Welcome, Guest. Please login or register.

Author Topic: Amiga hardware superiority  (Read 10992 times)

Description:

0 Members and 1 Guest are viewing this topic.

Offline Pentad

Re: Amiga hardware superiority
« on: December 01, 2010, 04:18:24 PM »
I think this chart really goes to show had badly the Amiga hardware was becoming by 1994.  Honest, I'm not trying to start a flame war but I'm just being realistic.   I really have fond memories of the Amiga and of Commodore but it just wasn't competitive by the mid 90s.  Some thoughts:

-AGA was just not good enough.  It was late to the party and only extended OCS/ECS for a short time.

-Do you realize that Paula was roughly 10 years old by 1994!?

-8 Bit audio was an embarrassment compared to the plethora of sound cards of the PC world and the Mac's built-in audio.

-3DFX was testing their Voodoo 1 3D cards in 1994 for the PC.  I remember seeing a demo of it on DOS/Windows 95 Beta at CES in Chicago in the summer of 94.  AGA seemed quaint and backwards compared to 3DFX.  

-Windows 95 would be released in August of 1994 and would kill Commodore and Atari and nearly kill Apple a few years later

-Apple's hardware was still stuck with System 7.x but the hardware was getting better and better

In the end, I think you can see the major problem here for Commodore and Amiga:  Technology Product Cycle.  The Amiga started theirs in 1985 and it was cutting edge but it was coming to an end in 1994/1995.  Apple/Microsoft/PC were just starting theirs at this time but the big deal is that it gave them something to grow into.  Apple was moving to PPC, Windows would get better and better (NT 4 was an amazing OS) and this would lead to Windows 2k/XP, and finally the PC world got PCI in '92 and folks like 3DFX were going to bring 3D acceleration to the masses.

What did Commodore have?  AmigaOS was very mature but had no place to go.   It was so tied to the hardware that any small change would kill legacy apps.  Look what a mess it was going from 1.3 to 2.x/3.x.

AAA (or whatever) should have been started in 1988 and pushed hard by R&D so that as one TPC ended (started in 1985) Commodore could have transitioned to a a new TPC for the 90s.

AmigaOS 4 should have been developed along with AAA and included modern features like memory protection.  By the late 80's, everybody in the computer world recognized that memory protection is a must for a stable, modern OS.  

Microsoft developed NT (based on ideas from VAX) with memory protection (among many other things) because its where you had to go.  Consumer Windows was always planned to intersect Windows NT and they did a great job of slowly getting everybody there.  

Apple and Commodore had the same problem with their OS.  You can't just 'bolt on' modern features and a rewrite kills your current apps that keep you in business.  Apple's soap opera like quest for a new OS was certainly amazing and they just got lucky with Steve Jobs and NeXT.

Commodore was...well, Commodore.   What worked in the 1980s for them failed in the 1990s...and here we are.

Cheers!
-P
Linux User (Arch & OpenSUSE TW) - WinUAE via WINE
 

Offline Pentad

Re: Amiga hardware superiority
« Reply #1 on: December 01, 2010, 07:29:25 PM »
Quote from: Franko;596091
I agree with a lot of what you say in your post, but remember it's wasn't the lack of ideas and R&D that brought Commodore to it's knees, it was greed & downright theft by the likes of Mehdi Ali that ended Commodore... :(


My friend, I could not agree more.  :-)
Linux User (Arch & OpenSUSE TW) - WinUAE via WINE
 

Offline Pentad

Re: Amiga hardware superiority
« Reply #2 on: December 01, 2010, 07:50:12 PM »
As somebody else mentioned, a HAL would have helped Commodore/Amiga move the hardware without killing the software.

The Amiga 1000 was designed with a video game mentality where the hardware really doesn't change.  If that is your foundation for development you can take many liberties that give you speed up front but cost you upgradeability down the road.

An AmigaOS that offered a true HAL along with memory protection and good support for virtual memory combined with upgraded hardware could have given the Amiga another ten years to grow.

At the time, I remember all the software was written in assembly and even coded for specific CPUs, like an 020, 030, 040, etc...  You would never do that today as a software engineer.  Granted, compliers are much better at optimizing code and CPUs are better at running code, but you don't want code tied to a specific CPU.

