Should game devs have to support new hardware?

[Dunny]

Originally Posted by roondar

Maybe, but where do you draw the line? I also mentioned the 68060, which has and still causes issues. Is that not a real Amiga?

For my part, that's as real an Amiga as any other, but if the software wasn't written to specifically support that particular model then that's just tough titties for the user. You can ask WHDLoad folks to try to patch it, or ask for the source.

[apex]

I see there no should. They should run on a stock A500 or A1200, they CAN support more modern hardware like 030 (biggest base), more ram/hd.
For everything else the buyer or player has to wait for a WHDLoad version.

[kamelito]

Isn’t how to code explaining how to code…
http://jvaltane.kapsi.fi/amiga/howtocode/
CBM guidelines http://amigadev.elowar.com/read/ADCD.../node000B.html

I kinda understand that 8-bit are easier to develop for and usually the 8-bit guys are nicer.

[grond]

Originally Posted by Minuous

If something like a DraCo or Vampire is incompatible with a real Amiga the fault is with those, surely.

This is not true for most perceived incompatibilities. Most of those are just the programmer assuming the blitter is done with its job while the CPU is busy doing other stuff and start the next blitjob without checking. If the CPU is very fast (060 or 080), this assumption will fail. This is just bad programming and certainly no fault on the side of the hardware. There are lots of similar bad coding techniques such as unclean self-modifying code which will break on some Amiga configurations which are not caused by the hardware but by a bad programmer.

And it is true that the original Amiga documentation explained all this in detail which underlines that it's the programmer's fault and not the hardware's.

[mcgeezer]

Originally Posted by grond

This is not true for most perceived incompatibilities. Most of those are just the programmer assuming the blitter is done with its job while the CPU is busy doing other stuff and start the next blitjob without checking. If the CPU is very fast (060 or 080), this assumption will fail. This is just bad programming and certainly no fault on the side of the hardware. There are lots of similar bad coding techniques such as unclean self-modifying code which will break on some Amiga configurations which are not caused by the hardware but by a bad programmer.

And it is true that the original Amiga documentation explained all this in detail which underlines that it's the programmer's fault and not the hardware's.

I'm sorry but this is exactly my point about poll option #3. Simply saying it's the programmers fault/bad programming is just a total cop out because the odds are not stacked in their favour.

Loads of games from back in the day probably used things like self modifying code or dumb CPU based timing delays, WinUAE has a specific option to force waits on blits so it must have been prevalent in games. If it worked on their targets configs then it was job done... no complaints, only later did this become a problem and the likely rise of WHD.

Are all those guys who made those games back in the day bad programmers? Is it their fault they couldn't test their code on all manner of different configurations? This is the same problem that exists today except that we have better access to information and tools but it still isn't perfect.

The hardware has some way to go and meet the programmers half way I think, having a way to programmatically disable their accelerators is one way it could be done.

[Daedalus]

Indeed, a missing option that I too would go for is that devs can develop for whatever target platform they wish. If they want to use tricks or techniques that means they limit themselves to a subset of Amiga users who only have stock machines, then that's their prerogative. So long as that's clearly documented in order to set expectations, there shouldn't be any issue coming from users who find it doesn't work. It works fine the other direction - if you say your software requires AGA and 4MB fast RAM then nobody can complain when it doesn't work on their A500. So do it the other way around - say it doesn't work on accelerated machines, and leave it up to the user to remove or disable the expansion, or use an emulator, pass it by or what not.

More and more users these days are using at least extra fast RAM and hard drive setups, many have faster CPUs too. Personally I wouldn't feel right excluding that use case as an option, but when you're developing primarily for yourself, all that matters really is that you're happy with it.

[roondar]

I do feel that some of the named issues aren't very prevalent any more these days, though. Almost all of the recently released games for the Amiga I've seen already seem to do quite a lot to keep compatibility on faster machines despite often being designed for entry level systems:

• Pretty much all of the recent games use proper Blitter waiting. Heck, I'm struggling to remember even a single released Amiga game over the past few years that didn't
• I've not even heard of recent games that used self-modifying code but didn't disable caches
• OS handling (de-activation/re-activation) seems to be done well in almost all these new releases
• Interrupt handling almost always just works on faster Amiga's, avoiding the A4000 incompatibilities
• Timing is virtually always done via either VBL/raster or via CIA
• VBR is usually handled correctly, as is move from SR
• Many games offer improvements to those running them on faster HW

In fact, the vast majority of issues I've personally heard about recently are exactly the kind I've been talking about in this thread: edge cases that exist only on exotic combinations of HW and SW. Either because of weird emulator setups, heavily patched OS configs, or expansion hardware that doesn't behave as expected, etc. Note: I'm not blaming the HW designers for making new stuff that may not be 100% compatible with what already exists. But on the other hand, blaming SW designers for issues that are in fact often created due to the HW/SW combination of the end user, rather than their coding prowess doesn't feel right to me either.

