Things the Amiga didn't get right from Day 1

[tolkien]

Originally Posted by rsn8887

Something else Amiga did wrong from day 1: The super ugly, crappy boot picture that looks like it was drawn in 30 seconds.

Reading this It looks to me no so ugly.
http://eab.abime.net/showpost.php?p=...6&postcount=52

[Thomas Richter]

Actually, it's drawn by an algorithm. The biggest problem is, as always, to fit the drawing instructions into ROM.

[paraj]

Originally Posted by Thomas Richter

Actually, it's drawn by an algorithm. The biggest problem is, as always, to fit the drawing instructions into ROM.

More discussion on SE for people curious (also references the post tolkien linked): https://retrocomputing.stackexchange...graphic-so-bad


I guess A1000 boot ROM contains simplified/hardcoded versions of the necessary drawing routines since it can't possibly embed graphics.library.

[TEG]

Originally Posted by tolkien

Reading this It looks to me no so ugly.
http://eab.abime.net/showpost.php?p=...6&postcount=52

In one hand this drawing gave the Amiga a real personality but in the other hand, yeah, it was not as clean as the MAC boot icon, perhaps giving the idea of a less polished machine but less boring. What was the Amiga.







So hard to say from my point of view. Anyway the surrounding was more decisive.

[Bruce Abbott]

Originally Posted by TEG

In one hand this drawing gave the Amiga a real personality but in the other hand, yeah, it was not as clean as the MAC boot icon, perhaps giving the idea of a less polished machine but less boring. What was the Amiga.

Pokey little monochrome icon on a pokey little screen - nowhere near the same impact as a hand holding up a huge brightly colored disk.

The only thing I remember about booting a Mac was seeing that Sad Mac way too many times. Makes me shiver just thinking about it. I got my revenge though - stripped that crappy thing down and melted the chips off the PCB with a heat gun.

[Bruce Abbott]

Originally Posted by AmiBoy

Atari GEM was not crappy. It was the best that could be done when everything is in ROM.

Yes, it was.

They could have done better, but they needed to get it out pronto - for obvious reasons. The Amiga would have wiped the floor with the ST if it had reached the market a few months earlier in a finished state.

I had Ataris and Amiga back in the days before I sell both (and reacquire them few years ago). I wish TOS/GEM on my Ataris were capable as the KS/WB was. The simplicity of TOS/GEM was amazing and made things easy but was just that.

'Easy' but oh so limiting, and it didn't look great either.

In my mind the thing that Amiga did wrong from day 1 was that its custom chipset was not an add-on card, an expansion so to speak,

And today we have... graphics and sound sound chips built onto the motherboard of nearly all modern PCs. The Amiga was too far ahead of its time.

with an API and standard commands for the various interfaces it provided, that any 3rd party or even Commodore could update without having to think about compatibility. But this was hard back in the days.

'Back in the days' the PC's idea of an API was the BIOS, which only supported one graphics card (CGA). So how did they maintain compatibility? By including CGA graphics in every card since. If Commodore had survived they would have done the same with OCS (and just like on PCs it would have become less and less relevant as the machines got more powerful).

[Thomas Richter]

Originally Posted by Bruce Abbott

'Back in the days' the PC's idea of an API was the BIOS, which only supported one graphics card (CGA). So how did they maintain compatibility? By including CGA graphics in every card since. If Commodore had survived they would have done the same with OCS (and just like on PCs it would have become less and less relevant as the machines got more powerful).

Actually, this is a bit more complicated. The VESA Bios can support more than CGA, and indeed, all chunky modes beyond 640x480 are way beyond what you could do with CGA. The mode request/mode selection through VESA bios is thus not CGA.


It is true, though, that the CGA modes are almost always supported by PC graphics chip sets, but mostly because they are a subset of the modes the VESA bios can do for you.


Some (probably most) PC cards support the legacy modes from completely different "silicon" than their "accelerated" modes. For example, the 3DPermedia chip (not really modern anymore today, of course) consists of two graphics engines, a legacy engine that had been used back then to boot the machine, and a "3D display" which provides all the chunky and 3D accelerated modes. Thus, in principle, it can do the planar CGA modes, just that almost noone uses them, probably except the bios boot screen and the bios menu.


Nowadays, all of this is being phased out, given that we no longer have bios, but uefi.

