Is the Amiga hard to program?

[Hewitson]

Originally Posted by mcgeezer

Believe me... compared to coding an Atari ST the Amiga is a dream.

I'm certainly not going to argue with you as you've spent a lot more time coding on the Amiga than I have. But in my limited experience I've found the opposite. Using the ST's traps is much easier and more convenient than fucking around with the Amiga's stupid libraries.

[StingRay]

Originally Posted by Hewitson

Using the ST's traps is much easier and more convenient than fucking around with the Amiga's stupid libraries.

Amiga libraries are pretty straightforward to use, the TOS trap calls use (almost) the same principle anyway. Besides, nobody forces you to use the "stupid libraries" at all.

[meynaf]

The ST is a simple machine. By such, it offers less possibilities, but things are more immediate.

[Hewitson]

Agreed, but you'd have to admit that opening a library, using it's functions, and then closing it, is far more complicated than simply doing a TRAP call.

Someone who knew basically nothing about the ST could easily write a "hello world". The same can not be said for the Amiga as you do require some knowledge of the operating system and how it works.

[Leffmann]

Originally Posted by Hewitson

Agreed, but you'd have to admit that opening a library, using it's functions, and then closing it, is far more complicated than simply doing a TRAP call.

Pretty much. Anyone who says AmigaOS is convenient to program has probably not programmed it themselves.

It's not just system calls, most of the system and its APIs are completely byzantine and convoluted, though I do understand that much of this came from the necessity to make it slimmer by removing conveniences.

[Daedalus]

But the discussion seems to be primarily about games, which didn't bother using libraries for the most part anyway. System-friendly software using the API is a different breed altogether, and while it has its own special ways of doing many things, I actually like the API for the most part.

[Ricardo]

Funny, I see you are thinking about programming the OS. When I think of "Programming the Amiga and ST" I automatically think about the hardware only. Hence programming the Atari was much more simple because you only had to deal with the 68k and a simple sound chip. At the time I never even considered using either operating system.

[NorthWay]

Programming the Amiga through the OS can be as complex as you can possibly make it - some have even made their own GUI support system simply to make their application.

Banging the metal directly does ironically enough also require a bit of OS understanding; if you don't know how to shut things down and what idiosyncracies you have to be aware of you will probably do a bodge-job that only works on your own machine.
Interestingly, there have been a number of good how-to sources since the growth of internet that has made this much easier.

Once down to the metal, I'd say the Amiga is pretty easy to program. You still need to know what is good and bad practices and look for ways to program "with" the hw rather than "against" it. For example, it took many years before using interleaved bitplanes became common.

[mcgeezer]

Just to explain where I'm coming from here with a couple of examples.

On the ST, if you want more than 200 lines of display you have to open the bottom border which the machine was never designed to do. It was only by fucking around with weird hardware that the ST boys of the day managed to do it... i think the technique is to wait for the last line of the display then put it into 60hz and then back to 50hz which causes the video chip to read more memory. Contrast that with the Amiga and you just have to set the appropriate registers in the copper and it's all taken care of.

Another example is rasters (copper bars)... try doing them on the ST is a pain in the arse... on the Amiga it's just ... well.. copper bars.

Or perhaps try setting some bitplanes at specific addresses in memory....no go on the ST as it doesn't have a low byte meaning you have to set displays on a 256 byte boundary, the Amiga has no such limitation.

I'm not really an Amiga OS programmer but the bits I do know like allocating memory seem every bit as straight forward as it was on the ST.

I could go on..... the Amiga is a dream compared to that piece of shit that Atari churned out.

[mc6809e]

Originally Posted by Shatterhand

What I believe could be a little hard is how to get the most of the Blitter. I am really surprised how slow the Blitter is, and it's amazing what some coders could pull it off with it, and then see some people discussing it with so much precision.

But even without being a blitter wizard, you still can do a lot of stuff pretty well.

If you're using blitz basic for blits then, yeah, the blitter is going to appear slow.

And if you're blitting and waiting and then blitting and waiting and on and on, it's going to appear to be even slower.

If you really want to see how much of a difference the blitter makes, compare Lotus 2 on the Amiga with that of Lotus 2 on the Atari ST.

The programmers use blitter interrupts to do a series of blits, starting each blit and then going back to do more work with the CPU.

I do agree it can be hard to get the most out of the blitter. To get the most out of it you need to make a sort of blitter queue by using the copper. The tricky part is handling the case where the copper list might be interrupted by vsync. It is possible, though. You have to set up a copper list that holds off on starting the blitter until the threat of a vsync is past and if vsync DOES happen in the middle of a blit, the copper list strobe address points to a wait or skip with BFD set to 0. Also, the CPU must be used to reload bitplane pointers, etc since the copper is doing all the blitter work.

Had the designers included a generic example of the solution I think blitter driving copper lists would have been more common.

[touko]

A little question, nowadays (only on the technical side),if you do a remake of shadow of the beast, what kind of technical improvements can we expect ??

[idrougge]

Originally Posted by mc6809e

If you really want to see how much of a difference the blitter makes, compare Lotus 2 on the Amiga with that of Lotus 2 on the Atari ST.

