Major shock indeed.
Back when he was in Sonicrealms, I remember him teaching me how to take that obnoxious fast chicane at WE1 flat, in some car or other. Amazingly fast driver...
RIP

Aha...

Someone called elsocom just joined the AWS server (and hit restart while I was on a lap, ffs :-/), but isn't signed up. Guess somebody has been sharing the password or poss there's a hack? (The racer in question has v v low online mileage on the stats.)

Quote from ArnoldSchwarzenegger :

My first S2 race!
I think for my first race I did pretty good.

Yes, well done

I was seriously lucky that I was running Aonio (little brother of Airio) as otherwise I would have run out of fuel
Dunno how, but 46% wasn't enough for my race (and it's not as if I was lapping at WR pace ). I noticed at the end of lap 21ish (Aonio told me I had 5.8 laps of fuel left, not 6.0, uhoh!), drove laps 23/24 rather slowly, and dropped around 4 seconds, so Falcon probably thought he was gonna catch me
Had saved about enough by then so was able to floor it for the last 3 laps and finally ran dry about 2 seconds before the finish line
Nice clean racing!

Quote from Specht77 :

On older patches it had the animation to looking behind, since it was 0º - 90º - 180º, not 0º - 45º - 90º like it's today, but since it didn't had the 45º, it was also only 5 stages:
http://www.youtube.com/watch?v=G9meSjjwNFM
(the first is patch X, the second is patch E)

Nice vid
However, it looks to me that the first one shows 0, +/-45, +/-90º, and the second shows roughly 0, +/-30, +/-60º.
I'm old enough for 90º to hurt so the second is more realistic for me
Neither video goes anywhere near 180º (looking backwards) so I'm possibly confused about what you mean.

In any case, 8 bits is indeed waaayyy more than we need for horizontal movement - I agree that 4 would be enough (that allows 16 positions, e.g. 7 left, 7 right, centre), and indeed the rendering could easily smooth the transitions between them at display-time. Vertical motion might be worth adding, not sure about that.

I doubt that Scawen will see this as a priority though

Sharing head movements - interesting idea.
I was initially thinking that doing it with enough time resolution to be smooth would be pretty difficult and that we'd basically have NO chance of getting it any time soon.

But... then I realised that this is very different from the car position & control updates (which often have very poor resolution - e.g. 3 times per second).

Head movements don't affect the car position, so latency isn't such an issue. I suspect it would work just fine if (say) a second's worth of updates got stuffed into one packet... And I *think* this would make it somewhat less demanding. So we could achieve something like 10 head position updates per second without much network traffic (IIRC, the short packets are dominated by packet header), as long as we're prepared to tolerate a short delay...

Scawen said he wasn't sure how many LFS users were still on XP.
I guess this means that the connection to the master server doesn't transfer any useful info of this kind - but perhaps it should?

I think most users play online at least sometimes - the AIs are too awful to be fun, and even hotlappers probably upload laps now and then, which could also (maybe already do?) have a field inside to record the OS version.

The master server and hotlap scans could then record data for a few months and we'd know which users are on which OS versions (some will use multiple no doubt).

This is not a tweak I'd want to see holding up the tyre physics, obviously!, but would help to guide the decision about moving on from XP.
(Btw, I still use XP myself )

Quote from sicotange :

Suppose 8 players are online and racing. If 3 players vote to restart and at some point 1 racer spectates then an auto restart could be triggered. That is if 50% (or more) of votes are required.

If you meant "7 players are online and racing" then yeah, I think that's what would happen (3/7 is less than half...)

I've personally seen races restart on a spectate event (caught me by surprise first time!). Indeed I interpreted it to mean that the vote threshold must reduce by 1...
And wth, it's always good to get bug reports so don't feel bad

Quote from Dandy Dust :

I am driving project CARS right now, it´s far from perfect, but there you get new cars and tracks and physics and stuff week by week.

How sick are we, to discuss the stupid scirocco any more at all?

It won´t save LFS even if it came out with Rockingham and new tyre-physics tomorrow... sorry.

Mmmm.
Almost sounds like you're saying that getting new stuff is more important than the quality of the new (and existing) stuff...?? It was before my time, but I think LFS went through a similar phase of rapid development in the early days, when it was basic/ropey but had enough promise to generate interest.

