Things the Amiga didn't get right from Day 1

[sandruzzo]

* 32k of fast mem for 68k
* Sprites should have had their own colors, as the capability to flip them.
* more memory dma slots to have 6 plane with hires.
* asynchronous copper, a better move operetation.
* chip register out of chip mem bus
* blitter with capability to do 6 blit operation without using cpu.
* better filling and line drawing to use less cpu.

[alewis]

Some people need to read up on the history of the Amiga's development

Commodore made no bad decisions regarding the A1000, as C= didn't develop it. The original Lorraine - or "Joyboard" - was funded by a group of doctors who wanted to break into the burgeoning home video game market. They hired Jay Miner et al to develop a 'killer' games console. The design team had the foresight to add keyboard and floppy drive interfaces, but that decision was external to the design brief.

Regarding gfx memory, even in 1985 the BBC Micro's use of 24Kb of its 32kb RAM for gfx memory was considered extravagant. The Amiga could use all 512Kb as gfx memory, albeit that needed a further 512Kb RAM expansion.

NO PC in 1985 came with an on-board hard disk controller. HD controllers came in the form of [expensive] add-in cards, and the user had to choose between a number of competing standards for the controller card, namely MFM and RLL, and a drive suitable for that standard of card. Hindsight is wonderful, but to expect the design team to have added massive cost to a games console, and found the space on the board to add HD controller functionality is unrealistic. Oh, and it would have locked the Amiga into that choice. Oh, and have you seen what an ST-506 (or 225) interface looks like? Cumbersome is such an apt word.

And which interface would they have used? ST-596 was introduced in 1980, ST-412 in 1981, ST-225 in 1982. IBM used WD1003, and WD/CDC intduced what would be termed ATA in 1986.

CPU. The design decision used a processor that belonged to a proven and broadly compatible processor family. The cost of using a 68020, in a games machine, would have added a huge price rise to the base cost. The 16bit 6592 was not wholly backwards compatible with the 6502, and had little guarantee of ensuring backwards compatibility with C64 software; one only needs to see the problems the C64C had with compatibility. Besides, such backwards compatibility limits the forward usefulness of the machine; whilst being compatible allows access to old software, it excludes the machine form competing with newer, faster, rivals.

MIDI. Debatable. Again, designed as a games machine...

Could C= have done more once they acquired Amiga in late 84? Yes, but... at the risk of having a machine that was in perpetual R&D; let's say they to,d the engineers to go back and add more registers, a fast RAM buffer, tweaks to the custom chip sets. Then new silicon is announced, faster CPU, better widget... the engineers want to use these new developments. Time is need to order and obtain samples, rejig manufacturing to incorporate them, test the new 'Amiga' works with them, etc etc. The another new/improved component is released... rinse and repeat. At some point one has to say "enough, we need to release a product to market".

What Commodore management did do wrong was not seek out partnership deals with Lotus for 123, Word Perfect Corp for Word Perfect, and Ashton-Tate for dBase, etc. This was the games machine that ran rings around every other machine in the world: PC was DOS based, no GUI. MAc had a gui, but it was monochrome. And the ST was, well, lame in comparison.

Regards the floppy argument. The HD drive didn't arrive until 1987, so the choose of a 1mb DD floppy was forward looking. The "non-standard" disk format was ahead of the PF disk for at, and OFS was designed to enable the complete recovery of data from a disk; a PC formatted disk was at risk of the directory block was damaged. With OFS, every sector is the directory.

Regarding incremental changes to the A500, to wit 6 channel audio, improved custom chipset, etc. The it would not have been backwards compatible with the A1000, and the software produced, resulting in a fragmented Amiga software market. The A500 was essentially the home version of the A1000, the B2000 the more practical A1000 form factor (think expansion slots).

Imho, incremental processor changes such as the 14Mhz 68000 would have been a bonus; yes, games would break due to timing loop issues. But it would have kept the Amiga relatively competitive until a new architecture was finished; the AAA chipset based Amiga.

Just my tuppence. But let's not forget the Amiga is a closed architecture; too much depended on custom chipsets and the MC68000, changes to any of the chipset or changing the CPU at all broke software.

