Arcade / Console --> Amiga conversion discussions

[OmegaMax]

By the way there's an interview with him for programming G'nG on amiga,It's interesting he talks about his experience and some of the challenges he had with this game.http://amr.abime.net/review_39272

[sandruzzo]

With Rygar we showed what A500 could do: Why don't try with GnG?

[dlfrsilver]

Originally Posted by mcgeezer

You're comparing Apples with Oranges.

At the time the ST was the biggest rival to the Amiga as lots of games were coded on the ST first because it was more established as a games machine given it had the luxury of more time on the shelves.

Biggest rival ? I would say that the ST was simply the 1st 16 bits machine, and the publishers had to get their grip on it.

Publishers had to port their games on Amiga because they were not selling enough copies on Atari ST. they needed to make more bucks by doing Amiga versions as well.

When the Amiga got traction software houses just did the ports of ST games cos they could get them out of the way in a matter of weeks, in reality all they had to deal with was the bitplane, sound, input and floppy differences - everything else was pretty much the same. The slightly faster CPU in the ST was probably negligible and perhaps is the reason that the odd port is more shit on the Amiga than it was on the ST.

The CPU is slightly faster on the ST because the Atari Engineers had a nasty surprise on the prototype they were testing before the ST went mass builded in factory : The quartz they choose at first had the same speed as the one used on the Amiga. Unfortunately, the ST to the contrary of the Amiga had no hardware acceleration, and the prototype was too slow, too slow to the point that it could not be sold. They changed the quartz initially used for a 8mhz one, so that the ST speed could be sufficient to be used. Then they gave green light for mass production of the ST.

This is why the CPU clock speed between both machines is bullshit. It's faster on the ST because the machine was at first way way too slow. The Amiga is the opposite, the machine had to be slowed down so that the CPU gets synchronised with the chipset.

And so far, since you're talking about ports on the Amiga, this was simple : the coders made the Amiga replicates with its chipset what the ST could do with its CPU alone. the result could just be weak in the end. The Amiga has its CPU and its chipset.

[Shatterhand]

To be quite honest I have the feeling many of the ST ports dont even do that. When I see games like UN Squadron or Dynasty Wars where the ST version actually has a smoother framerate, it makes me thing they didn't even bother using the blitter at all, all graphics are drawn by the CPU just like on the ST and then the slight faster CPU on the ST makes a difference.

[Galahad/FLT]

Originally Posted by Shatterhand

To be quite honest I have the feeling many of the ST ports dont even do that. When I see games like UN Squadron or Dynasty Wars where the ST version actually has a smoother framerate, it makes me thing they didn't even bother using the blitter at all, all graphics are drawn by the CPU just like on the ST and then the slight faster CPU on the ST makes a difference.

Plenty of ST ports that didnt use the blitter, some even drew a bitplane copperlist background instead of using a proper copperlist

Developers must literally have been laughing their asses off, being paid to do an ST and Amiga version, and the Amiga version was likely done in a couple of days, the longest wait would have been waiting for the 3rd party musician to supply at least an Amiga title tune.

I still maintain the ST did major damage to the Amiga because of the ease of porting ST games virtually unchanged.

[chb]

Originally Posted by dlfrsilver

The CPU is slightly faster on the ST because the Atari Engineers had a nasty surprise on the prototype they were testing before the ST went mass builded in factory : The quartz they choose at first had the same speed as the one used on the Amiga. Unfortunately, the ST to the contrary of the Amiga had no hardware acceleration, and the prototype was too slow, too slow to the point that it could not be sold. They changed the quartz initially used for a 8mhz one, so that the ST speed could be sufficient to be used. Then they gave green light for mass production of the ST.

Any source/reference for this story? Highly unlikely IMHO, a CPU clock difference of 10-15 percent is barely noticeable. They rather wanted to use the full 8Mhz the 68000 was rated for, on their way violating the DRAM timings specified by their manufacturers (still works because of tolerances).

[mcgeezer]

Originally Posted by dlfrsilver

Biggest rival ? I would say that the ST was simply the 1st 16 bits machine, and the publishers had to get their grip on it.

Publishers had to port their games on Amiga because they were not selling enough copies on Atari ST. they needed to make more bucks by doing Amiga versions as well.

I think we are saying the same thing so I agree with you.

If people weren't buying Amiga's they were buying ST's at £100 less. Those days the developers didn't have any incentive to push the hardware capabilities of the Amiga, it was just make a game available on both platforms and they get access to both markets on the school playground.

The fun started when the true developers started to unlock the capabilities of the Amiga and ST was lagging behind on the playground. I remember playing Hybris on my friend's Amiga when I had an ST, I new then what that extra £100 would have got me.

Happy days.