Like I tried to point out before, I feel this has little to do with poor coding and everything to do with real-world difficulties in getting compatibility across a large section of different configurations. Difficulties that everyone who's involved with them struggles with - all the way up to today's giant IT corporations.

The position that it's all just about poor programming IMHO doesn't do justice to the actual issue here.

[grond]

Originally Posted by mcgeezer

I'm sorry but this is exactly my point about poll option #3. Simply saying it's the programmers fault/bad programming is just a total cop out because the odds are not stacked in their favour.

No, it is not. Those programmers contravened Commodore's official documentation on how to program the Amiga and, more importantly, on how NOT to program the Amiga. Using your arguments you could just as well code your program to jump into fixed OS routine addresses because, well, everything after Kickstart 1.2 is just incompatible and not the programmer's fault (which early Amiga programmers did do although this was obviously also contravening Commodore's programming guidelines).

As for your point about disabling accelerators: how is an 040 or an 060 being the only CPU in an A3000 or A4000 going to deal with the broken code?

[grond]

Originally Posted by roondar

• Pretty much all of the recent games use proper Blitter waiting. Heck, I'm struggling to remember even a single released Amiga game over the past few years that didn't
• I've not even heard of recent games that used self-modifying code but didn't disable caches
• OS handling (de-activation/re-activation) seems to be done well in almost all these new releases
• Interrupt handling almost always just works on faster Amiga's, avoiding the A4000 incompatibilities
• Timing is virtually always done via either VBL/raster or via CIA
• VBR is usually handled correctly, as is move from SR
• Many games offer improvements to those running them on faster HW

If all of these points are met, then it is probably a hardware issue or some strange OS incompatibility and the programmer is likely not at fault. What is being discussed here seem to be much greater flaws of software. Of course, as long as no money is involved, anyone can do just anything they want (perhaps not viruses...) but I will still consider them bad programmers if they do not take care of what you listed above without need. Need may be if you really, really absolutely need the very last clock cycle to achieve what you are doing e.g. in an impressive stock OCS demo. Most often this is not the case.

[mcgeezer]

Originally Posted by grond

No, it is not. Those programmers contravened Commodore's official documentation on how to program the Amiga and, more importantly, on how NOT to program the Amiga. Using your arguments you could just as well code your program to jump into fixed OS routine addresses because, well, everything after Kickstart 1.2 is just incompatible and not the programmer's fault (which early Amiga programmers did do although this was obviously also contravening Commodore's programming guidelines).

So because they didn't follow Commodore's official documentation on making games it makes them shite programmers does it? My last look at Commodores documentation on how to make games was to do everything through the OS... and that was never going to happen.

Originally Posted by grond

As for your point about disabling accelerators: how is an 040 or an 060 being the only CPU in an A3000 or A4000 going to deal with the broken code?

A3000 or A4000 Amiga's were never mainstream (at least where I lived)...but besides, I accept that at least for my productions those stock configurations are tested (in WINUAE).

[mcgeezer]

Originally Posted by grond

[/LIST] but I will still consider them bad programmers if they do not take care of what you listed above .

I don't call them bad programmers, I call them inexperienced and there's a massive difference.

[roondar]

That does sound reasonable, yes. We all started somewhere, doesn't make us bad per se

[Reynolds]

One thing is that in the heydays the A500 was the most common config. Other is, we're not in the heydays anymore and the possibilities are vastly extended. Ah yes, and third aspect is the development process. I mean noone is chained down to a basic Amiga setup to create state-of-the-art games, demos or whatever, but most stuffs are coming to see the light of day on modern pcs. So targetting a base A500 (with 1 megz of ram ) is a nice challenge, but as storage devs are cheap as dimes nowadays, at least this should be taken into consideration, but doesn't happen. At least extended versions could come out for high(er) spec machines' owners. While more and more users are expanding their configs with 030+ CPUs and other bits & pieces, they have to accept that these extra resources will not be supported. Even nostalgic factor is forming strongly by using chiptunes in fresh Amiga games... but come on, we already had the chance to enjoy Project-X, Agony, Pinball Dreams and other amazing games with proper audio - why should we bother ourselves with chiuptunes? those should be left for C64 and Spectrum kind of machines where it really belongs to. Even on a plain A500 with hdd music could be even streamed or rendered anims could be used... but I know I'm just dreaming... but hey, if we consider to develop and buy new stuff, it should deliver quality, even if we are talking on hobby projects. And one more thing.... Classic 68K, OS4, MOS, AROS... but why should we leave PowerUp/WarpUp? We have more options still available in parallel to target, but have to stick to the A500? Can't see the point. We should take the advantages.
Saying that a developer can't be forced or requested to test on most/every hw available, is kinda nonsense. Many of us have various configs and can offer them to make tests about practically everything. If a developer afraids of his/her stuff getting stolen or leaked, any limitations he/she could include the program gave for tests, to cease it's inappropriate spread and/or theft.
I have dozens of hw and always when I see a new project, and see any chance to offer my service, I do that, because I had the time and money to buy the hw, now as I can't code, I can only support developers with betatesting and buying games.

