do a lot of 0.1s segments to keep them busy while you move past them
doesnt host_framerate .001 really slow down npcs?
yes but that CVAR isn’t legit in a speedrun so you gotta raise your fps with fps_max, it does the same thing but harder to achieve high fps
Except we’re not talking about realtime speedruns right now. Thing is, grunts’ shots are random, and their movement can also be random. Although if it takes them a lot of time to rotate, there may be not enough time for them to actually shoot.
Artjom TJ’s Power Up Speedrun HD Test
Power Up TAS in 50 sec
We did a TAS using the intended route and we got 50.025 seconds. Coincidentally, this run also took 50s.
We used Zezima’s run as a reference from time to time.
This took us a few days to lay down the plan and about a week to script. Very often we only had some vague idea about how some of tricks were to be carried out. We did not know exactly how to bring the crate out of the turret room until we reached that part, for example. We didn’t even know if we can dodge all bullets from the turret.
[hr]
Now, you may object to us climbing over the obstacles, but what mattered is the fact that the obstacles were smashed by the tram before we entered the next chapter, as verified by setting scr_ofsy -1000 at the end. If you really dislike this, just add 1-2 seconds to the run time.
Absolutely amazing. I hope you intend to run the full game.
What would be the estimated time for a full TAS of Half-Life? I find it hard to picture something more optimized than HL21, but this is just insane.