[Predseda]

Yes, there is a thread where original author of Hand holding floppy picture was talking about how that screen was made. I think it is okay for Amiga 1000, but silly for KS1.2 and 1.3.

[Bruce Abbott]

Originally Posted by Thomas Richter

Actually, this is a bit more complicated.

With PCs it got more and more complicated as time went on. Had Commodore survived the Amiga would have too. At what time would they have decided to drop OCS support completely? The same time that we decided the latest 'Amigas' were not actually Amigas at all - and this did actually happen!

Nowadays, all of this is being phased out, given that we no longer have bios, but uefi.

I have given up trying to keep up with the latest PC developments. Not interesting anymore.

The Amiga however...

[Clockodile]

Originally Posted by Bruce Abbott

With PCs it got more and more complicated as time went on. Had Commodore survived the Amiga would have too. At what time would they have decided to drop OCS support completely? The same time that we decided the latest 'Amigas' were not actually Amigas at all - and this did actually happen!

IMO, we kinda have the answer to that question. Right before Commodore's bankruptcy in 1994, they were planning the Hombre chipset, which had no planar modes, practically no sprites, and lacked other traditional Amiga hallmarks. It was intended to be only compatible with the older Amiga's applications on the OS library level and nothing else. And, using a PA-RISC CPU, it would have been quite different from the M68K based Amiga's in both the CPU and surrounding components.

[AmiBoy]

Originally Posted by Bruce Abbott

Yes, it was.

They could have done better, but they needed to get it out pronto - for obvious reasons. The Amiga would have wiped the floor with the ST if it had reached the market a few months earlier in a finished state.

'Easy' but oh so limiting, and it didn't look great either.

They could have, they didn't have the time. They maybe had done better or worse if they had the time. Let's argue on what has actually happened.
It was indeed simple and limiting, in terms that you needed a 3rd party tool to edit a file. Quite resembling the Amiga philosophy here to make things modular and let people create the tools they need. But IMHO some tools should have been integrated.
Didn't look great, so true again. Actually it was looking basic. So basic that anyone could start using it directly but if it had been more like Apple Finder things would be better.

WB/KS was obviously more featured and advanced and for the same features you had to pay a lot of money on the Atari.

Originally Posted by Bruce Abbott

And today we have... graphics and sound sound chips built onto the motherboard of nearly all modern PCs. The Amiga was too far ahead of its time.

'Back in the days' the PC's idea of an API was the BIOS, which only supported one graphics card (CGA). So how did they maintain compatibility? By including CGA graphics in every card since. If Commodore had survived they would have done the same with OCS (and just like on PCs it would have become less and less relevant as the machines got more powerful).

The graphics/sound chips built onto the motherboards today is not an enterprise decision for a platform but instead something some companies offer on a wide open market for people (read customers) to get the basic set of features without paying much especially for features you don't care about, eg in an enterprise environment. It is not mandatory for the architecture. And I don't see the Amiga as the equivalent of a modern cheap PC.

The Amiga was a closed platform, meaning that only Commodore could decide what the Amiga would feature on its motherboard. And it feels like the chipset that made it great, the lack of development on it to be more precise, was the decisive factor for the platform decline.
My 2c

[NorthWay]

As I was just reminded from another thread here:
Different (and crashing) behaviour between CLI and Workbench startup. That should have been possible to handle uniformly and/or "behind the scenes".

[Overdoc]

I won't comment any longer on the Amiga system flaws I have already explained in every detail, so that a 3 year old kid would understand, because there are some guys with serious comprehension deficites like daxb...it is simply too tedious.

But there are other flaws to mention, which are really funny.
The Amiga 500 manual mentions that the OS is able to acces up to 4 drives, DF0 to DF3.
But nobody knows why the developers did even bother to implement support for a DF2 or DF3, if it was not even possible to connect more than one external drive, because Commodore's external Amiga 500 drive model 'A-1011' lacked a connector for another drive!

Also the fact that every connected external drive eats up ram of the computer is quite laughable.
But still, they didn't bother to implement a switch for turning off that A-1011 drive....

Same with the original A-501 ram expansion missing a on/off switch.
Obviously developers were too sure about software still being 100% compatible with a ramexpansion connected.

Good thing is that 3rd party developers recognized those issues, so we had good and cheap alternatives to those stupid original Amiga accessories.

[Chucky]

A connected drive takes ram due to buffers. it is "by design" nothing special by that.

and memexpansion was 100% compatible. do not blame hardware on coders that did not follow the rules that was.. it was the software that wasn't compatible. not the opposite.

