Scali wrote:
Yes, but you'd obviously pick the high-end to cutting-edge parts. This is gaming we're talking about. I could even buy an 8088 machine today, but that's hardly relevant, is it?
Relevancy to the topic at hand? Yes, it is. Regarding the topic, "Objectively fast" over "Properly dated"?, it is missing a 3rd option: "Widest scope?", which is taking the intrinsic benefit of the PC platform itself, i.e. making the most of backwards compatibility over the widest timeframe.
For example, if my requirements are to have as few builds as possible to run as many games as possible using real hardware from 8088-class to fastest possible, then I can build such a system. I can do this right now from an 8088 to Pentium II-class and everything in-between with one build.
Scali wrote:
Why are you trying to overcomplicate things? Hardware upgrades came in waves, and so did the games and their system requirements.
I mean, new CPUs would only get introduced once every 3-5 years.
Also, for some games it's more relevant than for others. Some games aren't very speed-sensitive and/or work exactly the same on a wide variety of hardware.
I get that you are simplyfying 'Period Correct' builds by choosing an era and speccing parts for that era based around game system requirements. I'm not trying to overcomplicate things - in fact, with hindsight, I can simplify builds to the fewest possible over the widest timeline. So rather than applying restrictions to builds (Period Correct), this is celebrating the flexibility of the PC.
Scali wrote:
If we're talking about my proposal of a kind of 'virtual museum' that displays all games as the developers meant them to be played, then the answer is: as many as it takes. Which is probably not that many. I guess 4 to 5 machines would cover the DOS era at least (the one I'm interested in). With another 4-5 you could probably cover the Windows era as well, since Windows is not that sensitive about speed or exact hardware features. Eg, pick the best videocard for every era, and build a machine around it. A VooDoo machine, a TNT2 machine, a GeForce3 machine, a Radeon 9700 machine, a GeForce 8800 machine and by then you've pretty much arrived at 'modern' machines.
I see how you are differentiating the Windows builds with video cards, but what about the DOS builds? CGA, EGA, and a couple of VGA builds?
Scali wrote:
j^aws wrote:But if your build covers those two reasons, then build whatever works.
I already explained the reasons, so I don't see why you're still debating this.
I don't see your point: "The point is just that you don't go overboard."
What is overboard? Going past the restriction imposed by being 'Period Correct' and/ or past the recommended guideline suggested by the developer'? I understand your point for your 'museum', and in that context I get it. I also get it for playing safe for compatiblity. I've gone beyond developer guidelines and found games to run better on many occasions, so I'll proceed to ignore said developer guideline as limited to their timeframe. If the game works though, so what if I go overboard?
Scali wrote:It could be virtual. I bet between all the users of Vogons, we've more than got every interesting configuration of real hardware […]
Show full quote
It could be virtual. I bet between all the users of Vogons, we've more than got every interesting configuration of real hardware for all games ever made covered. People could just make YouTube videos of the games playing on real hardware.
Ever seen the YouTube channel of TheShadowsNose? He records the same game running on various platforms, using real hardware. His videos give very good insight in how the games really play, look and sound on real hardware. For example his video on Titus the Fox: https://youtu.be/hqgsUemLi84
His recording really shows off just how smooth the Amiga's scrolling is, and how the Atari ST isn't. Emulators on PC generally don't get the scrolling perfectly smooth, so if you play it in an Amiga emulator, you may get an experience closer to the Atari ST, without knowing just how good the game really is.
Also, he records from a real CRT, so you see the graphics as they were designed, rather than the blocky 'perfect' output from most emulators, that don't quite do the original graphics justice.
That's the sort of stuff I'd want for PCs as well.
I agree emulators are not perfect, and a 'virtual museum' of sorts via YT is another way to get insight. Neither can replace real hardware, especially with tactile feedback missing regarding specialised control systems (e.g. Arcade) - you won't get that from YT, but you can get a sense from emulators using similar control systems.
I can detect uneven scrolling/ panning easily, and that YT example of scolling reminds me of early home computers using flip-screen scrolling. Other things I can detect are laggy inputs, missing low res scanlines etc... I'm sure VOGONS users could contribute via YT (some already have) and highlight their insights, too. I'm struggling with time to build and test, nevermind play and post.