[OmegaMax]

Originally Posted by mcgeezer

I think we are saying the same thing so I agree with you.
If people weren't buying Amiga's they were buying ST's at £100 less. Those days the developers didn't have any incentive to push the hardware capabilities of the Amiga

No company pushes the limit of any system until years later,that's not solely an amiga issue.

Originally Posted by mcgeezer

The fun started when the true developers started to unlock the capabilities of the Amiga.

Every system has companies competing,with different skill of programmers,they all push the system later in it's lifespan,programmers learn the system through projects and from other programmers where they're employed,again not solely amiga issue.


Not apples and oranges...Red apples,Green apples

[Shatterhand]

Originally Posted by OmegaMax

No company pushes the limit of any system until years later,that's not solely an amiga issue.



Every system has companies competing,with different skill of programmers,they all push the system later in it's lifespan,programmers learn the system through projects and from other programmers where they're employed,again not solely amiga issue.


Not apples and oranges...Red apples,Green apples

I believe it took longer with the Amiga than with other systems, mainly because of the easy ST ports.

[OmegaMax]

Originally Posted by Shatterhand

I believe it took longer with the Amiga than with other systems, mainly because of the easy ST ports.

And your most likely correct but to say programmers didn't have enough time to learn only amiga and everyone had it better"console programmers" is incorrect.I think I proved that with the facts I posted"the two comments from programmers" instead of opinion.The amiga is difficult to learn in my opinion,coming from tile based systems.The other main fact is,this programmer then landed a job at team17 correct?those guys knew the amiga very well and he likely learned from them"programmers" which would be why assassin was quite good.

[mcgeezer]

Originally Posted by OmegaMax

The amiga is difficult to learn in my opinion,coming from tile based systems.

If you find the Amiga hard to learn you should try dealing with the ST and its stupid GLUE/MMU chips. Programming the Amiga in comparison is/was much easier.

[Galahad/FLT]

Originally Posted by OmegaMax

And your most likely correct but to say programmers didn't have enough time to learn only amiga and everyone had it better"console programmers" is incorrect.I think I proved that with the facts I posted"the two comments from programmers" instead of opinion.The amiga is difficult to learn in my opinion,coming from tile based systems.The other main fact is,this programmer then landed a job at team17 correct?those guys knew the amiga very well and he likely learned from them"programmers" which would be why assassin was quite good.

No, David Broadhurst already independently improved with Dojo Dan which was before Assassin, and Dojo Dan was leagues ahead of his early work.

Team 17 published Assassin, but I very much doubt they had any direct programming input into it.

[OmegaMax]

Originally Posted by mcgeezer

If you find the Amiga hard to learn you should try dealing with the ST and its stupid GLUE/MMU chips. Programming the Amiga in comparison is/was much easier.

I have no Idea,never programmed St

[OmegaMax]

Originally Posted by Galahad/FLT

No, David Broadhurst already independently improved with Dojo Dan which was before Assassin, and Dojo Dan was leagues ahead of his early work.

Team 17 published Assassin, but I very much doubt they had any direct programming input into it.

Even so that was a few years after the amiga release correct?

[Shatterhand]

His work became insanely better after he begun working with the Amiga, instead of the ST.

All his games before Dojo Dan were Atari ST games which were ported to Amiga. That's the main point IMO. Maybe if he had started with the Amiga all along, both Bionic Commando and G'n'G would be good ports.

He also did Bubble Bobble which IMO is a big dissapointment, playing it after coming from the MSX2 version, I couldn't understand how the Amiga version could be worse than the MSX2 one.

[TuRRIcaNEd]

Originally Posted by Shatterhand

Just checked on youtube, the background has buildings and then "shadow" buildings behind scrolling at a different speed of the buildings in front. For me thats "Parallax"

It's been so long that I'd completely forgotten about it - and to be fair, in the YouTube longplay video I could barely see it. I then looked at the ST longplay vid and I noticed the colour bars are significantly brighter. I wonder if that's an example of the coders forgetting to fix the 3-bit ST palette values for the 4-bit Amiga equivalents in just that section?

I have a lot of fond memories of that game - I got it with the Batman Pack (like many kids my age in 1990) and it was the first game I was able to complete without resorting to the cheat mode ("JAMMMMMMMM", remember?).

Originally Posted by OmegaMax

Your making it sound like only amiga programmers weren't given enough time to learn the hardware,console programmers had to learn the hardware as they programmed it also.

I don't think that's what people are saying (and it's certainly not what I'm saying) - rather that publishers used the superficial similarity between the ST and Amiga platforms, along with the fact that the ST was first to be successful in the European 16-bit home computer market, to make the developers write games to the ST's spec and port that to the Amiga almost as an afterthought - which in turn meant that some developers didn't get a chance to get to grips with the Amiga's powerful - but rather unique - hardware arrangement in the course of their work.

