Some thoughts on why the Amiga didn't thrive

[Frogs]

Originally Posted by robinsonb5

The thing to remember with the iPhone, though, is that Apple already had a very strong foothold in the MP3 player market with the iPod - so when the two classes of device merged into what we now recognise as a smartphone Apple were uniquely well positioned to take over that market.

If the Amiga had taken over, we would have likely said the same thing about Commodore. In 1985, the C-64 was the world's best selling computer.

Someone made a really good point earlier that what people didn't know at the time was that the market was starting to split between specialized gaming devices (consoles) and computers for work.

Whether specialization for electronics was inevitable or not I don't know. What we do know is that when the Amiga first launched, Commodore was trying to straddle both worlds with the Amiga being the computer that could play games and do business work.

Unfortunately, the *hardware* wasn't well suited for supporting business work as well as its competitors and thus it was the competitors that got the software.

As persuasive as I find Bruce's arguments, I don't subscribe to the inevitability of the PCs becoming so dominant if there had been alternatives that could do what PCs did but better. Otherwise, we'd be using Blackberries or Windows CE phones today.

Commodore in 1985 had the largest market share of personal computers. (the C-64) Bigger than IBM. Bigger than Apple. It was in a position to capitalize on that with the Amiga if the Amiga had been a superset of what the PCs and Mac could do. But it wasn't. Each made trade-offs with their hardware.

To use a D&D analogy, the Amiga was a much better overall character. Its stats, overall, were much better than the PCs and Mac. But they weren't better in all areas and the areas where the PC and Macs concentrated on turned out to be the areas that mattered for businesses (namely, on-screen clarity via resolution and/or via monochrome).

[grond]

I find it strange that the Amiga was originally intended to become a console and then was turned into a computer because of the US American console crisis but for some reason the popularity of consoles for games in the US is supposed to be the reason for the Amiga's lower popularity in the USA as compared to Europe. Either the situation around consoles changed a lot in a relatively short time or this doesn't seem to make much sense.

Perhaps Commodore should have turned the A500 back into a console and only made big box Amigas that could also play the games for that console. Console sales could have driven Amiga technology for the big box Amigas which would have given nice margins on top of the console business.

[Bruce Abbott]

Originally Posted by dreadnought

Debunking myths which plague your posts could be a full time job, so I'll just make a guest appearance now and then

You debunked nothing.

Lotus 123 was released in January 1983. By that time the IBM PC had sold well over 1 million units, and already had Visicalc (from the Apple-II) and Multiplan (precursor to Microsoft Excel).

So it wasn't so much people buying PCs to use Lotus 123, as Lotus producing an app for the PC. They did so because the existing PC market was large enough to justify developing a better spreadsheet for it. The idea that Lotus Corp would go to all the trouble of porting their x86 assembly language code to an entirely different platform with no hope of achieving the sales required to justify it is laughable.

Of course, eventually PC compatibility became a must, but in the beginning, not so much.

The IBM PC became a runaway success right from the start, and anyone who had one needed software for it. Obviously such software needed to be 'IBM compatible' and the corollary was obviously true too. With the PC greatly outselling other desktop computers it became a magnet for software development.

IBM Personal Computer

the IBM keyboard provided good ergonomics, reliable and positive tactile key mechanisms and flip-up feet to adjust its angle.

Public reception of the keyboard was extremely positive, with some sources describing it as a major selling point of the PC and even as "the best keyboard available on any microcomputer."...

Third-party software support grew extremely quickly, and within a year the PC platform was supplied with a vast array of titles for any conceivable purpose...

Reception of the IBM PC was extremely positive. Even before its release reviewers were impressed by the advertised specifications of the machine, and upon its release reviews praised virtually every aspect of its design both in comparison to contemporary machines and with regards to new and unexpected features.

