Was running in front of a rocket only possible with pentiums?

Sopwith (see Fraggle's avatar) was the first game where I noticed the effect, going from an antiquated Amstrad PC1512 to a entry-level 386DX. You ran the game and it was game over five seconds later as your plane took off on its own (waited too long I guess) and crashed immediately after.

Oh hey. Quake is also affected. Good thing that DOSBox kinda limits it. But it was ridiculous under high-end Windows 98 machines.

I remember playing around with this years ago on Odyssey of Noises with the -turbo parameter. I was surprised that my projectiles passed through me, but I knew next to nothing about the technical aspects of Doom back then. Still, it was fun to race my rockets and watch a barrage of my own plasma coming at me.

printz said:

Oh hey. Quake is also affected. Good thing that DOSBox kinda limits it. But it was ridiculous under high-end Windows 98 machines.

Define high-end?

printz said:

Oh hey. Quake is also affected. Good thing that DOSBox kinda limits it. But it was ridiculous under high-end Windows 98 machines.

Wait. You mean that Quake, a late DOS-era game, had no adequate CPU speed compensation? Or that timing was all over the place if running under Windows 98? Under "pure" DOS or in a window?

Those old CGA era games were easily affected because they relied on increasing a delay loop based on an internal benchmarch vs a "known performance" CPU. This method wasn't very future proof, either resulting in too little delay (hitting a delay ceiling) or too much (an infamous case: the Bubble Bobble conversion by Novalogic, which was unplayable on anything higher than a 486 DX/33, giver or take).

But Doom uses the system's PIO interrupt chips to determine timing and uses only a single wait loop while waiting to process the next tic (which was usually skipped in most machines, heh).