Did Commodore give Amiga 1000 developers enough support?

[rhester72]

Originally Posted by Megalomaniac

Looks like they were more helpful in the A1000 era than the A500 era, surprisingly.

That isn't true at all. CATS was a _very_ big deal to commercial (hardware and software) developers and wasn't even a thing until after the A500 and went on through the end of Commodore as a company.

Granted, it wasn't free, but good developer support never is.

[ImmortalA1000]

Originally Posted by rhester72

That isn't true at all. CATS was a _very_ big deal to commercial (hardware and software) developers and wasn't even a thing until after the A500 and went on through the end of Commodore as a company.

Granted, it wasn't free, but good developer support never is.

But were the CATS team capable of writing Lotus Turbo 2.5D quality routines, Beast 1 overland/intro level parallax scrolling routines that race the beam? This is the sort of level of coding talent and developer support the A500 needed to match the little PC Engine of 1987.

Books are of zero use unless the coder is already as talented as Martin Edmondson et al

They needed something like Sony's Playstation1 Analyser used by Namco to write Ridge Racer IV for something as complex as the Amiga chipset+68000 combo used to get its pixel pushing grunt.

Amiga != C64 simple to hack and push type hardware, you need genuine multiprocessor coding skills to make it do bugger all remotely jaw dropping IMO. If all 2.5D racing games and side scrolling Rastan Saga knock-offs were Lotus II/Beast 1 quality a lot more Amigas would have sold IMO. I think Commodore development team members need to accept that as their failing as much as the coders who weren't ever up to Martin/Shaun standards of OCS coding genius etc

[Megalomaniac]

Reflections were able to design Beast 1 from the ground up to use the Amiga to its limits, with no consideration for other formats (although it did end up on the 8-bits as well as the ST), no deadline and no publisher pressure or expectations. Lotus 2 at least partly the same, though I expected Gremlin wanted it ready for Christmas, and probably insisted on the ST version too. If you're doing an arcade conversion or film license you don't get those luxuries.

[ImmortalA1000]

These, along with Turrican 3 make up the holy trinity of A500 games, the peak of development.

For whatever reason, be it complexity of extracting 100% out of OCS chipset vs SNES/MD hardware or lack of support, these were really rare games. If only publishers sought out the developers with the best fit for the type of game engines needed it might have been OK. Ultimately though the software houses never chased the talent IMO nor did Commodore help those inferior coders get up to that "Gold standard" for various game engines.

Not sure what to make of that.

[TCD]

Originally Posted by ImmortalA1000

Ultimately though the software houses never chased the talent IMO nor did Commodore help those inferior coders get up to that "Gold standard" for various game engines.

Not sure what to make of that.

Software houses chased sales numbers, no matter if the game was well coded and designed or not. In the same way Commodore cared about sold machines and while I'm quite sure they knew that a good number of their machines sold for gaming, they also didn't care how well those games use the hardware. I don't think that quality control for games happened on any computer at that time (or today for that matter). That was and is done only for consoles.

[Phantasm]

Most of the top developers didn't rely on anyone to figure out how to get the best out of systems in those days. They were making their own development tools and figuring out how to push systems to the limit. The official documentation and tools will only get you onto an equal footing with everyone else who has access to them. If you want to do better you have to figure it out for yourself

[TCD]

Originally Posted by Phantasm

They were making their own development tools and figuring out how to push systems to the limit.

Always loved to read the 'behind the scenes' kind of articles in magazines where you got to see some screenshots of those tools that they used.

[Bruce Abbott]

Originally Posted by ImmortalA1000

But were the CATS team capable of writing Lotus Turbo 2.5D quality routines, Beast 1 overland/intro level parallax scrolling routines that race the beam? This is the sort of level of coding talent and developer support the A500 needed to match the little PC Engine of 1987.

Commodore's software engineers weren't writing hardware banging games, they were doing OS stuff. The idea that they should have built up the expertise in house and spoon-feed it to clueless developers who would cut and paste it into their code is ridiculous. No other home computer developers expected that kind of support.

Books are of zero use unless the coder is already as talented as Martin Edmondson et al

I disagree. Any competent coder should have had no problem getting to grips with the Amiga's hardware. I had a copy of the hardware reference manual in 1986, a year before I purchased an A1000. Had I wanted to develop a game it wouldn't have taken much to get into it, and I am just an average coder.

They needed something like Sony's Playstation1 Analyser used by Namco to write Ridge Racer IV for something as complex as the Amiga chipset+68000 combo used to get its pixel pushing grunt.

Why? Poking Color00 will tell you most of what you need to know.

Amiga != C64 simple to hack and push type hardware, you need genuine multiprocessor coding skills to make it do bugger all remotely jaw dropping IMO.

There's your problem. A good game doesn't necessarily need 'jaw dropping' coding skills, just a competent game engine. The hard work is in all the other stuff.

I think Commodore development team members need to accept that as their failing as much as the coders who weren't ever up to Martin/Shaun standards of OCS coding genius etc

It wasn't Commodore's job to make up for developers' poor coding skills. They did provide support where it was needed, particularly on the OS side which was quite complex.

I was a registered CDTV developer, which only cost US$25 per year and had few strings attached. Unlike other Amiga models you had to submit your product to Commodore for quality control before it could be released. They tested our title and asked how I was able to get the loading so fast. I pointed out that I just used the technique they suggested in the CDTV developer notes!

As I said before I am just an average coder. I have never tried to make a hardware banging game engine for the Amiga. But for the stuff we needed I found the OS functions did the job fine. Many games do not require 'pulling out all the stops' to get the desired effect. Most stand or fall on the game design, artistry and playability, not frame rate and size/number of objects moving around.

A good game will not try to be 'jaw dropping' at the expense of other factors. Shadow of the Beast is a good example of a game that failed there. Some others such as Outrun should never have been released if they couldn't make it fun to play.

[Megalomaniac]

Originally Posted by Bruce Abbott

I was a registered CDTV developer, which only cost US$25 per year and had few strings attached. Unlike other Amiga models you had to submit your product to Commodore for quality control before it could be released. They tested our title and asked how I was able to get the loading so fast. I pointed out that I just used the technique they suggested in the CDTV developer notes!

Does this imply that other CDTV developers were overcomplicating things with their own slower loading routines? I know that registered developers on the Amiga side had to only do system-legal code, which probably prevented most of the fancy hardware tricks seen in the most technically impressive arcade games, but presumably limited compatibility issues with later models.

I love the OP's idea that a programmer who achieved something impressive should have shared his routines with rival programmers, potentially with rival publishers. This wasn't communism, and Magnetic Fields weren't a registered charity. And imagine how bored you'd be if 10 other companies did games with the same routines as each other.

Incidentally, did Commodore know that the Amiga was more complicated to code to its limits than the C64 was when they launched it? Even when they launched the A500, did they perceive how far beyond the ST its ultimate potential was? The general public seemingly didn't.

[desiv]

Originally Posted by Megalomaniac

Incidentally, did Commodore know that the Amiga was more complicated to code to its limits than the C64 was when they launched it? .

I'd be willing to bet Commodore provided more support options to developers for Amiga DEV than C64 for that reason...
I think both got the basic hardware programmers reference manuals, but the Amiga also had all the OS dev information...

That said, I doubt most computer companies spent a lot of time focusing on hardware banging support for game programmers...

These are computers and not game consoles... ;-)