Originally Posted by eXeler0
The Dreamcast is one unified platform.. Amiga OTOH is spread very thin across a whole crapload of different configurations, resulting in very varying performance. And even the OS was branched off into 68k Classic, Morph OS, OS4 PPC, AROS etc..
So no wonder that any given config is going to have precious few followers.
If you sort models by sales then there is much less fragmentation.
The original user base is mostly composed of A500 models followed by A1200 ones and this is likely to be reflected in the current user base. With these two models you are guaranteed to reach a very large fraction of the Amiga owners (not to mention that the A2000 is just an expandable A500). Of course, the passionate ones will tend to own more high end models but given the high second hand price of these it's reasonable to think that they represent a sliver of the total.
With two models to target remains the choice of the configuration to support. There too, looking at the requirements of softwares sold for these machines one gets a pretty good idea of what to support: 1MB chip for the A500 and likely stock for the A1200 (accelerators costed one genital and a leg at that time).
With this selected I'm fairly confident that one maximizes their potential public.
Once the market gets some traction and if/when programs start reaching the limits of the machines then it's always possible to up the requirements and require some Fast RAM. But anything more than that would lead to fragmentation again so it's probably a good idea to keep concentrated on barely expanded stock models. (1MB chip, storage and network.)
So if you were an indie developer, what would you target? Ok, lets go for 68k Classic OS. Next choice: Minimum CPU requirement? Minimum RAM? RTG only or AGA?
I am a developer and following my above reasoning I am currently targeting the stock A500 + 1MB. In the future I would love to release two versions of each game I write: one O/ECS and one AGA but for now I'm concentrating on finishing the first one.
If I were to realize that a particular game would require Fast RAM to run efficiently I would probably try to bundle it with modern RAM expansions but frankly if that would ever happen this would be years in the future so this is not my concern at the moment.
Something I would be considering however would be to put the game on cartridges since these are fairly cheap to produce these days and this would work around the problem of aging floppy disks and drives. (As well as loading times which newcomers would find hard to tolerate.)
Amiga has -by modern standards- very little in the way of Dev -friendly tools that would make your life easy.. And even if they were, the results would probably not be very impressive.. because the most impressive stuff was always done in 68k asm. ;-)
This can and (I think) will change. It's in the air.
Also Dungeon Master is 100% C code and SkidMarks is 90% Blitz Basic.
Assembly was the most used because of the inertia of coding practices of coders from the 8bit era where this was the only choice. It is possible to make a fast Amiga game with 90%C/10% assembly. And anyway 68k assembly is really not that hard and quite fun to use.
A year or so ago, there was a campaign for the AmiDark engine (which I supported at the time) but the campaign failed, and it got me thinking... How come that there are so many ppl out there that still wanna pay £300-500 for 20 year old Blizzard cards, £100 for a single 060 chip, etc.. but when it comes to supporting software.. nope...
Why do people pirate software but do not steal games at the store? The cost of the physical product in stores is not even 1/10th of the sale price so this barely generates a loss for the vendor.
Because our monkey brains have evolved to give more value to the physical than the intangible. We build sophisticated arguments about a copy not generating a loss but the fact that more expensive console games sold more disproves this reasoning: when copy is impossible, even less wealthy people buy.
MegaDrive/Genesis owners were a much poorer demographics than Amiga owners, this console's games costed two to three times the price of an Amiga game, yet the same game sold 5 to 10 times more on the console than on the Amiga.
I had to literally train myself to not use cracked software for almost one year before it became natural to me to pay for software. Unless one has voluntarily opted to respect these intangible rights one will revert to our default nature which is to give less value to non physical property. That's how we are wired.
Also we tend to promise ourselves "I'm testing it and if I like it I'll buy it" but do not follow through afterward.
(That's part of what is called the discounting effect in scientific terms: we always overestimate the amount of self control we are capable of in the future.)