Was anyone else disappointed with the A1200?

[oscar_ates]

Originally Posted by grond

Wasn't that the reason why they bent away one pin of the connector? When I exchanged the harddisk in my A1200 for a CF card, I found that the people at the shop where I bought all my upgrades had done that and was a little shocked.

I am not sure but it could be this reason

[TCD]

Stumbled over this German price list from 1993 today on a1k.org. Might explain why some people didn't get an A1200 and rather switched to PC:

[oscar_ates]

There was an AA+ design on paper. Lew Eg. admitted the shortcomings of the A1200 and they wanted to change it. HD floppy support, chunky modes, 2mb fast ram, 16bit audio, etc.
If they could release this machine in 1993 with a 68030 33mhz it could be a game changer. Amiga 1200 itself was too little without aforementioned improvements.

https://en.wikipedia.org/wiki/Commod...et?wprov=sfla1

[Gorf]

Originally Posted by TCD

Stumbled over this German price list from 1993 today on a1k.org. Might explain why some people didn't get an A1200 and rather switched to

This price list from Commodore given out to dealers in Germany might also be a hint:

http://www.computerpreisarchiv.de/CC...omo-12_92.html

They did not even offer the smaller Amigas here because even Commodore did not consider them "professional". But they did try to sell the A3000 here … but also at not very competitive prices.

Interesting here:

 386SX-25  D 386SX/25  4   80  3  --       A     1.375  Ä

A 386SX @25MHz with 4 MB RAM and 80 MB hard disk for just 1375 DM…
compare that to a A1200HD with 030 card and some FastRAM … suddenly the A1200 is not so cheap anymore.

(And this is also the proof that it is not really more expensive to put that board in a bigger box …)

Also all the other offers for PCs from 1500-2000 DM - at that range Commodore was totally missing out with a competitive Amiga model
(no the A2000 was really not competitive any more)

[Karlos]

Why are you comparing the 386SX with the 68030? Every 68030 card for the A1200, whether EC or not at least runs a 32-bit wide bus and 32-bit address bus. By comparison, the 386SX had 16 bit data bus and a 24 bit address bus. I'd be surprised if it was only on par with the 68EC020 at the same clockspeed.

[Gorf]

Originally Posted by Karlos

Why are you comparing the 386SX with the 68030?

Because even with its 16bit data bus a 25MHz "AMD 386SX/SXL-25" is significantly faster than a 14MHz 020.

This Commodore PC did also include a "Cyrix FasMath CX83S87-25-JP" FPU

So comparing it with an 030 upgrade is more than fair!

[oscar_ates]

Well CD32 was 32 bits on paper. I have not seen any better than 16 bit games on it. So bit count can be tricky

[Karlos]

Originally Posted by Gorf

Because even with its 16bit data bus a 25MHz "AMD 386SX/SXL-25" is significantly faster than a 14MHz 020.

This Commodore PC did also include a "Cyrix FasMath CX83S87-25-JP" FPU

So comparing it with an 030 upgrade is more than fair!

I used to use the exact 16-bit 386sx you are describing at university and recall it's real world performance pretty well even if it was decades ago. At 25MHz it performed faster than the 14 MHz 020 in a stock A1200 on some compute bound tasks. Running some benchmark code compiled for both. The gap versus the A1200 with fast ram was much less and in some cases the 020 did better. Scaling for clockspeed I'd say the 020 is a match. A fact I was able to confirm years later with a 28MHz Blizz 1220.

The 25MHz 386sx is not a performance match for any 32-bit 68030 at the same, or higher clockspeed that you could obtain on an accelerator card those days.

[itsmedoofer]

Jeez, either you had used something better and were disappointed or it was a step up. What if, maybes or perhaps are never going to be. We have the machines that were left to us,there are never going to be any more, chill and enjoy what we actually have rather than pontificating about what could have been .

[Gorf]

Originally Posted by Karlos

The 25MHz 386sx is not a performance match for any 32-bit 68030 at the same, or higher clockspeed that you could obtain on an accelerator card those days.

Depending on FPU and MMU usage.
And I did not specify any clockspeed for the 030 upgrade. An EC68020 is just not cutting it ...

You can dive down into the specifics all you want:
the point I made with this comparison should be quite clear to everyone.

Commodore was trying to persuade the PC customers by all kinds of configurations and models throughout a wide price range with their PCs, while treating the Amiga-line like an unwanted child.

