Amiga OS - Why on custom architecture?

[gafstu]

Guys, You may be  able to answer a question that’s bugging me.

Why does the architecture for OS4.0 have to be custom?  This makes it expensive and possibly less reliable than mass produced alternatives.  Why not build the Amiga OS to run on a IBM PC style architecture, then all the hardware availability issues will go away??

[DonnyEMU]

I have been asking this question on AW.NET myself. Apparently there are just too many people that love the PowerPC platform, and don't care about cost or competitive pricing to keep the platform alive.. It's ashame..

Donny,

Don't forget, there's the cursed "pact of the holy triumvirate" which keeps them from even thinking about saving the platform without selling closed-end hardware.

Humourously enough, I am starting to believe that AROS is the only one that'll survive.  It is definitely the only one with the right idea.

Wayne

[Prmetime]

There has been all kinds of talk about why would anyone use the Amiga OS on PC hardware when they have windows?  I have heard those arguments before but I don't really believe them.  Now it is kind of a moot point.

If I understand correctly, there won't be any new PPC architecture after the current boards unless there is a large order from a commercial concern. That pretty much leaves the x86 hardware or nothing. Either that or the hope that China or somebody else decides they don't want to be beholden to Microsoft.  We can only wait and see.

[alx]

I'm afraid that this has been asked more times before than I care to remember.  There are some very good reasons why AOS should be ported to commodity PC hardware, but here are a few on the other side of the argument:

1) The A1 is definitely a lot less custom than previous Amiga hardware - it basically is "IBM PC style architecture", but with a PPC CPU.

2) If OS4 was on PC hardware, people would be able to use Windows applications, so there wouldn't be an incentive to develop OS4 applications.

3) The Amiga has been migrating to the PPC anyway - OS4 will be able to run WarpUP applications natively.

4) It would be impossible for Hyperion to support all possible x86 mobo configs - and if they only supported certain mobos/cards etc, people would buy it and complain it didn't work.

5) OS4 will have to go on embedded devices to pay it's way - PPC is a far more common platform for these than x86.

6) OS has to be released for Classic Amiga PPC at some point.  Hyperion can only concentrate on one architecture at the moment, so PPC it is.

7) OS4 is now based on a fairly portable HAL (ExecSG is all C).  If the custom hardware was to stop being avaliable, we wouldn't be too tied in as it could be ported across to another platform.

[x56h34]

Ideally, AmigaOS could become open source and it could run on whatever you port it to. :-)

I wonder if that would stimulate upgrades of the classic 68k OS3.9 as well? Now that would be extra cool.

[Varthall]

x56h34 wrote:
Ideally, AmigaOS could become open source and it could run on whatever you port it to. :-)

I wonder if that would stimulate upgrades of the classic 68k OS3.9 as well? Now that would be extra cool.

I don't see much point in having AmigaOs4 ported on x86, since there's already Aros for that, and better yet it's free.

Varthall

[gafstu]

Would it not be sensible to base the OS design to work on a selected range of x86 mobos, and cards etc, giving the benefit of cheap/common x86 architecture whilst still having a fixed hardware platform?

Sounds like they have to sort something which is comercially viable soon to keep this project going, as i think everyone will admit - the amiga enthusiasts are the only people who are going to buy any new amiga systems for the forseable future?

gafstu wrote:
Would it not be sensible to base the OS design to work on a selected range of x86 mobos, and cards etc, giving the benefit of cheap/common x86 architecture whilst still having a fixed hardware platform?

Sounds like they have to sort something which is comercially viable soon to keep this project going, as i think everyone will admit - the amiga enthusiasts are the only people who are going to buy any new amiga systems for the forseable future?

Look how we can talk about this subject on here without fear of some banana waving imbecile with a bad grasp of the English language calling us names.

@Wayne
Thanks. :-)

[billt]

>Why does the architecture for OS4.0 have to be custom? This
>makes it expensive and possibly less reliable than mass
>produced alternatives. Why not build the Amiga OS to run on a
>IBM PC style architecture, then all the hardware availability
>issues will go away??


Why has Apple gone Intel, but will still only run on "custom" x86 hardware and not commodity PC hardware??

http://it.slashdot.org/article.pl?sid=05/06/14/1442212&tid=93&tid=3


There was a marketing reason given some time ago, that I feel is believable.

I've noticed my own computer usage slowly shifting to more PC than A4000T things over the last couple years. If it was all the same computer, it'd be even more "convenient" to stay in Windows for more things rather than keep rebooting, as if AOS could run on my commodity PC, why whould I have two commodity PCs under my desk via KVM??? I wouldn't.

I'd dual-boot. As most stuff I'm doing these days is easier with Firefox or Thunderbird than IBrowse or YAM, and I'm getting into apps not present on Amiga today, such a dual-boot machine would most of the time be in Windows ready for use. I'd reboot to run AOS, then reboot back to Windows. The marketing people among us would fear that over time, inconvenience of dual-booting would erode away the AOS usage more and more, until the user really doesn't get into AOS very often at all, and thus not want to pay for AOS apps and games.

It would also be too easy to be impatient waiting for AOS ports and just buy Windows versions of things, and then why bother buying the AOS port later on when you already have the same program for Windows, have already beaten and gotten bored of the game, etc?

