Scawen the dynamic lighting thread got locked, and you don't take PMs, so I'll suggest here: https://twitter.com/inovae_keith has been working with a small team to make a game which has some of the best atmospheric rendering techniques. He periodically retweets papers and tweets on the subject.
I dunno if you would gain something from those different more high-end implementations, but maybe something applicable could come from sifting thru that collection. IIRC many of them are first principles analyses not yet actually translated to code.
You may have to go back a ways in his tweet history, as they're now pretty far along in the game's pipeline.
All it means is that you cannot see it, not that it won't happen. They're two different things.
The car and track suggestions thread do not have to cause all of those specific cars and tracks to be added, to have been worthwhile. Maybe you can chew on that yourself to see how that makes sense.
That's a pretty pretentious assertion. Unless I'm misreading it. I really doubt you could walk up to Scavier at work and seriously tell them they're not serious about it. They've been nothing if not serious about their work style choices and about LFS itself.
I reckon you have some perceptual bias that would not be solved without some serious intervention, e.g. shooting the shit in earnest about work philosophy with Scavier over beers.
Even if that were how it turned out, saying something like that, this far ahead of time and in spite of everything Scavier have said and done, would have been a serious lack of deference to the people actually working on getting the game done.
You may disagree with them but you do not get to disrespect them.
All that said, I will concede that I have almost exactly the same type of work style preference. It makes things absolutely 1000x better, for me and for the quality of work. And the company I started on those principles is still working and attracts more of the same kind of people who appreciate that work ethic. Win-win.
That is so the best icebreaker ever, should one meet Scavier in person.
/s; emph mine
Actually that thread could be restarted by someone who's willing to stick around once a month or something, to keep it up to date. That way any new/returning readers have a TLDR right off the bat. I'd be willing to do it (and stick around for 10 more years or whatever) if none of the mods/devs object.
It'd have 1 line per car and ≥ 1 link -per to a pic and/or wiki type page.
Time to get a new crystal ball. Real world motorsports never underestimate the importance of tires. LFS is no different. The real litmus test will be once the rubber revision is done -- how much or little updating LFS gets afterwards.
[ ... ]
Last edited by Scawen, .
Reason : removed part of the post that was probably just quoting a typo
What I did back then was entirely legal, using Tweak. That's part of why it seems so feasibly to crowd source a rough first batch of new car candidates. The impact on cars performance from their VOB-dependent physics is AFAICT marginal. The only major obstacle I recall was not being able to change the suspension types. E.G. the XFG's crappy rear end.
That's the point. That regardless if he's said something or even decided anything WRT to it, his word will trump yours.
Which could just as well be a distinction without a difference: the modded cars are no less cars than official cars are cars. Quality cars will out-compete crappy cars. It is not comparable to expecting that LFS could not survive tools enabling or helping free roaming and/or drifting because that would trigger a landslide player migration to those modes and completely starve conventional racing. It's not anything so black and white, with a fair share of all server types getting players, with the higher quality servers regardless their type being most popular.
That's not an inherent consequence as you imply it is, and aren't they demo because that's all Tweak can edit - the unencrypted cars?
Taking an unproven assumption for granted, and more evidence that for you having the last word trumps everything else, so really I think this discussion is done. Man, I forgot how so many in motorsports "communities" are such ...
IMO that's the real problem during these long stretches without updates.
Anyway... A invite-only and/or NDA (by Scavier) pool of car builders using tweak like programs could cover the "proliferation" issue. It could additionally have some kind of server-side baking of user-submitted "tweak" parameters. But that really reduces the breadth and depth of user-made permutations. Ultimately that's one of the major missing parts of LFS - way too little volume and variety in the carset.
We could hash out a lot of the basic playtesting using "crappy" tweak approximations... Then Scavier pick out whatever car formulas they think best fit into the carset.
The bodywork is mostly independent of the stuff available to Tweak apps, right? IIRC the rims are also part of the physics. But AIUI the difference in physics that a different car model makes, is in the later part of design. IE fine tuning.
Right now, and this is an assumption, the LFS car classes have plenty of rooms for more varied new cars. I'm pretty sure of it because I did it myself. Added a few cars to the LRF class and they were even enough that lap times were competitive over a few tracks (IIRC we'd tested them on SO4, Blackwood, and Fern Bay black). Changed the XFG into an AWD turbo kit car like in amateur rally, the XRG into a big Muscle/Corvette-ish (IIRC because an exact Corvette replica was too fast), and the XRT into an F430 replica. The intent was to produce cars that were
2) As different as possible from what LFS already had
And as a proof of concept, it worked IMHO.
So maybe it ultimately (when Scavier and/or all players evaluate it) wouldn't work well enough , but I made it work that well on my own. If I can do that, and the idea isn't wrong (I've seen no reason to think so) 10-50 players could definitely do it well enough. And of course those specific cars were just examples. It is just a proof of concept.
With all due respect I'm going to take Scawen's word over yours.
You can read my mind! Haha!
How do you not see that that's what I described. If there's a few things that suck about LFS, one of them is this sort of teenage level pseudo conversation. Having the last word completely trumps actually getting to the bottom of things.
If you can't see how what we have can be used to get to what I described, then I dunno. I'm not gonna connect the dots for you.
It's a straw man because you invent this loaded premise of LFS modding tools that completely open up game-killing vulnerabilities, and then pretend that that's the only way LFS could possibly be programmed. That you can't imagine something, is not evidence that it can't be done.
Why isn't LFS server list full of Tweak servers, Right Now?
...And there it is. Good old snide rhetoric. Houdini would be proud. LOL!
I don't see why user-made and official content are mutually exclusive.
We don't need much quality content. Even half as much user content as official content would be a very good situation. The main problem (or one of them) with user content is player dilution by the sheer variety of content. But if (e.g.) Scavier restrict things a bit with only new cars that fit within the official car classes, things quickly are more manageable. So user-content isn't inherently hopeless as you and others seem to argue.
I do think there'll be enough who both care and know about quality, for the best user-made content to be really good. For instance I edited an LFS car to as close as possible to a Lamborghini as LFS Tweak allowed, back then, and Flotch fairly quickly made a really good base setup. Now, I'm not saying that ripping off real world car designs so blindly is ideal, but IMHO it does show that
1. There are players around who know quality and care about it enough to produce high quality content
2. LFS is a good enough platform that it's possible to so readily make new and fun cars
3. LFS carset has so many holes in it that we're nowhere near the point of diminishing returns
Another example: Bob Smith's NASCAR mod. Again the underlying platform that is LFS shows how flexible and potent it is.
You might say tried and tested. New physics and crowd(ish)-Q.A.'d new cars to add variety to each car class on their own would IMO really make LFS hard to resist at least trying again. Even if I'm partial to LFS, I think that's a pretty fair assessment.
I don't understand the part in bold. It seems like a contradiction.
I think some intermediate formula would work. A way for player-made content to be added after a long time, is better than no way at all. E.G. player-made content submitted to some more or less gated QA system, where (e.g.) submissions are filtered up thru increasingly strict evaluation, with Scavier at the top and a larger group of testers at the bottom of the pyramid. It doesn't have to be a pyramid; the point is only that even a very long wait to approve only the highest quality submissions would be better than nothing at all.
But something tells me there's some reason that even that's not an option to Scavier, so far.