I mentioned in another post that Atari TOS was compiled strictly for the 68000.  You couldn't even run TOS on an 010, 020, 030 because they were using instructions that were not certified by Motorola to be in next generation chips (010 and beyond).  Motorola did not include them and so TOS was stuck with the 68000.

Atari had to rewrite TOS for their TT (030) line which caused compatibility issues.

The Amiga had the same problem to an extent with the OS and hardware.   The OS was so tied to OCS/ECS that even moving to AGA lost apps written under 1.3 for ECS.

Which leads me to this...

I know AAA wasn't very far along in development but I wondered if there were any discussions on how to break the OS from the hardware and move to AAA.  RTG was have been baby steps to a full HAL but the OS would have to have handled older software making calls to hardware that was no longer there.  I wondered what they were going to do...

Does anybody know when Carl Sassenrath left Commodore?   Who took over the AmigaOS Kernel after Carl left?  Was it Bryce Nesbitt?  

Cheers!
-P
« Last Edit: December 01, 2010, 07:52:44 PM by Pentad »
Linux User (Arch & OpenSUSE TW) - WinUAE via WINE
 

Offline Pentad

Re: Amiga hardware superiority
« Reply #3 on: December 02, 2010, 02:00:59 AM »
Crumb,

I'm honestly not trying to give you a hard time but allow me to comment:

Quote from: Crumb;596219
That's an urban legend... DRACO ran AmigaOS3.1 quite fast without the need of custom chips.

I'm not sure what you mean by this statement.   AmigaOS requires a number of custom chips to function properly.

Quote from: Crumb;596219
BTW, 2.x to 3.x software transition was quite smooth. You had professional high quality software like ImageFX, AdPro, Photogenics, Lightwave, Cinema4D, Real3D, Imagine, Caligari Truespace, Bars'n'pipes, Pagestream... you could even paint with TrueBrilliance, and there were professional and affordable video solutions not available for any of the listed systems. Final Copy, Final Writer and Wordworth were excellent packages too.

You are mixing a number of arguments here in this section of text.  While I would agree that the transition from 1.3 to 2.x was decent it was by no means smooth.  If I cared more I could offer you a number of articles on OS 2.x that Amiga World ran at the time but suffice to say it was a painful move for all involved.

Commodore tried to motivate publishers by running ads for OS 2.x with the catch phrase "...and the list keeps growing..." or something to that effect.  OS 2.x was a major shift in the AmigaOS technology and it was growing pains.  The shift was good for Commodore and the Amiga at the time.

As you stated, there were many fine applications that upgraded and worked well with 2.x and beyond.  In the end, some 1.3 applications did not work and their owners decided not to move beyond 1.3.

I think many developers saw the Amiga platform stagnate compared to PC and Mac and decided not to fund the upgrade of their applications.

As for 'affordable video solutions', yes, there were many of those available for the Amiga well before other platforms.

Quote from: Crumb;596219
In 1994 we were happily multitasking and most computer users didn't know what that meant and even claimed it was useless. And our applications were quite professional and most of them more affordable than similar programs in other systems.

I'm not sure you are remembering 1994 correctly or your use of the word 'multitasking' is incorrect here.  These are a few the operating systems/computers that allowed multitasking in 1994:

Windows 3.x
MacOS (First there was Switcher and then MultiFinder)
DOS (many application switchers)
CPM
Atari (Atari TT and Falcon with MultiTOS)
VAX
Unix

With the exception of VAX and Unix, a major program crash would take out the entire computer.  However, people were happily multitasking on their computers.  They were running more than one program at a time, cutting and pasting between them, and cursing when one program crashed and took down the entire system.

I'm not saying that every crash could bring down the system but Atari, Apple, and Commodore did not offer memory protection so its very easy for one program to bring down the whole machine.

If you want to completely trash AmigaOS 1.0 to 3.x, just write to memory location $4.  That's it.  Memory location 4 is the only absolute location in the system.  You destroy that pointer and the AmigaOS is dead.

AmigaOS is a pre-emptive OS where just about everyone else on the list is co-op (excluding Unix and VAX) but we're just splitting hairs.




Quote from: Crumb;596219
really? Win95 crashed much more easily than AmigaOS. And it crawled in hardware way faster too. And needed 8MB to be useable. By the late 80's most of people used monotask-OSes like MacOS or MSDOS+Windows. But most of them didn't have a clue about what multitasking meant and as you can suppose memory protection was an even more strange word for them.