Those who can decide such things have made their decision considering possibilities that would affect their sales, even in this tiny market. You don't have to like the outcome or the reasons behind such decisions, either accept them and live with it, or invent something yourself that you think is better and present your proposal to the businessmen in a professional way to encourage them profit can be had your way.

Either way, there may be opportunities hiding around here someplace.

[Paul_Gadd]

The last remaining companies cannot see the forest for the trees. Something good could be put right in front of their faces on a silver plate but they would ignore that and find a way of making a quick buck.

Every single company after Commodore has learned nothing.

[Rooster]

See, my solution would be this:

A custom card, be it PCI-eXpress, or standard PCI add-on card.  This card would among other things, be fitted with a CF (or similar, CF fan here) adapter and card, which would hold OS files.

Several possibilities exist, so let me share some:

-OS BACK on a chip. Be it a CF (easily updatable, and huge in terms of capacity) or other flash technology.  This is one of the hottest things the Amiga had going for it, IMO.  And if a network card can steal boot priority on a x86 system, so could "The Amiga" card.  

-Backward compatibility.  OK, not likly to be popular, but one of many adapters that could hook up to the add-on card:  serial drive adapter for all us external DD drive owners.

-Controlled envirnment:  Let's face it - one reason the mac platform is currently a good choice is becuase of the physical differnce in number of 3rd party vendors.  For a company to support/create drivers for all the possible COMMON cards you might have plugged into your system, could bankrupt a company, or lead to non-stop technical support calls.  However, via the "core" of this Amiga Card, they could again attain control over the x86 platform, should they choose.  A/V COULD be onboard.  There could be headers for other cards to plug into PCI slots(mainly for power) but physically communicate with the Amiga Card either through the system bus, or via a direct connection.  

Doing this would place the Amiga/AmigaOS in the realm of the common people with x86 systems, while ensuring that it isn't simply a "Linux-like OS", that you dual-boot to instead of Windows when you "need" to.  You would have a "true" Amiga system.  

[Rooster]

See, my solution would be this:

A custom card, be it PCI-eXpress, or standard PCI add-on card.  This card would among other things, be fitted with a CF (or similar, CF fan here) adapter and card, which would hold OS files.

Several possibilities exist, so let me share some:

-OS BACK on a chip. Be it a CF (easily updatable, and huge in terms of capacity) or other flash technology.  This is one of the hottest things the Amiga had going for it, IMO.  And if a network card can steal boot priority on a x86 system, so could "The Amiga" card.  

-Backward compatibility.  OK, not likly to be popular, but one of many adapters that could hook up to the add-on card:  serial drive adapter for all us external DD drive owners.

-Controlled envirnment:  Let's face it - one reason the mac platform is currently a good choice is becuase of the physical differnce in number of 3rd party vendors.  For a company to support/create drivers for all the possible COMMON cards you might have plugged into your system, could bankrupt a company, or lead to non-stop technical support calls.  However, via the "core" of this Amiga Card, they could again attain control over the x86 platform, should they choose.  A/V COULD be onboard.  There could be headers for other cards to plug into PCI slots(mainly for power) but physically communicate with the Amiga Card either through the system bus, or via a direct connection.  

Doing this would place the Amiga/AmigaOS in the realm of the common people with x86 systems, while ensuring that it isn't simply a "Linux-like OS", that you dual-boot to instead of Windows when you "need" to.  You would have a "true" Amiga system.  

[Doppie1200]

Who needs a true amiga system these days?
Something like that would only appeal to a small number of people namely; amiga fanatics/hobbyists.

Simple rule IMO; an economically viable product is a product that can offer something the competitors can't offer.

That may be in terms of price/quality balance and or functionality added to that.

For the computing masses these days Windows XP on a x86 is practically unbeatable (perhaps in price is where somthing can be achieved). For people wanting an alternative there is the MAC, or Linux for those who like that. But amiga in the realm of 'common people' with X86 systems? Not for a price you will. For free people might take a look (WINUAE) but it isn't worht the 'common people' a dime.

If you are really determined to get amiga back into everybodies homes you need to look into new markets. Markets where the product can offer something others don't.

[Noster]

Hi,

@Doppie1200
> Who needs a true amiga system these days?

Who needs a computer at home? It's nice to have and to play/work with it. And to work with a real Amiga is somehow more fun/uncomplicated than using Win#*?&%$§" XP.
I've never bought a computer in the last 20 years, because it was usefull in any way. I always bought computers, because the machines and their usage/programming is so interesting. The Amiga is much easier to program than a Win#*%&§$ machine. I could do anything on an Amiga, no restrictions by the OS.
Ok, call me a fanatic but I still think the Amiga is the better system for home-use.

When I have the money, I will buy an A1 to work and play with it.

My only problem is the browser, more and more Internet sides are not readable with the Amiga any more, even eBay has changed their sides in the last week, so the sides are not displayed correct any more :-( Why do everything has to be changed in the IT-world, even the already perfect running things?

@gafstu
> Why not build the Amiga OS to run on a IBM PC style architecture?

Thought about big endian <-> little endian?

Noster