[Gorf]

Originally Posted by itsmedoofer

Jeez, either you had used something better and were disappointed or it was a step up. What if, maybes or perhaps are never going to be. We have the machines that were left to us,there are never going to be any more, chill and enjoy what we actually have rather than pontificating about what could have been .

Well - that is probably not what this thread is about - like it or not.

Sure: the past is the past. But that does not make me really satisfied with the present, does it?
Contemporary computing has become quite boring - and the demise of our favorite platform is part of the dilemma.
And yes: one can still enjoy retro-computing and live in the past..
.
I in contrast try to imagine what went wrong, what could have been and what all this would mean for today.
I know - that's a stretch.

[Karlos]

Well contemporary computing may be boring, but it did let me write a 64-bit virtual machine that runs code written in 68K style assembly language that can hit up to 1000 MIPs while running interpreted on a mobile i7 that's maybe 4 or 5 years old now. I didn't even start work on a JIT for it.

[Gorf]

Originally Posted by Karlos

Well contemporary computing may be boring, but it did let me write a 64-bit virtual machine that runs code written in 68K style assembly language that can hit up to 1000 MIPs while running interpreted on a mobile i7 that's maybe 4 or 5 years old now. I didn't even start work on a JIT for it.

Did you now...
Sorry for my skepticism, but that sounds a little bit too good to be true.
Ok: you say "up to" - that leaves room for interpretation.
So what is your average result? How many cycles per 68k instruction?

Edit:
"68K style assembly language" also leaves a lot of room for interpretation ...

[Cyprian]

Originally Posted by Karlos

Why are you comparing the 386SX with the 68030? Every 68030 card for the A1200, whether EC or not at least runs a 32-bit wide bus and 32-bit address bus. By comparison, the 386SX had 16 bit data bus and a 24 bit address bus. I'd be surprised if it was only on par with the 68EC020 at the same clockspeed.

I see that in case of signed div/mul 386 is a bit faster than 030:


Signed Multiplication 32bit
386 38 cycles
030 44 cycles


Signed Division 32bit
386 43 cycles
030 90 cycles


https://www2.math.uni-wuppertal.de/~.../opcode_i.html

https://oldwww.nvg.ntnu.no/amiga/MC6...s/68030it.HTML

[Karlos]

Yes, but what percentage of normal application code is signed integer division? Answer, not much. Most code is move, followed by basic addition and logic, followed by multiplication and division way down at the bottom. Even conditional branches are more common.

In case it wasn't obvious, my point was I actually don't think it's fair on the 386SX to be compared to the 68030 on an accelerator card with local 32-bit fast RAM, especially if it's zero wait stare.

Flipside, it also doesn't help the case that the A1200 suddenly wasn't that much cheaper when putting an 030 card inside, given that the disparity between the 386SX and the 68030. If anything it makes the 386 seem overpriced.

[Karlos]

Originally Posted by Gorf

Did you now...
Sorry for my skepticism, but that sounds a little bit too good to be true.
Ok: you say "up to" - that leaves room for interpretation.
So what is your average result? How many cycles per 68k instruction?

Edit:
"68K style assembly language" also leaves a lot of room for interpretation ...

You can try it out for yourself, it's on GitHub here: https://github.com/IntuitionAmiga/MC64000

Myself and IntuitionAmiga started it out of boredom. He created the repository to throw down the gauntlet after we'd been discussing the idea for a while. I've been tinkering building it in rare pockets of spare time ever since.

Here are some very trivial examples of "coding just for fun" https://youtube.com/playlist?list=PL...3wHOUo3BuhNly5

[Bruce Abbott]

Originally Posted by paul1981

The A3020 (1992) came with an on-board 2.5" IDE interface.

Aha!

But the A3010 (also 1992) didn't, and even though it only had 1MB RAM it cost £100 more than an A1200.

Neither the A3010 nor the A3020 had enough space for a 3.5" hard drive. But we'll give Acorn a pass because they weren't Commodore.

[Karlos]

The whole thread has turned into bashing as much early/mid 90s era tech as possible.

Bloody minidiscs, eh?

[grelbfarlk]

The A1200 should have had a tape deck. Since it didn't have a 5 1/4" bay to install a CD Drive, we could at least use a tape adapter to play CDs from our Discmans.

[Bruce Abbott]

Originally Posted by grond

That's like saying "I can't shake your hand because my left hand is holding my right hand". Why didn't they make the case a little bit bigger? Why didn't they orientate both the FDD and the HDD bay horizontally?