The programmers use blitter interrupts to do a series of blits, starting each blit and then going back to do more work with the CPU.

I think this is a fine example of how the Amiga is hard to program. You need to juggle the blitter and the copper, choose the correct screen modes (and write a corresponding copperlist) and then add some sprite tricks to that.

Some say that the Amiga should not be compared to the later 16-bit consoles because they are, well, later – but the fact is that we have been comparing our Amigas to those consoles ever since they came out. We don't fancy comparing our dear platform of choice to the C64 or ZX Spectrum, and no-one has hitherto established some kind of intermediate hardware generation to which the Amiga and ST would belong.
At least back in the early nineties, a lot of Amigans considered the Amiga superior to the Megadrive and SNES, or at least on par with the 16-bit consoles. However, the MD and SNES map more closely to the arcade hardware available at the time, and map more closely to the most popular genres of the time. The Amiga does not.

Just as porting an Amiga-style game to the ST requires a lot of extra work, opening the borders, making a stable raster interrupt for copper rainbows and enabling fast scrolling and screen updates, porting a console-style game to the Amiga with sufficient fidelity requires a lot of planning and set-up.

The Amiga hardware is a bit more general than the Megadrive, so you basically need to program the software equivalent of what is made in hardware by Sega and Nintendo, just as an ST programmer needs to make a software copper and software sprites before implementing the actual game.

[roondar]

Originally Posted by idrougge

Some say that the Amiga should not be compared to the later 16-bit consoles because they are, well, later – but the fact is that we have been comparing our Amigas to those consoles ever since they came out. We don't fancy comparing our dear platform of choice to the C64 or ZX Spectrum, and no-one has hitherto established some kind of intermediate hardware generation to which the Amiga and ST would belong.

I would say that just because it's been done this way for ages doesn't actually mean it's realistic (or smart). We have better information these days than hype in magazines and refusing to use it feels wrong to me.

Case in point: we don't compare the C64 to the Amiga and expect the former to win precisely because everyone understands that for most metrics the C64 will never beat the Amiga (assuming competent programmers for both). In reality, the same goes for the Amiga vs Megadrive/SNES/486 PC. The Amiga will never be able to push as many objects on screen in a single frame as any of these systems (assuming you want to keep them at the same size/fidelity and all systems are competently programmed) and as such comparing it to them anyway is IMHO rather counter productive.

All it seems to accomplish is that people get the idea the Amiga can/should do a lot better than is realistically achievable (just look how often people claim perfectly fine Amiga games that actually use the hardware fairly intelligently are 'lazy' because they don't run in full overscan/64 colours/lack parallax/etc). Alternatively it seem to lead to people coming to the conclusion that the system just wasn't very good because the Megadrive/SNES can display 3-4x the objects in a frame the Amiga can - and with more colours to boot.

And to me, neither of these is fair. The 'mid generation' idea you mentioned is actually far closer to what the Atari/Amiga actually occupy in history than anything else: they're clearly much better than their predecessors yet clearly worse than what came after.

The Amiga was a wonderful machine for 1985. It was still a pretty good/great machine in 1987. But by 1989 it was, well, obviously dated. And I say this while I love the machine, more so since I've become more aware of what makes it tick.

[idrougge]

I would like to say that I wholeheartedly agree, Roondar, but I am an Amiga user too, and it hurts my pride to put the Amiga in a lower stage than the consoles. I prefer to think that the Amiga is not inferior, but that it simply is different because it is not optimised for any single display architecture.
I know I'm deceiving myself, but then again, I'm an Amiga user.

[roondar]

Well, it's not all doom and gloom. The Amiga is objectively better suited to certain games (though the PC will also do well for those) than the consoles and it is incredibly flexible. Flexible enough that it did manage to compete to a point with all that newer hardware.

And if you want to feel even better, there is no doubt in my mind it's possible to create an Amiga game that simply cannot be replicated on the consoles without having to cut back on some elements (though we might have to cheat just a tiny bit here by designing the game to fit the hardware). Just don't expect it to be the rule

[trydowave]

Originally Posted by roondar

I would say that just because it's been done this way for ages doesn't actually mean it's realistic (or smart). We have better information these days than hype in magazines and refusing to use it feels wrong to me.

Case in point: we don't compare the C64 to the Amiga and expect the former to win precisely because everyone understands that for most metrics the C64 will never beat the Amiga (assuming competent programmers for both). In reality, the same goes for the Amiga vs Megadrive/SNES/486 PC. The Amiga will never be able to push as many objects on screen in a single frame as any of these systems (assuming you want to keep them at the same size/fidelity and all systems are competently programmed) and as such comparing it to them anyway is IMHO rather counter productive.

All it seems to accomplish is that people get the idea the Amiga can/should do a lot better than is realistically achievable (just look how often people claim perfectly fine Amiga games that actually use the hardware fairly intelligently are 'lazy' because they don't run in full overscan/64 colours/lack parallax/etc). Alternatively it seem to lead to people coming to the conclusion that the system just wasn't very good because the Megadrive/SNES can display 3-4x the objects in a frame the Amiga can - and with more colours to boot.