Maybe this last thought from me will be kinda opposite, but one thing also valid: IF a developer plans to make something on any platform, it is his/her decision. Like an artist, when does somewhat extraordinary because of the way he/she sees some feels to express him/herself. His/her right is to do what & how he/she want. but then don't be surprised if the audience won't be satisfied due to different thinking.

[Reynolds]

WHD should have to be left for the old games and demos with trackloading or other "cheating" of the system but not stuff created nowadays.

[phx]

Hmm. Maybe I misunderstood this thread. The question is about supporting new hardware, not all classic Amiga hardware? This depends on the hardware, I would say.

A new 030 or 060 accelerator board should be supported automatically. There is no excuse as there is no difference to other accelerators from the software's point of view.

But imagine somebody builds new hardware, calling it Amiga-compatible, but it isn't, because it requires some specific initialisations to make it work. Then the hardware is at fault, because it is not really Amiga-compatible.

I mostly agree what Daedalus said. Make a list of the hardware your software is tested on, or intended to support. If the list is very short and you want to sell your product you might get some questions, or less sells, but that's all. So there is not really a problem.

[Foebane]

Originally Posted by mcgeezer

Why do they have to be compatible with faster hardware though? Please elaborate.

Optional. As an example, Frontier: Elite 2 needs a beefy processor to run smooth, not an A500 or an A1200, even though most played it on that.

[Amiga1992]

Originally Posted by mcgeezer

My opinion is that if the Amiga scene wants to attract new developers then the devs should not have to test/support all of the different boards that people have in real Amiga's these days.

Most accelerators were a terrible idea to begin with, stemming from a futile urge to "keep up" with technology that was clearly surpassing the Amiga come the 90s and the user base's fanatical loyalty that would not let go of the platform and do whatever if it was even to create a dim ray of hope of the Amiga remaining alive and "relevant".

I mean, I understand some professional applications benefited from them since the beginning of the platform's existence, but to *need* them for every day use or, even worse, for gaming, was absolutely stupid and to this day a fanatic push to take the Amiga where it, in my opinion, should NOT go.

The thing that makes other scene productions often more impressive, and also more attractive especially to new coders, is what mcgeezer said: the platform is fixed, and it gets milked. This is what you have and you have to make do with it. From a creative standpoint, it usually pushes people to think differently and come up with far more creative ideas to resolve problems with the limited capabilities of the platform.

Also, it simplifies things a lot to know that you don't have 1789312 target configurations to support when you deliver your product. A C64 game will basically run on any C64 that exists out there, the exceptions are few. Meanwhile the Amiga platform is so fragmented, that if needing to cater for all these myriad configs is a brick wall in people creating software, I would very much be glad if people stopped caring about these things.
Another problem is that not one accelerator works like another, it's not like all you have to do is "support 680x0", it is well known that there are different (in)compatibilities among them. Personally, I have a 030 that really gives me a lot of headaches on my 1200.

Not comparing to other platforms, look back at the Amiga demoscene in the past, at least, 5 years (probably more), where the focus seems to have re-centered in making OCS/ECS productions, and most times, these new productions for even older hardware, are far more impressive, interesting and innovative than anything that targets accelerated hardware that only very few own (looking at you 060 only, 3D flyby demo crew).

tl;dr, my vote is, and has always been, for targeting stock hardware. If doing something really tied to that stock hardware breaks compatibility with accelerators, so be it. Take that shit off the trapdoor and enjoy a good Amiga game in the 21st century.
You can always stick it back any time to play your favorite game, SysInfo

[Reynolds]

I disagree. Of course the accelerators works slightly different from each other, but that can't be the reason. For example the Apollo 1260 is a terrible card if you want to use it with a properly installed system. Except, if you install OS3.9 as that runs smooth as silk from the beginning of the install, while it is a real pain if you want to build a WB manually bypassing the OS3.9 disliked by many. So everything depends on the will of making something to work equally on any accelerated systems. Just check the late commercial releases. No CD games came out which refuse to work on any dull and dumb base OS on any CPU, and decades ago also there was dozens of accelerators and other expansions. But hey, even now there is game under development to run from floppy... come on, seriously... Noone could say that hooray, this is the way for us, turning back in evolution.

[roondar]

Amiga's of any sort are not up to date hardware though, so that argument doesn't really work for me.

[mcgeezer]

Originally Posted by Reynolds

But hey, even now there is game under development to run from floppy... come on, seriously... Noone could say that hooray, this is the way for us, turning back in evolution.

You know why they run from floppy?
Because it's easier for the programmer to implement their game on without complaints.


Does anyone seriously do productivity (*) work on an Amiga these days? Surely not.


* Web Browsing, Word Processing, Accounting, Presentations, Email, Digital Imagine, Video Processing, Cloud Computing.

Edit... hang on...

I'm now starting to wonder what people do actually do with the extra speed in their Amiga systems.