a joke, but the people that use it we are more tolerant that the people than use Mor.... because if your words were to the inverse and the Mor.. was the joke in other forum, no, you have an abuse report, but I'm not so reactionary and radical as in the Mor..

[B00tDisk]

I'd buy OS4 for x86 in a heartbeat.  Even if sales only approached, say, BeOS for x86 levels compared to what it brings in now monetarily Hyperion would be drowning in cash (comparatively speaking).

[Digiman]

Well let's assume a certain specification of hardware shall we? We don't need to support every PC out there, just a subset. This can be done by specifying supported hardware up front, and deviances aren't guaranteed to work.

In this case, how about targeting AMD's forthcoming Llano processor and surrounding platform?

i.e., 2, 3 or 4 x86-64 cores, 160,320 or 400 AMD/ATI Radeon 6xxx series shaders, UVD3 (video decode), AMD/ATI SB9xx southbridge features (USB3, USB2, SATA3, ...).

No, and I will stop you right there. Isolating sections of the x86 'standard' people own would be like making a fuel that only runs in 2% of cars on the road. Your example would not cater for 8/9 of my machines....all of which play 1080p video files so not exactly ancient.

This is why Windows needs about 1gb of drivers on DVDs, and that humongous registry file.