And to me, neither of these is fair. The 'mid generation' idea you mentioned is actually far closer to what the Atari/Amiga actually occupy in history than anything else: they're clearly much better than their predecessors yet clearly worse than what came after.

The Amiga was a wonderful machine for 1985. It was still a pretty good/great machine in 1987. But by 1989 it was, well, obviously dated. And I say this while I love the machine, more so since I've become more aware of what makes it tick.

I think most here would agree that the Amiga could not push objects around as well as a snes or megadrive. But what we got was far from what the Amiga was capable of.

Turrican 3 and Flink arent as good as the Md versions but only by a margin. T3 shows that the amiga could of handled competent versions of say rolling thunder and contra.

On another note. When it comes to the homebrew scene. We see lots of c64 games being released now that are not only as good as the games released in the peak 64 years but are even better. Sams Journey, Steel Ranger, Caren and the Tangled Tentacles. Some are better that what the amiga has to offer. Case in point; Sams Journey, in my humble opinion is better than any platformer to grace the Amiga and I say that in terms of game design. The homebrew scene on the amiga has lots of backbone stuff. im not sure if the amigas harder to program for than the 64 or if the latter has the more enthusiastic users?
Worthy was a great release and Reshoot R is looking pretty amazing though.

I would love to see solid conversions of Final Fight and Strider. I know it wont happen. I can dream.

Happy Christmas everyone

[roondar]

I'd say it's a bit of both - the Amiga always had a much smaller user base than the C64 and this likely translates into fewer coders interested in programming for it. And secondly, the C64 just has a lot less hardware features & hardware registers to learn about to get it to do something useful.

And that difference can get big enough to matter - especially when you're just getting started. For example, try to make a simple 'hello world' program without using the OS in C64 assembly and Amiga assembly. You'll quickly find the former to be rather easier to get started on.

Now, once you get to grips with both that difference does become smaller (though I don't feel it goes away completely). However, I feel these differences might discourage people starting out when they see just how much they need to do/know compared to simpler systems to get the basics working.

[idrougge]

The C64 also has better development tools nowadays.

[Juz400]

Originally Posted by roondar

The Amiga was a wonderful machine for 1985. It was still a pretty good/great machine in 1987. But by 1989 it was, well, obviously dated. And I say this while I love the machine, more so since I've become more aware of what makes it tick.

Its kinda funny that you say the Amiga hardware dated by 1989, I think thats about the time when the better coders had learnt how to `conduct` (that guy with the stick at the front of the orchestra) the hardware to get it to perform.


Anyone heard of Grime z80/6502 by Keith S (I dont know him personally)
Its a game he has converted from MS-DOS to pretty much all z80 and 6502 based computers/consoles.
He has just start working on converting it to 68000 based machines (Amiga, Atari, Megadrive, X68000 and NeoGeo)
There isnt much activity on those videos yet so have a look

Dont criticize what he is doing off the bat, it seems logical and how you would convert a game from one platform to another. Get it working first and THEN fine tune/optimize the code!
THAT there my friends is where the shit/lazy software houses/programmers make or break a game

As the videos progress (new one today 29th Dec 2018) I guess we see the differences between the hardware and how its programmed.
https://www.youtube.com/playlist?lis...hYTIrRDsfOc0sf

I look forward to his channel progressing, it would be interesting to see a 68000 conversion of his game ChibiAkumas just to see the difference in hardware again.
I was quite impressed at the amount of stuff moving on the Amstrad CPC!
[ Show youtube player ]

[Tigerskunk]

Originally Posted by roondar

I would say that just because it's been done this way for ages doesn't actually mean it's realistic (or smart). We have better information these days than hype in magazines and refusing to use it feels wrong to me.

Case in point: we don't compare the C64 to the Amiga and expect the former to win precisely because everyone understands that for most metrics the C64 will never beat the Amiga (assuming competent programmers for both). In reality, the same goes for the Amiga vs Megadrive/SNES/486 PC. The Amiga will never be able to push as many objects on screen in a single frame as any of these systems (assuming you want to keep them at the same size/fidelity and all systems are competently programmed) and as such comparing it to them anyway is IMHO rather counter productive.

All it seems to accomplish is that people get the idea the Amiga can/should do a lot better than is realistically achievable (just look how often people claim perfectly fine Amiga games that actually use the hardware fairly intelligently are 'lazy' because they don't run in full overscan/64 colours/lack parallax/etc). Alternatively it seem to lead to people coming to the conclusion that the system just wasn't very good because the Megadrive/SNES can display 3-4x the objects in a frame the Amiga can - and with more colours to boot.

And to me, neither of these is fair. The 'mid generation' idea you mentioned is actually far closer to what the Atari/Amiga actually occupy in history than anything else: they're clearly much better than their predecessors yet clearly worse than what came after.

The Amiga was a wonderful machine for 1985. It was still a pretty good/great machine in 1987. But by 1989 it was, well, obviously dated. And I say this while I love the machine, more so since I've become more aware of what makes it tick.

Great post, where I concur with every single sentence.