Thanks for the link to Chris Shrigley's page - there's a lot of interesting reading there. As far as the console thing goes, I think when people talk about developers knowing the hardware earlier on they're mainly referring to in-house developers at Sega and Nintendo (along with Nintendo's close partners) as opposed to third-party firms such as those Chris was working for. However there's also the fact that Japanese console hardware tended to work along the same principles (i.e. tilemap, sprite and text layers), with the capabilities increasing as the hardware got more powerful. As I said above, the Amiga's hardware arrangement (based around the Copper and Blitter with relatively limited hardware sprite capability) was very different - almost unique in some ways. This could certainly have taken some time to grok - particularly if one were not being paid to do so.

In addition, when it came to arcade conversions, the in-house console development teams would have had direct access to the original coin-op's resources (i.e. code and graphic/audio data). That didn't really start happening with the European home computer conversions until the early '90s (though there were a couple of exceptions - notably Ocean France).

Originally Posted by Galahad/FLT

Developers must literally have been laughing their asses off, being paid to do an ST and Amiga version, and the Amiga version was likely done in a couple of days...

In fairness, I'm not so sure the devs themselves tended to be that cynical - though I don't doubt there were some who considered the Amiga as just another platform they had to work on. I do think that some publishers certainly were that cynical - particularly in the earlier days when most of the money was still in the 8-bit market.

The story of ZZKJ apparently using his pay from the ST version of Super Hang-On to buy himself an Amiga and spend the next several months learning the hardware on his own time - with no guarantee of being paid - is a really cool and interesting tale. Of course, the result of that effort was the markedly improved Amiga conversion of Super Hang-On; which as I recall stood as the undisputed benchmark for 2.5D driving/racing games on the Amiga for a couple of years.

I still maintain the ST did major damage to the Amiga because of the ease of porting ST games virtually unchanged.

That I do agree with. It certainly seemed to take games like Shadow Of The Beast along with the Cinemaware and some EA titles - i.e. those written specifically for the Amiga - for publishers to start realising that shovelware from the ST wasn't going to be tolerated for much longer.

Originally Posted by OmegaMax

No company pushes the limit of any system until years later,that's not solely an amiga issue.

Actually the first part of that statement's not entirely true, and it goes back to what I was saying about the consoles largely sticking to evolutions of the "tilemap, sprite and text layers" formula. To take Sega's consoles of that era as an example, the Mega Drive hardware was somewhere between an evolution of the Master System and a cut-down implementation of their System 16 arcade hardware.

As I said above, this had the advantage of developers being able to use the same principles and techniques they used on older-generation hardware on the newer generation consoles (with more layers, sprites, colours and effects available to them), but it also meant that the hardware was guaranteed to be able to show most of it's potential from the start of the machine's lifecycle. In addition, the similarity of the Mega Drive and Super Nintendo's architecture to the arcade hardware of the mid- to late-'80s meant that the machines could provide a near-arcade experience by design, without the need for some of the lateral thinking necessary to make the Amiga architecture do the same.

From what I've read, the tendency with home computers (which had much more varied architectures) versus consoles was that it took some time to figure out the various tricks which could make the hardware shine. It's certainly the case that enthusiasts were finding novel ways of taking machines like the C64, Amiga, (though it pains me to admit it) Atari ST - and even the humble Spectrum - and making the hardware "dance" in ways the original designers never envisaged; years (and in some cases over a decade) after the platforms were considered obsolete.

Specifically referring to the Amiga, I recently saw YT video of a great presentation by Joe Decuir and Ron Nicholson where they explicitly said that they designed the hardware and architecture in such a manner that they expected developers to find ways of making it do much more than it appeared on paper. Obviously this could be a double-edged sword with the benefit of hindsight.

[Shatterhand]

I spent the last week scratching my head coding a scrolling routine with a tilemap for the Amiga. I ended up with like 3 different routines with different memory/blitter usage (and a particular one that is great for games like Super Mario Bros where you only scroll forward and nowhere else, very fast and using very low memory). The Amiga is interesting because you can try to do stuff in many different ways... the same time with the tilemap/sprites stuff you have basically one way to do things, but this actually make easier to do things since you already know what you have to do from the start.

[OmegaMax]

I've also played with tilemaps for amiga,there are a few ways to handle things I'm not sure what's the best as of yet.

[AnimaInCorpore]

Originally Posted by Galahad/FLT

I still maintain the ST did major damage to the Amiga because of the ease of porting ST games virtually unchanged.

The Atari ST didn't do major damage but minor at most. Damage was done by the limitations of the Amiga system itself. More than 30 years ahead we have no proof of a different view since the Amiga homebrew community wasn't able to deliver like others did on "lower spec" systems.

[AnimaInCorpore]

Originally Posted by mcgeezer

If you find the Amiga hard to learn you should try dealing with the ST and its stupid GLUE/MMU chips. Programming the Amiga in comparison is/was much easier.

This makes no sense at all.