[amiga_Forever]

Originally Posted by ajk

The 512kB extra RAM is actually one of the few things Commodore did provide (as the A501) from the beginning

I don't think anyone disagrees that Commodore made poor marketing choices, especially later on. They should have pushed upgrades to the systems much more. But as the topic was "from day 1", high-density floppies, hard drives and such simply were not a realistic option for the A1000.

The technology was pretty much all there, just not utilized to the full extent by the market.

I had my Commodore A501 with Real-Time battery backuped up clock..... That was probably the biggest 'WOW' factor at the time., or perhaps the 1MB memory increase.

[Miggy4eva]

1) Commodore should have put midi ports on the Amiga from DAY ONE. It was unacceptable that a superior machine with great sound was beaten in the music market by an inferior machine which had them included. Yes, adding midi ports is very simple and very easy but musical types like everything to be easy out of the box.
2) 68000 should have been at least 8mhz but 10mhz ideal with 32k of fast mem.
3) NON-INTERLACED 640 x 480 4 colour mode at 70hz for productivity, office, midi software usage. Something else an inferior machine managed to offer which helped with midi software, crisp image better for staring at for long periods than normal CRT's of the day.
4) 256 colour mode from palette of 262,144 colours, supported by all hardware sprites, blitter and copper
5) PC-compatible disk format and A500 from 1987 onwards should have introduced 1.44MB HD drive.
6) A1200 should have had a DSP, 030, more colours, and some kind of built in modem to propel Amiga into the internet age. That was a BIG miss by Commodore.
7) EDIT of course! Just one, ONE chunky graphics mode to make Wolf3d and Doom possible

Originally Posted by alewis

What Commodore management did do wrong was not seek out partnership deals with Lotus for 123, Word Perfect Corp for Word Perfect, and Ashton-Tate for dBase, etc. This was the games machine that ran rings around every other machine in the world: PC was DOS based, no GUI. MAc had a gui, but it was monochrome. And the ST was, well, lame in comparison.

Ahahaha at your comparison to other machines. yes it's true, Commodore should have got better partnerships, even with Microsoft for early Office versions. This would have positioned them much better to compete against the PC in the business and office market.

[lordofchaos]

Originally Posted by alewis

Regarding incremental changes to the A500, to wit 6 channel audio, improved custom chipset, etc. The it would not have been backwards compatible with the A1000, and the software produced, resulting in a fragmented Amiga software market. The A500 was essentially the home version of the A1000, the B2000 the more practical A1000 form factor (think expansion slots).

I don't think it would have made much difference in the long term, the Commodore approach was already fragmented from the very start, including their overall approach, Commodore was all about being fragmented, just look at the machines they released over the course.. There wasn't much rhyme or reason to anything they did. Not to take anything away from the actual brilliant people that built the machines.

To the posters original question, what they didn't get right from day 1? A good business plan. It's clear they had something very special on their hands, they just couldn't focus it's vast capabilities. Was it a games machine? Was it a business machine? Was it both? The market wasn't even mature enough at that time to appreciate a machine that did both, and they failed to turn that into an advantage. Probably just talking rubbish, my take on it anyway.

[Juz400]

One thing we seem to be ignoring when we talk about `compatablitiy` between machines is those (in the early days at least) often sneered at things like `graphics.library`
Everyone hit the hardware like they used to on the 8bits, we dont need no stinky graphics library, MY code is better! it often was

Over and over CBM said use the librarys to keep things compatable, hardware will change!

If you look in the CD32 documentation it states USE THE LIBRARYS! as the next chipset will NOT be the same!
It also mentions that the librarys execute FASTER than code based in the chip memory, the memory address is outside the custom chip bus and I guess would execute the same speed as FAST ram (Why no standard FAST ram CBM? )

CBM should have got the lead programmers from the A standard game developers together maybe once a year for a few days/a week to WORK on things like graphics.library
Improve the code, add new functions, figure out ways of implementing real hardware banging stuff into easier to use code.
This way would have encouraged the use of librarys and stuff could have stayed compatable.
Beween 1.2 , 1.3 , 2.0 then 3 were 8 or so years, 4 releases of possible optimization, growth and compatablity

I guess this also could have been a bit of a thorn in CBM side as too much software relied on the hardware not changing...

