View Single Post
Old 30 January 2015, 05:54   #169

Join Date: Dec 2014
Location: Montreal
Posts: 129
Originally Posted by Galahad/FLT View Post
From my perspective, software companies really didn't do themselves any favours whatsoever.
They 'predicted' that the likes of the Megadrive and SNES would last until 1998 as profitable machines, so put more efforts into those machines..... needless to say, they died quite a bit sooner than that with the advent of the Playstation.
It's a repeating problem. They completely missed the Facebook game wave, then the mobile one, they will soon miss the (small IMO) VR one, etc.
The video games industry is terribly conservative: we don't do any real research, we copy something when it has been proven to work, we would rather issues sequels than take risks developing new IPs and original gameplay.

What surprises me though is how this was true even in the early decades of "home gaming". You would expect a lot of wild innovation and risk taking in the early phases of a new market but what happened was tamer than for other industries. I sometimes think that the high-volume-low-quality producers such as US Gold, EA and such are responsible for killing the risk taking willingness of early developers and forcing them to be more conservative. US Gold alone almost sterilized the market by capturing an enormous portion of the arcade licenses and flooding stores with horribly low quality games. (Sorry Richard, yes Final Fight was an achievement given the constraints but it was also a sub par game because of these constraints.)

On the other hand, I lost count of the amount of times people complained that Sega were no help at all, that their various libraries to write effects and get the Saturn to do whatever it needed to do was woeful, one of the programmers from Gremlin rewrote a particular library for Slipstream 5000, and as a result, the Saturn version was the same as the Playstation version, but he received no help from Sega to achieve it.
That's probably true, but I would be willing to bet that the reason the Saturn was not liked by developers was the same one that they used to develop bad ports for the Amiga: more complex than just "draw these pixels". During the era of assembly coding there was little room for libraries and frameworks flexible enough to accommodate several games, especially given the crazy crunch times. (I entered the industry in 98 and even at that time this was still crazy so imagine how worse the 90s and late 80s were!). So any machine which looked like it required preliminary analysis and proper abstraction of the hardware was frowned upon. Hell, the Saturn is not complex at all if you take time to look at it carefully: the PS2 was way worse but by that time the industry had learned that they needed some code reuse to move forward so abstracting the hardware didn't seem so scary anymore.

This was also the advent of C++ programming. Take a PC title, and in a couple of days/weeks, have something up and running on a Playstation and Saturn, massively reducing development time.
Let me stop you immediately here.
Unless it was a very low resources PC games, fitting in inside a PS1 was almost always a challenge: 2M of RAM, 1 of which for the frame buffer, textures and palettes and very slow IO via the CD meant that enormous work was usually needed to 1) make it run smoothly, 2) fit everything inside without incurring long loading times.
Moreover usually the code was so shitty (even in C/C++) and all modules all coupled together that just isolating the hardware dependent parts would take time.
The adaptation of Heart of Darkness on PS1 is a work of art by itself and took several very intense months of work. If you ever can interview Philippe Paquet (the guy who did it) and ask him to list all the tricks he used, by all means do it, it's both a harrowing and fascinating story.

It's true however that the attractiveness of c++ made the Amiga look a bit more foreign to many big companies but there were plenty of good c/c++ games on the Amiga. Dungeon Master is almost full C for example.
I don't think assembly is useful for the entirety of an Amiga game. As usual, you only need 20% of your code to be optimized: the tight inner loops which are repeated thousands of time per frame. But your main loop and once in a frame code are fine if coded in c/c++.

Before that, it was ASM programming for different processors, different hardware, different musicians needed for different machines.

The Amiga was partly on the industries hitlist because it was never going to work as part of this new ethos of easier development.
This is probably true but only because coders as a group were too stupid to realize and communicate that you could code most of your game in c++ on the Amiga and still have it run fast as long as the code was modular enough. What I mean is that this fact only reflect the lack of technical analysis in the industry, not that the Amiga was not capable of going through that transition.

The pretence that an Amiga game couldn't sell very well was destroyed by the likes of Mortal Kombat, which according to the programmer, he did very handsomely out of MK2.... in 1995.... on the Amiga!!!!

How many years did the likes of EA, Gremlin, Ocean, US Gold take to actually produce a REAL Amiga game?
I cry every time I think of it.

I've just done another ST conversion, it took me from start to finish, 2 HOURS to have the game fully up and running on the Amiga where I could play it, and I don't have the source code.
Experience is a fantastic tool!
Tell me again why you aren't coding *new* games for the Amiga instead of just converting existing ones?

So for me, the biggest problem the Amiga had was always the influence of the Atari ST, because if that machine didn't exist or failed sooner, programmers would have been more keen to utilise the Amiga better and get more from it.
I agree 100%. The ST was a enormous splinter in the foot of the Amiga for half of its career. If only Commodore had not fucked up with the A1000 and released the A500 immediately the sales curves would have been massively different.

Personally speaking, the Amiga had a run of 10 years which not many machines get close to, so it did well, I just think we all realise that had Commodore been a little more savvy, the Amiga in some form could have either still be going as a machine in its own right, or lasted into the millenium.
The Amiga should have been what the Mac is now becoming. And frankly when you look at the evolution of OS X it's very Amiga like in its use of the hardware to do very practical stuff (moving screens, natural trackpad gestures, etc.). The Amiga could have been it 15 years earlier but Commodore lacked the vision of Steve Jobs (among many other things).

And whilst I moan about the Atari ST doing more damage to the Amiga, I think had the Amiga been in Ataris hands, it would probably have been safer, certainly the A1200 would have been more Falcon like with a DSP and that really could have been something.
I don't know about that. The STE is a demonstration that Atari either did not have good enough hardware engineering and/or very bad market understanding, the machine is riddled with original ST issues and the blitter does not even add much to the table given how constraining its use is, moreover it's still less powerful than the A500's!

Great software sells, but if you're phoning your shit in, don't be surprised when people aint buying it.
Absolutely. Software is the great differentiator. Apple proved it when they released the iPhone: 0% hardware innovation, but in a completely integrated software package and good UI -> hotcakes.
Nekoniaow is offline  
Page generated in 0.13330 seconds with 9 queries