Testing to race distance, and then some, is always super important if you intend to do a no-stop strategy. You find out how much margin you have before tyres pop.
Oooh. If I may inquire about specific examples, does this include the flat-out (in the FO8) chicane on FE2's front straight, and KY3's transition between the oval that necessitated Boothy's chicane? Maybe AS3's final chicane, too?
And this absolves the game and developers of all criticism? Are we to simply take the developers' claims of historical accuracy at face value? What makes you think the reviewer didn't play the game?
Is the exclusion of people of colour not a historical cleansing in a way that suits their white supremacist, far right agenda? This is to say nothing of the other points RPS makes about class and gender.
Doesn't that sound just a bit like a fascist ideal?
This is only half correct. People often leave out the first part, which then becomes:
Perfect practice makes perfect.
If you're pushing at 100% all the time and sliding everywhere with little control, then you're practicing sliding everywhere with little control.
Take a slow car (XFG/XRG) and take it slow, get the car under control, keep your inputs smooth, figure out exactly where and when your reference points are. Once you get some consistency, then you can start increasing your pace/aggression a little. If you start sliding everywhere again, you increased it too much.
Pure aggression will get you to a certain point, but no further. You need to use your brain too.
If there is no option for a no-stop strategy due to insufficient tyre life and/or fuel tank capacity for race distance, it's an endurance race.
If a no-stop strategy is the always the fastest option, it's a sprint race.
There may be some vague middle ground between the two
You might say that a race where 1-2 stops are viable is called a grand prix, and anything with 3+ stops is an endurance race, but I feel this distinction is unnecessary. Alternatively, a race where one driver per car is practical and viable is a grand prix, and a race with viable/necessary driver swaps is an endurance race.
I think this is the series you're looking for:
5 different riders winning (9 last season).
3 different manufacturers winning (4 last season, but only just).
No push-to-pass.
Long story short, ignore anyone who talks about wheel speed rather than torque when talking about clutch-pack LSD. Clutch packs are torque-sensitive. The one in LFS is modeled after a Salisbury LSD. Look up something called Torque Bias Ratio.
The viscous LSD is speed-sensitive.
What is the result you want? Why not try experimenting?
I think there's a special level of hell where you have to drive a damaged car for endless laps.
I got caught up in a couple of other people's crashes early on, but fortunately they only cost me time, not damage. I had a couple of my own crashes later in the race, neither of which really cost me much time directly, but did cause a fair bit of damage to my car. That chicane is soooo treacherous, and the consequences for being just a little off are huge.
Just by virtue of the way LFS works, with its 100 Hz physics loop rate, 100 fps is actually smoother than 120 fps on a 100Hz monitor. You really want your framerate to be a multiple of 100 at all times, and 100 is the easiest to maintain.
It appears v-sync is to be avoided. With v-sync on, I see some frame stutter every 5 seconds or so, represented in the physics graph as a series of alternating missed and duplicated frames lasting a half second.
After playing with the settings it appears the following are optimal for achieving a solid line in the physics graph with a 100Hz refresh rate and minimal input lag:
- v-sync off
- frame rate limit on
- max frame rate 100 fps
- sleep every frame off
- max buffered frames 0
If I have a monitor that runs at 100Hz refresh rate, and knowing that LFS's physics loop runs at 100Hz, is there a way to ensure that each monitor refresh cycle displays a completely unique frame (i.e. no dropped or duplicated frames)? In other words, I want what I see on the screen to be in lock step with the physics loop.
Does enabling v-sync accomplish this? How does this impact input lag, if at all?
I would measure this myself but I lack the means to do so. Has anyone else performed such an analysis? I thought Scawen might be able to provide some insight as well.