The HDD bay had to be horizontal (or vertical) because hard drive manufacturers insisted on it (something about excessive bearing wear when spinning on an angle). Most 3.5" hard drives at that time were not slim, and they were heavy too. I have a Conner 425MB 3.5" drive that weighs 457g (nearly half a kg), while my Toshiba 351MB 2.5" drive only weighs 138g.

Having the floppy drive on an angle is more ergonomic. This was (AFAIK) a unique feature of the A600 and A1200. It also contributed to the 'wedge' shape, with that brilliant step you could stack up floppy disks along. An A1200 without these features wouldn't be an A1200, it would be just another boring 'flat box with sloping front for the keyboard' like all the others.

Yes. Which is why Commodore would have had to spend a dollar more on the power supply if they had sold the A1200 with the capability of taking 3.5" harddisks.

More expensive, bigger, bulkier, heavier, and last but not least - more expensive to ship. But not only that. People complained about the size and weight of the A500 PSU (not me though, I Iove it) so there was an aesthetic reason to make it smaller too.

Commodore saved a dollar, the user had to shell out a couple of hundreds more for the harddisk.

Commodore saved a dollar, the customer saved $10. Might not seem like much, but you would apply the same principle to everything else - and then it would be more expensive than an Archimedes. Result? Poor sales, and disgruntled customers who paid more for an unpopular machine. Commodore made the A1200 cheaper so you could get it cheaper and get more out of it. The more sales they made the more popular it would be, and the better it would be for them and you.

I don't know why you didn't expect this. Commodore always built their 'consumer' models down to a price, and the strategy worked. The VIC-20, C64 and A500 were all top sellers because they were cheap. Which is the way most people liked it. Making the case and power supply bigger would raise the cost, compromise the ergonomics and make it look ugly. For those who didn't need a hard drive that would be a lose-lose situation - to satisfy the whims of a few who wanted a cheap hard drive no matter what else it affected.

And then Commodore found itself in the same trap with the A600/A1200HD models.

Not sure what you mean by this, but the A590 needed a separate power supply because the A500 PSU (despite its size) wasn't deemed powerful enough. Commodore obviously wanted to avoid that with the A600 and A1200. The obvious answer was to use a 2.5" drive.

The other answer would be a separate case with PSU built in, but that would be a totally different machine.

I have never heard of anyone with a 3.5" harddisk in their A1200 that had a problem because of that.

But I bet you heard of people whose hard drives started getting errors and eventually died, right? This was a common problem with 3.5" drives, particularly if they were roughly handled during operation. I remember one day in my shop when I was installing Windows on a new hard drive. Outside they were working on the road using a compactor, and the vibration was going through everything. I didn't think anything of it - until the hard drive crashed.

These days I am generally very careful not to bump the computer while it's running. But my A1200 sits on the coffee table and sometimes gets knocked. 2.5" drives are designed to withstand being handled roughly, and their small size makes them inherently more robust due the lighter mass of the heads etc.

This is not true. There were 3.5" harddisks that fit into the A1200 case but you had to look for them. The floppy disk drive in fact is a 3.5" device, right?

There's no room to put a 3.5" drive in horizontally. So you have to install it on an angle, which voids the warranty. Maybe it will work fine like that for years, but Commodore couldn't do it without voiding the manufacturer's warranty. Just image if they did - you would never hear the end of it from Amiga fans.

Of course as you say, only a few very slim models would fit. I know because I have done it myself. Didn't like to though. I would rather pay a little extra and have a 2.5" drive properly installed.

The case only needed to be a few millimetres bigger to house any 3.5" harddisk instead of only some slimline models.

A bit more than 'a few' I reckon. But so what? Commodore decided they wanted a smaller case, and many of us love what they did. That a 3.5" drive won't fit was not a deal breaker for us. I hated 3.5" IDE hard drives. I have seen so many fail over the years it makes me sick. They were large and heavy and power hungry, and yet delicate and unreliable.

Just having a harddisk in any Amiga makes it feel magnitudes faster in daily usage.

Depending on what your 'daily' usage is. If it's playing pirated game disks then floppies are fine. 'Power' users would want a hard drive for sure, and they got it - either built in or as an upgrade. Many users would start at the low-end and then move up as their needs changed.

Are you sure about these prices/capacities? I paid several hundreds (800?) DM for a 330MB 2.5" harddisk for my A1200 in 1995.

See FAX image below (received using GPFAX on an Amiga 1200, in 1996!)