[Daedalus]

Originally Posted by Overdoc

I won't comment any longer on the Amiga system flaws I have already explained in every detail, so that a 3 year old kid would understand, because there are some guys with serious comprehension deficites like daxb...it is simply too tedious.

Aw, diddums, substituting personal insults for making a convincing argument.

But there are other flaws to mention, which are really funny.

Indeed, it's far from a perfect system, but much of what you see as flaws are simply design decisions that you don't seem to understand.

The Amiga 500 manual mentions that the OS is able to acces up to 4 drives, DF0 to DF3.
But nobody knows why the developers did even bother to implement support for a DF2 or DF3, if it was not even possible to connect more than one external drive, because Commodore's external Amiga 500 drive model 'A-1011' lacked a connector for another drive!

For potential future hardware. That's quite clever forward thinking from the developers, which proved useful when drives later became available with passthrough. There are many other aspects of the OS that show this sort of forward thinking too - it's quite an impressive feature of the OS, rather than a flaw.

Also the fact that every connected external drive eats up ram of the computer is quite laughable.

As was explained, drives use buffers to greatly increase performance in certain activities. Would you prefer this wasn't the case? Or would you prefer if the OS allocated this RAM for all 4 drives, even if no external drives were attached? Do enlighten us as to how you would have made it better.

But still, they didn't bother to implement a switch for turning off that A-1011 drive....

If you're that stuck for 30kB of RAM, you probably need a RAM expansion, not an on/off switch.

Same with the original A-501 ram expansion missing a on/off switch.
Obviously developers were too sure about software still being 100% compatible with a ramexpansion connected.

Again, as was explained, that's an issue with a few badly coded games. Any software that follows the rules and allocates RAM properly will work just fine, as the majority of games do.

Having to flip the machine over, open the trapdoor cover, stick fingers in and flick a fiddly little switch is hardly user-friendly. Or worse, running a wire to the back of the machine, drilling a hole in the case and mounting an ugly switch is hardly a user-friendly solution either.

[Overdoc]

Originally Posted by Daedalus

Aw, diddums, substituting personal insults for making a convincing argument.

No, I listed arguments, even gave extremely simply examples (listing disk contents) in difference to others who childishly just say 'no, no no'.
User daxb started asking if I had mental problems, talked about 'stupid' etc, etc., because he ran out of arguments? Just read his comment....

If you like we can get proof of my example. I bet for something that I will be faster listing a disk's contents on the C64 than on the 'so well designed and perfect' Amiga system.
the C64 beats the Amiga in facts in this case, no matter what Amiga fanatics talk.

Originally Posted by Daedalus

Indeed, it's far from a perfect system, but much of what you see as flaws are simply design decisions that you don't seem to understand.

And, what's the difference?
if those issues were already designed as that, then it is exactly what I have already mentioned: bad design!

Originally Posted by Daedalus

For potential future hardware. That's quite clever forward thinking from the developers, which proved useful when drives later became available with passthrough. There are many other aspects of the OS that show this sort of forward thinking too - it's quite an impressive feature of the OS, rather than a flaw.

Yes, for future hardware which did not suit to their own design?
The A-1011 drive was designed to be the extrenal drive for the Amiga 500, and afaik came AFTER the Amiga 500 release, but did not have a passthrough.
It was simply stupid design of this drive.

If there had not been 3rd party companies helping Commodore out, then Amiga 500 users would have never ever seen a DF2 or DF3 drive.

Originally Posted by Daedalus

As was explained, drives use buffers to greatly increase performance in certain activities. Would you prefer this wasn't the case? Or would you prefer if the OS allocated this RAM for all 4 drives, even if no external drives were attached? Do enlighten us as to how you would have made it better.

Again, compare this to the C64.
I can connect up to 4 drives without that stupid buffer issues. None of the drives uses up any ram.

Think about how crazy this is in fact! As a programmer, you need to know how many diskdrives your customer/user has connected, otherwise your program might not run. (which actually was the case with some games, and gladly the 3rd party drives had this switch to turn them off, but not Commodore's own A-1011 drive....really funny )

Originally Posted by Daedalus

If you're that stuck for 30kB of RAM, you probably need a RAM expansion, not an on/off switch.

Haha, so I need to buy another extra hardware (ram expansion) to get the other badly designed hardware (diskdrive) working without issues?