Praise was directed at the build quality of the PC, in particular its keyboard, IBM's decision to use open specifications to encourage third party software and hardware development, their speed at delivering documentation and the quality therein, the quality of the video display, and the use of commodity components from established suppliers in the electronics industry. The price was considered extremely competitive compared to the value per dollar of competing machines.

Two years after its release, BYTE Magazine retrospectively concluded that the PC had succeeded both because of its features - an 80-column screen, open architecture, and high-quality keyboard - and the failure of other computer manufacturers to achieve these features first:

"In retrospect, it seems IBM stepped into a void that remained, paradoxically, at the center of a crowded market"

Yes, one of the major reasons for the PC's popularity was... its keyboard! Or to be more precise, journalists' praise for it. They didn't care what it was attached to, just that they all loved their IBM typewriters so the PC must be even better!

So what the Amiga needed wasn't a killer app, but a killer keyboard. And a time machine set 4 years in the past.

[defor]

Originally Posted by grond

I find it strange that the Amiga was originally intended to become a console and then was turned into a computer because of the US American console crisis but for some reason the popularity of consoles for games in the US is supposed to be the reason for the Amiga's lower popularity in the USA as compared to Europe. Either the situation around consoles changed a lot in a relatively short time or this doesn't seem to make much sense.

Tons have been written on this topic. Just a few (quickly found) articles. Type "how Nintendo saved video game industry" to your favourite web search engine yourself to get more, much more.
This, this, this and this (interesting book, highly recommended).
It's a very well known story. Meanwhile Nintendo, at that time, was not that big in Europe. I don't know why. Marketing strategy? Strong focus on US market? Popularity of 8bit home computers in Europe?

Originally Posted by grond

Perhaps Commodore should have turned the A500 back into a console and only made big box Amigas that could also play the games for that console. Console sales could have driven Amiga technology for the big box Amigas which would have given nice margins on top of the console business.

CDTV, CD32?
Too little, too late.
My personal opinion is that Commodore wouldn't have a chance with a dedicated gaming machine anyway. Why? Because Commodore was always weak in a software support. And video game consoles are solely about the software - the entertainment software. So many manufacturers tried to repeat a Nintendo's way to success. So many of them failed miserably.

P.S.: Idea of a big box Amiga compatible with a "console" Amiga is IMHO unrealistic because the video game economy is based on a closed platform principle (a walled garden) in which publishers must pay "an entrance fee". It also allows to have some anti-piracy protections. You can't have both at the same time: A closed system in a form of video console, and an open system in a from of home computer.

[TEG]

Originally Posted by Bruce Abbott

Lotus 123 was released in January 1983. By that time the IBM PC had sold well over 1 million units, and already had Visicalc (from the Apple-II) and Multiplan (precursor to Microsoft Excel).

So it wasn't so much people buying PCs to use Lotus 123, as Lotus producing an app for the PC. They did so because the existing PC market was large enough to justify developing a better spreadsheet for it. The idea that Lotus Corp would go to all the trouble of porting their x86 assembly language code to an entirely different platform with no hope of achieving the sales required to justify it is laughable.

And so? We are talking about buying a licence. You can rewrite it from scratch. With the Amiga graphical API, I don't think it would have been a big deal. However it's true that you had to wrote it in assembly language I guess.

These editions of 1-2-3 for DOS were primarily written in x86 assembly language.

Release 1 was the first release for DOS-based PCs. Introduced in January 1983.[34]
Release 1A in April 1983[34][35] Officially supported ASCII, unofficially supported the IBM extended character set (but not LICS).[36][37]
Release 2 brought add-in support, better memory management and expanded memory support, supported x87 math coprocessors, and introduced support for the Lotus International Character Set (LICS).[37][36][38][39] Introduced in September 1985.[34][35][40] The Japanese version Lotus 1-2-3 Release 2J for NEC PC-98 computers was released on 1986-09-05.[41][42]
Release 2.01 in July 1986.[34] Introduced an option to switch between LICS and the IBM extended character set.[37][38]
source

[ Show youtube player ]

It lead to another question: when Lattice C was release on the Amiga? If someone have the answer... The Wikipedia page only mention the date of the original release for the PC (1982).