[Miggy4eva]

I don't really agree, if you make killer hardware then software can adapt. Concentrate on the great hardware and the rest will follow

[Predseda]

Originally Posted by Miggy4eva

I don't really agree, if you make killer hardware then software can adapt. Concentrate on the great hardware and the rest will follow

It is not so true. Till today all the new PD games are written to work on a basic A500 config. No one makes anything that would benefit from AGA for example. Something that I can not understand. If I would be capable to produce a SW, I would do it for the best Amigas widely available - A1200 - as a base.

[drHirudo]

Originally Posted by Predseda

It is not so true. Till today all the new PD games are written to work on a basic A500 config. No one makes anything that would benefit from AGA for example. Something that I can not understand. If I would be capable to produce a SW, I would do it for the best Amigas widely available - A1200 - as a base.

Basic A500 config gives you the widest audience. If something works for A500, it will work on A600, A1200, A2000, A3000, A4000. While if you make the game for Amiga A1200, it will not work on A600, A2000, A3000. Also it's much easier to program for ECS chipset.

[Megol]

Originally Posted by Miggy4eva

1) Commodore should have put midi ports on the Amiga from DAY ONE. It was unacceptable that a superior machine with great sound was beaten in the music market by an inferior machine which had them included. Yes, adding midi ports is very simple and very easy but musical types like everything to be easy out of the box.

IIRC it would only have required a DIN connector. But it wouldn't be used by most people and at least some software included a MIDI dongle.
Don't think it would make any difference.

2) 68000 should have been at least 8mhz but 10mhz ideal with 32k of fast mem.

What? Don't think you know much about the Amiga hardware - everything is synchronized so one can't change the processor clock freely without making the interface asyncronous.

Changing to a 14MHz 68k would be relatively easy but require $$$.

Don't know what you mean by 32k mem. It just doesn't make sense unless you are thinking of a cache - which really wouldn't make any sense!

3) NON-INTERLACED 640 x 480 4 colour mode at 70hz for productivity, office, midi software usage. Something else an inferior machine managed to offer which helped with midi software, crisp image better for staring at for long periods than normal CRT's of the day.

The Amiga was designed for video standard signals and not expensive monitors.

And again: the Amiga uses syncronized components. Adding this would require a major redesign or an external graphics card.

4) 256 colour mode from palette of 262,144 colours, supported by all hardware sprites, blitter and copper

When the computer was designed the difference between 32 palette entries and 256 was significant in cost. 256Ki colors too - the DAC would become more expensive.

5) PC-compatible disk format and A500 from 1987 onwards should have introduced 1.44MB HD drive.

Guess you are young or new to computers? I'm not and remember the premium price of HD floppy drives.

Commodore didn't like to spend money.

6) A1200 should have had a DSP, 030, more colours, and some kind of built in modem to propel Amiga into the internet age. That was a BIG miss by Commodore.

Now it's becoming ludicrous.

7) EDIT of course! Just one, ONE chunky graphics mode to make Wolf3d and Doom possible

Would have been relatively easy to add but require a redesign of the chipset. $$$.

Ahahaha at your comparison to other machines. yes it's true, Commodore should have got better partnerships, even with Microsoft for early Office versions. This would have positioned them much better to compete against the PC in the business and office market.

The competition was lost when PCs became the standard.

[Retro1234]

Should definitely had better support for PC stuff - PC disks, BMP images Word docs etc
and a good text editor - notepad was terrible.

In this respect PC disks and word processers Atari ST was better.

[Amigajay]

Exactly a lot of comments are crazy, I think people forget the price of things back then, the Amiga was a budget computer first and foremost when, sure add what you like to desktop models, but the price always reflected what you got, the main things missed was a little fastram and a lot better custom chips in the A1200/4000 models in 1992, who the bloody hell wants midi support in a games/budget computer!?

[ajk]

When the Amiga came out (the A1000) in 1985, it was hardly a budget computer though. It was quite expensive and advertised as a multimedia system, for audio, graphics design and other productivity and of course games too.