Man, seriously, this cannot be true!!
You all keep telling me that this is such a well thought design which was ages in advance of it's time ?!
What a joke!

Originally Posted by Daedalus

Again, as was explained, that's an issue with a few badly coded games. Any software that follows the rules and allocates RAM properly will work just fine, as the majority of games do.

Ok, but then it is not 'hardware compatible' at all.

Originally Posted by Daedalus

Having to flip the machine over, open the trapdoor cover, stick fingers in and flick a fiddly little switch is hardly user-friendly. Or worse, running a wire to the back of the machine, drilling a hole in the case and mounting an ugly switch is hardly a user-friendly solution either.

Actually, what you describe was the case when you had the original A-501 expansion. You had to remove the expansion.
At least with 3rd party expansions you had the switch taped to the Amigas side, or you simply had to drilled that hole. (which is by far more easy than removing the expansion all the time)

Commodore designed a ram expansion which was not fully hardware compatible, but yet didn't bother to install a switch to make it backward compatible.
If developers had really been such great minds thinking about future expansions and all, then they might have implemented a nice looking switch in the case by design, which then could have been used by future expansions.
They also could have designed this 'trapdoor slot' a bit better, so that you would have been able to connect and remove the expansion easily from the right side of the Amiga. Then it wouldn't have been a big issue, you could have just attach and remove it easily like you do with a cartridge on the C64.
But no, they made it at the bottom!! of the machine....

Like I mentioned so often before.....user friendly design obviously was not the Amiga's strength

But it was a very good machine because of it's special chips - that's what gave us all those great games and demos etc.

[Chucky]

Again, compare this to the C64.
I can connect up to 4 drives without that stupid buffer issues. None of the drives uses up any ram.

well on a machine that had max 64K of adressable ram. there was no room for luxery as buffers!

Haha, so I need to buy another extra hardware (ram expansion) to get the other badly designed hardware (diskdrive) working without issues?

It is not a "bad design" just as you do not understand it.

Ok, but then it is not 'hardware compatible' at all.

It sure is. software however was not written by the set rules. again.. do not blame the hardaware when people programmed it as if it was a C64.
(oh seing your arguments here. it makes sense. they as you belived this was a c64...)

Commodore designed a ram expansion which was not fully hardware compatible, but yet didn't bother to install a switch to make it backward compatible.
If developers had really been such great minds thinking about future expansions and all, then they might have implemented a nice looking switch in the case by design, which then could have been used by future expansions.

It was fully compatible! it was just bad coders doing shit. nothing else. again. do not blame hardware for bad programmers.

[Daedalus]

Originally Posted by Overdoc

<snip>

It really sounds like you want people to take you seriously, but all you're doing is ignoring anything that doesn't fit your skewed views, and instead repeating the same nonsense that has already been shot down over and over again, presumably because that helps you believe that nonsense is true. So you either have delusions of your own infallibility, or you're a troll. And either way, you can't expect anyone to have a normal, grown-up discussion with you.

[grond]

Can't we cut out the acid talk?

I think it is a design flaw that the Commodore external FDD didn't have the passthrough port (I had never heard about this before, I'm pretty sure the external FDD that is somewhere in my basement has one but it's probably not Commodore branded hardware). In any case it would be typical of Commodore: save a few pennies for a port and connector at the expense of the customer.

As for the buffers, I agree with the others that this rather shows how much advanced the AmigaOS really was. When did DOS/Windows become as flexible as that? When did DOS/Windows ever allocate only those resources to the system that it really needed at this point in time? If it ever did, then only at a point in time when 30KB more or less didn't matter at all. The C64 doesn't have any buffers so no wonder it doesn't need any extra RAM to connect more devices. It doesn't offer you buffered accesses either which do serve a purpose.

[ajk]

The earlier A1010 does have a pass-through so it makes sense for the OS to have support for multiple drives. Why it was omitted from the A1011 is anyone's guess, probably just a cost-saving measure or perhaps some attempt to make the design smaller in size. In any case there certainly were drives available that could make use of the multi-drive support (just not the A1011).

If one of these systems suffered from poor disk drive design, it was most certainly the C64, not Amiga. The 1541 was ridiculously complex and expensive, mechanically unreliable, extremely slow (although 3rd party fast loaders could help), ran super hot and even the support in CBM Basic was very lacking (no way to copy from drive to drive, for example).