[robinsonb5]

Originally Posted by Bruce Abbott

IBM Personal ComputerYes, one of the major reasons for the PC's popularity was... its keyboard! Or to be more precise, journalists' praise for it. They didn't care what it was attached to, just that they all loved their IBM typewriters so the PC must be even better!

Another example of recognition from a tangential market being a significant driver of adoption.

(To be fair, though, those keyboards were excellent, and still sought after today.)

[Bruce Abbott]

Originally Posted by Frogs

Commodore in 1985 had the largest market share of personal computers. (the C-64) Bigger than IBM. Bigger than Apple. It was in a position to capitalize on that with the Amiga if the Amiga had been a superset of what the PCs and Mac could do. But it wasn't. Each made trade-offs with their hardware.

You forgot about another important factor - price. The C64 was a success partly due to better gaming hardware, but mostly because it undercut the competition on price. To do that some hardware compromises had to be made.

The Amiga was a superset of what home computers had, and the price reflected it. This was a major stumbling block for sales. Had they added even more hardware to make it a superset of business computers as well the price would have been prohibitive. Why would any business pay extra to get a lot of stuff they didn't need? This is where IBM's genius came in. Their 'business' package had a 10MB hard drive, a monochrome text adapter with printer port, a business-quality printer and their famous keyboard - ie. just what a business user wanted and no more.

So what should Commodore have done to match that? Firstly they should have made the Amiga's custom chipset an optional 'multimedia' plugin, so people didn't have to pay for it if they didn't need it. But everyone needs high resolution text so they should have integrated MDA/CGA/EGA compatible video onto the motherboard, along with industry standard serial and parallel ports (not those backward A1000 connectors - what were they thinking?), floppy disk controller and DMA channels. The keyboard port should have allowed you to plug in the famous IBM keyboard that everyone wanted. They should have kept the hardware mouse port, but dropped the joystick port as this could be provided on the multimedia card.

The machine should have had an A2000 form factor, but without the extra slots (by which I mean the Zorro-II and CPU slots), and it should have taken PC cards directly so you could install new video cards etc. when they became available. They should have used an Intel CPU to make porting Lotus 123 practicable, and separated the OS into a BIOS and DOS portion so users could load alternative OS's off disk. Oh look, they did!

[Signman]

I just ordered a Unicomp Model M keyboard.

[Bruce Abbott]

Originally Posted by TEG

And so? We are talking about buying a licence. You can rewrite it from scratch. With the Amiga graphical API, I don't think it would have been a big deal.

Just buy a license, then rewrite it from scratch on a new and unfamiliar platform with totally different hardware? Piece of cake!

But aren't we forgetting something? If this is to be the 'killer app' that people will buy an Amiga for, how do we convince them not to just buy a PC and run the DOS version? I know - since we already paid more for the license and development than we will ever get back anyway, we'll just pay them to use the Amiga version! There's always a way to make sales when profit isn't the goal.

It lead to another question: when Lattice C was release on the Amiga?

Lattice C V3.03 for the Amiga appears to have been released in December 1985 (that's the file dates on the disk, disk label says copyright 1985).

[TEG]

Originally Posted by Bruce Abbott

Just buy a license, then rewrite it from scratch on a new and unfamiliar platform with totally different hardware? Piece of cake!

When you see the simplicity of the 1985 DOS version, I don't think it would had been a big deal to do it on the Amiga. You had Intuition.
On the contrary, on the PC you had to be a genius for sure. You had to address the memory directly.

But it make me think of another catch we did not broach: the incompatible floppy disk format. CrossDOS appeared only in 1989. And at the time it would have been a real problem to infiltrate the professional world.

Originally Posted by Bruce Abbott

But aren't we forgetting something? If this is to be the 'killer app' that people will buy an Amiga for, how do we convince them not to just buy a PC and run the DOS version? I know - since we already paid more for the license and development than we will ever get back anyway, we'll just pay them to use the Amiga version! There's always a way to make sales when profit isn't the goal.