Later on the A500 could be considered a budget computer, although it still wasn't that cheap when launched.

[duga]

Barrel batteries.

The release of A300/A600.

No fast mem in A1200.

[Samurai_Crow]

The mistakes at the beginning? There weren't many but there was a few annoying bugs in the polygon plotting capabilities because they were rushed to production. If they had been corrected, the simplicity of polygon plotting would have made 3D games easier and more powerful.

The line draw mode would plot one too many pixels at the bottom making a single pixel that would screw up the fill mode afterward.

Secondly, the two fill modes would either leave the border bits alone making it impossible to make single pixel widths and the other one deleted both border bits making it impossible to create single pixel spaces in the fill mode. Both should have been replaced by a fill mode that would delete only the rightmost border bit. In that way the circuit would have been simpler and still created less undesirable results.

Finally, the sprite multiplexer should have been copper based instead of setting sprite heights to a maximum of 254 pixels and requiring that all images be stored sequentially in the sprite DMA channels. Most sprite images are duplicated somewhere else in chip RAM thus this violates the "don't repeat yourself" principle that was one of the founding policies of the design.

Ultimately though, the biggest mistake was not fixing bugs but leaving things alone to avoid software breakage. This policy prevented AAA and AGA from being successful later on because there were no system-friendly code bases and everybody was mercilessly banging the hardware independently of the OS functions. Kickstart bugfixes could have been used to patch bugs in the chipsets, but instead later patches were required like WHDLoad, KillAGA, Degrader, Relokick and Tude. MrgCop() in Graphics.library was buggy enough that very few people ever used it in production software and as such, more powerful versions of the copper coprocessor became impossible.

[Juz400]

If you look at the computer chronicles episodes that had Amiga in the early days, it was a whole new level compared to what was before.
(I do find it cringeworthy when they do the `look we can change the colours of workbench` and repeat again with the 3000. It seems like scraping the barrel of Ideas to interest potential customers)

The levels of power (Mainframe/Minicomputer/PC) each had a market and the machines were built to cater for that market.
Home computers were just something to play with and learn the basics before putting on your big boy pants at college and using `real` computers.
Amiga and Atari were looked down on by the `big boys` they were both powerful machines but nobody new who they catered for or what market to aim them at.
In those days if you were going to spend big bucks on a computer you would have been steered to a PC compatable or Mac as they were for grown ups.

VGA didint appear until 1987, the chipset was already set in stone(silicon) by then so 256 colour BMP images were a kinda new thing then. AGA should have had something to speed up conversion.
Amiga floppy was programable, 720k DOS compatable from the start just a bit slow.

RE graphics.library - fresh sets of eyes help bug fixes, getting together with programmers who write for your OS would have helped immensly.

One problem I overlooked, I Hated the 3D stuff on the Super Nintendo (Mode 7 IIRC)
All the games looked the same with just different colours, got old VERY quickly!
The good thing about computers is you generally had to write EVERY part of your game
Using set routines to do your games kinda makes them all cookie cut...

[Foebane]

3. HAM mode was a hack, mostly useless for other than static pictures!

[ Show youtube player ] would beg to differ with you, as would [ Show youtube player ], along with many other realtime HAM animated Amiga demos.

[idrougge]

Originally Posted by Juz400

One problem I overlooked, I Hated the 3D stuff on the Super Nintendo (Mode 7 IIRC)
All the games looked the same with just different colours, got old VERY quickly!
The good thing about computers is you generally had to write EVERY part of your game
Using set routines to do your games kinda makes them all cookie cut...

You could say the same thing about copper rainbows on the Amiga.
It's not as though console developers at the time were given large support libraries with ready-made routines either. They used the hardware that was available. Check out Albert Odyssey for the SNES for an innovative use of Mode 7.

[plasmab]

As previously said.. Commodore didn't design the Amiga.

Biggest gripe from me is that the disk filesystems sucks. So much so that DiskSalv exists.

Oh and a comedy disk controller built into the chipsets when there were off the shelf solutions.

[idrougge]

Did any good ones exist?