I think MS did a pathethic job with Win95. They should have marketed a NT workstation version as Win95 instead of creating that "thing". OS2 was simply superior and even allowed running Win3.1 apps too. It was not until WinXP that peecee users got a stable Windows system. Until WinXP you could hang Win95/98/ME as easy as AmigaOS3.x. Win95 with 4MB was unusable. Swapping floppy disks in my A500 was a less painful experience and usually more productive.

Let's talk about what you've said here:

First, I am no way a fan of Microsoft but I also try to be fair person.  When you take all things considered, Microsoft did an admirable job with Windows 95 (excluding ME).  You may scuff and mock my post but at least hear me out.

The programmers at Apple and Commodore had it insanely easier compared to the programmers at Microsoft.  Apple and Commodore controlled both the OS and the core hardware.  If you have any idea how hard it is to write a kernel, imagine how hard it would be not knowing what type of core system it will be installed on.

The programmers for Windows had to write an OS that sat on top of DOS, had to work with thousands of different hardware configurations, remain backwards compatible, and unify a driver set (DirectX) for the very first time.

Seriously, that is an amazing set of goals to aim for.  I'm not saying that you have to like Microsoft or that Windows 95 was the greatest OS ever.  I'm saying that for all that it had to do, they did a decent job for their first time out.

To be honest, I used to bash Microsoft and Windows 9x just as much as anybody else.  Then I had to write a kernel for hardware in college and boy does that help you to see the world a bit differently.

I would absolutely agree that the AmigaOS was much more stable than Windows 9x.  However, IT SHOULD HAVE BEEN!  Writing a kernel for a handful of platforms should be a piece of cake compared to the zillion configurations of the PC world.

I guess in the end I feel that if you are going to compare Operating Systems you have to take into account the hardware it has to run on.   Apple and Commodore never had to address the issues that Windows developers had to address.

Quote from: Crumb;596219
AGA in 1994 was not as bad as you may think, it allowed you to watch ham8 pr0n and animations smoothly. They should have improved more the CDXL format to take advantage of 030/040. Amiga was very cost effective solution.

Amiga also sported Autoconfig(tm) and it has worked very well until today.

A3000/4000 16MB limit was not really important until many years later. With 2MB of chipram you could do many things at once while other systems had to spent money in both gfx and normal ram. Even soundcard ram in some cases.

Amigas used to sound much better connected to a 1084s monitor than the old and crappy yogourt-like speakers used by 90% pc users in the 90s.

In 1994 AmigaOS was simply superior

It may not seem like it but I am a huge fan of the Amiga and of Commodore.  I owned most of the best machines by Commodore and wonder 'what if' like most of you.  AutoConfig was brilliant but by 1994 Paula was outdated...

In 1994 AGA was too little too late.  Doom and Wolfenstein 3D were huge and for the first time PC games made the Amiga look dated.
« Last Edit: December 02, 2010, 11:09:16 AM by Pentad »
Linux User (Arch & OpenSUSE TW) - WinUAE via WINE
 

Offline Pentad

Re: Amiga hardware superiority
« Reply #4 on: December 02, 2010, 10:58:52 AM »
Quote from: fishy_fiz;596308
I think what he means is exactly what he said. Draco ran OS3.x without custom chips ( as does amithlon). Not sure how that could be made clearer.

Your selection of examples is very interesting:

Are you suggesting that Amithlon does not emulate the custom hardware the AmigaOS needs to function?  

You might want to research how it (and UAE derivatives) work:

"The emulator, developed by Bernd Meyer, is based upon the authors' experience with the WinUAE JiT emulation, but features some dramatic changes to increase emulation speed (at the loss of compatibility). The slim-line ISOLinux distribution is used to boot directly into the Amiga emulation, removing the need for users to interact with a host operating system. This simple, yet effective change resulted in many users favouring Amithlon over AmigaOS XL as the emulation of choice."

Please read further about it here: http://www.amigahistory.co.uk/emulators/amithlon.html

Draco did use custom chips:

2 CIA chips
1 Kickstart Rom
Paula (according to a post on Usenet that I found)

I would consider these custom chips.
You can read further here:

http://www.amiga-hardware.com/showhardware.cgi?HARDID=43

and here:

http://amiga.resource.cx/mod/draco.html


Cheers!
Linux User (Arch & OpenSUSE TW) - WinUAE via WINE