No, you paid because you can do what others can do and more. For example: having several spreadsheets on your screen or even the words-processor simultaneously and displaying colourful images.

Originally Posted by Bruce Abbott

Lattice C V3.03 for the Amiga appears to have been released in December 1985 (that's the file dates on the disk, disk label says copyright 1985).

Thanks. So if Lattice C was really released in 1985, it mean Lotus could had been rewrite in C. Still less a big deal to make a clone if you had the cooperation of the original team.

Now I wonder if Commodore made a deal with Lattice to have the C ported or if Lattice did it by itself. Having a compiler for the platform was just crucial.

[TEG]

Originally Posted by Bruce Abbott

So what the Amiga needed wasn't a killer app, but a killer keyboard. And a time machine set 4 years in the past.

What was the A1000 keyboard quality? Commodore had a rather good reputation about keyboards since the start. PET, C64 and C128 have good keyboard.

[robinsonb5]

Originally Posted by TEG

What was the A1000 keyboard quality? Commodore had a rather good reputation about keyboards since the start. PET, C64 and C128 have good keyboard.

https://deskthority.net/wiki/Commodore_Amiga_1000


If that's anything to go by, it was very good - but once again lacked the brand recognition factor.

[Bruce Abbott]

Originally Posted by robinsonb5

If that's anything to go by, it was very good - but once again lacked the brand recognition factor.

I thought the A1000's keyboard was pretty good, but in hindsight the cursor keys were a bit too narrow and the 'cross' configuration not as ergonomic as the 'inverted T' that later Amiga keyboards used.

But of course it wasn't IBM and therefore an unknown quantity. Considering the pedigree (the C46 and Atari ST both had terrible keyboards), I can understand why potential buyers would shy away from it.

As for the IBM keyboard, I had one for a while and while it was nice I prefer Amiga keyboards. The IBM's keys were too 'clicky' for me. Since few other keyboards duplicated it I suspect most users actually prefer less noisy keyboards.

[robinsonb5]

Originally Posted by Bruce Abbott

As for the IBM keyboard, I had one for a while and while it was nice I prefer Amiga keyboards. The IBM's keys were too 'clicky' for me. Since few other keyboards duplicated it I suspect most users actually prefer less noisy keyboards.

Maybe - but I suspect mass-producing keyboards more cheaply was the driving force that led ultimately to the $3 rubber-dome-and-membrane $3 keyboards that have come with off-the-shelf computers for the last couple of decades.

(The infamous Model M is technically a membrane keyboard, mind you, despite the buckling spring clickiness.)

Keyboard preference is, of course, another "Holy War" kind of topic!

[Foebane]

Americans called the Amiga a "console".

My own father (who got it for me for my birthday) called my A1200 a "toy".

Sums it up, really.

[redblade]

Maybe the development kit for the Amiga was too pricey and not enough assembler examples. If commodore had supplied an intuition builder to speed up gui development maybe that would of helped? The new interface builders today (based one the origional by NeXT) really does help speed up development.

[Bruce Abbott]

Originally Posted by TEG

Thanks. So if Lattice C was really released in 1985, it mean Lotus could had been rewrite in C.

Irrelevant. Professional developers used cross-compilers, which were available well before 1985 (that's how Amiga OS was developed). But since Lotus 123 wasn't written in C, and both the hardware and OS were very different, porting it would be rather difficult even with the aid of a C compiler.

Still less a big deal to make a clone if you had the cooperation of the original team.

A license with no usable source code is not much use, and the 'original team' wouldn't be much help for figuring out how to produce an Amiga version from scratch.

What was a big deal was getting it out as soon as possible. Commodore figured that having Lotus 123 on launch would be a good selling point, and they had a way to do it without having to spend an unacceptable amount of time and money on porting it - software emulation! They contracted Simile Research to create the Amiga Transformer which could run many IBM titles including Lotus 123 and Wordstar (slowly of course), and bundled it with the A1020 5.25" floppy drive for compatibility with PC disks. User response was (understandably) meh.

The November 1985 issue of Amiga World magazine has a contemporary take on it:-

Amiga's Trump Card:
IBM PC Emulation

By C. Graham

Although the Amiga itself was clearly the star performer at the July 23 launch at the Lincoln Center in New York, Commodore's software emulation of an IBM PC drew its own gasps, skeptical headshaking and rounds of applause, as a Lotus 1-2-3 spread-sheet unfolded across the Amiga screen projected on center stage.

IBM's continuing consolidation of the PC and PC-DOS as defacto standards for mainstream microcomputer applications forces every new computer on the market, no matter how remarkable in terms of price and performance, to confront the issue of PC compatibility. Indeed, one of Macintosh's biggest drawbacks in business environments has been its lack of a bridge to the PC/MS-DOS world, a problem exacerbated by the Mac's closed architecture.

The Amiga, as most reviewers now agree, is a next generation micro, offering performance (speed, multitasking) and impressive features (color graphics, stereo sound and genlock capability) that will open up new markets. Despite the Amiga's advantages, however. Commodore needed to address the IBM compatibility issue for several reasons. These included the need for a mass of software available on launch to give the Amiga strong initial sales momentum and the desire to shake off Commodore's "home" computer image in the U.S. by assuring that Amiga owners have access to the best library of serious business software on the market.

That last point is debatable. The 8088 CPU used in the PC was designed on the same principle of having access to a 'vast library' (of CP/M software), but this did work out in practice. Early ports to the PC quickly gave way to new apps that played to its strengths.

The emulation idea is intriguing, but I had no desire to run Lotus 123 on my A1000 - and I suspect most Amiga owners at the time felt the same. If we wanted to run PC apps we would have bought a PC!

Commodore obviously needed to address the IBM compatibility issue, but attempting to ride on its coattails was bound to fail. They should have aimed towards producing unique software designed for the Amiga that played to its strengths, not just create clones of PC apps. But this was difficult when they didn't have machines for developers to work with. So they had to release the A1000 even though it wasn't finished - then got criticized for not having sufficient titles on launch.

Now I wonder if Commodore made a deal with Lattice to have the C ported or if Lattice did it by itself. Having a compiler for the platform was just crucial.

Good question. Commodore did rebrand it as 'Amiga C' at about the same time, but they used the Greenhills compiler on Sun workstations to produce Amiga OS.

[Bruce Abbott]

Originally Posted by redblade

Maybe the development kit for the Amiga was too pricey and not enough assembler examples.

It would have been pricey for sure, just like development kits for other platforms were. This was not a problem for commercial developers.

Assembler examples were not needed for apps. Intuition itself was written in C, so C was the obvious language for developers to use. Also (unlike earlier 8 bit machines) C on the 68000 was actually quite efficient.

If commodore had supplied an intuition builder to speed up gui development maybe that would of helped?

Perhaps, but building GUIs in earlier versions of Amiga OS was not that hard (fixed resolutions, no locale, low expectations). Developers usually created in-house tools for such things anyway so it wasn't a big deal.

The new interface builders today (based one the origional by NeXT) really does help speed up development.

Modern OSs are far more complex and difficult to write low level code for. However I did program a Win32 app using the same techniques as I do on the Amiga (macros, cut&paste, type in all the coordinates!), and it wasn't that hard once I figured out how Windows did it.

[Foebane]

@ Bruce Abbott:

Is that you in your avatar? You remind me of a fisherman at best, or a gnome or dwarf at worst.

[Bruce Abbott]

Originally Posted by Foebane

@ Bruce Abbott:

Is that you in your avatar? You remind me of a fisherman at best, or a gnome or dwarf at worst.

Or a ... troll?

Yes, it's me. Here's the full picture for context.