Personally I don't really give a toss about the Scirocco itself. I agree that this wouldn't "save" LFS.

A new, real, track would be cool though, and updates/fixes to the other tracks are also long overdue. (I have no idea why anyone cares about the car interiors, mind you.) They might revitalise LFS.

But I reckon new tyre physics could make a huge difference and get lots of old blood back in, as well as plenty of new blood...

To be brutally honest, I am pleasantly surprised that LFS hasn't already imploded. Not wishing to slag off the devs here, but progress has been stunningly slow.

Are there enough LA racers to try for your own multiclass server? (See earlier suggestion - I doubt Dave would be too keen to manage it, but you might twist his arm into helping you to set it up...)

I can see why someone wouldn't be keen to face an angry mob - in person.

But posting a message to say "sorry guys, business went tits up, money's all gone, my apologies but I can't refund you; you better look elsewhere for hosting" would have been so much better than silence.

So yeah, maybe his body and/or head are not in a good place...

Quote from Squelch :

There is another possible reason for ISP's dropping/blocking UDP, and that's for comercial reasons. VoIP is in direct competition with telephony, RTP video streaming etc. are now becoming "premium" services, and Torrents are a grey area, so certainly count as don't time critical. Perhaps gaming is considered low priority for mainstream internet access, and falls into the premium category alongside video? Overall you make some fair points, and hosters need to pay particular attention to their UDP service.

I've been reading that the standard speed and ping tests are not a good indicator of UDP throughput as they rely mainly on TCP. I'm not quite sure how this can be tested for properly, but RTP (real time protocol) jitter is probably a more worthwhile metric.

Well, I came up against "traffic management" when I was with Plusnet. Their service started out excellent, but (popularity?) went downhill after they tweaked their traffic management settings.
However, all recognised gaming packets were given high priority - the bandwidth used is generally tiny and gamers who can't play their favourite games are highly prone to leaving for another ISP (as I did).
They struggled for several weeks to find a way to recognise the LFS packets sequences but ultimately failed to find a reliably recognisable signature. Some times the game would work OK, sometimes stupid lag + disconnects. Thus I ceased to be a Plusnet customer...

But in any case, UDP packets are definitely the only way to go for ephemeral data like position updates. (I can't now recall whether TCP or UDP was more troublesome in my case.)

Missing skins: my vote would be with Scawen - spectate 'em

Agreed. Basically, everybody hopes he's OK, while at the same time wishing for their servers to be back up or at least to know what's going on and when they will be back up.

It's precisely the "one man band" thing that's the problem. (Occasionally he has had one or two others step in, tho IIRC they've been silent this time.)

But... I reckon most of the people who get their servers from him - or at least the ones who are posting here - have been aware for some time that he is a one man band, so it's not like this can have taken them by surprise...

Quote from MrSam :

I understand better thank you, I had not noticed.
but the current is really bad, a little change for the next update would be nice
example reduce the size numbers so it does not overlap

Woo, you mean you actually use the speedo while driving hotlaps? If so, then either I'm unusual (cos I never do that) or you are... How does the speedo help you?

Quote from capris :

3 days now and counting without a server and as to what the problem is were is the support when franky is not around to sort things out,we need some feedback and information [...]

Well, what it's really about for me is simply giving customers the service level they have paid for. If you tell 'em up front: "this is a dirt-cheap service, so you can't expect quick responses and sometimes it's gonna take a week", then cool... Or if you make any promises about prompt support (but does Franky do so?) then clearly it's bad form not to keep those promises. Franky does provide a rather cheap service folks...

Quote from proteus :

And whats 500Servers gift to all its customers this year their lives back for the festive season lol

Nice

Quote from E.Reiljans :

If you knew who creators of Kebrum.com are, and what/whom it was made for, you'd know that this statement in their terms is there just to scare away script kiddies.

Ah.
(And em, it could be argued that we're veering off topic now so I think I should end this strand of the thread )

Quote from E.Reiljans :

Then again, there are prq.se, kebrum.com and other decent VPN services that don't rat their users out.

Picked kebrum.com at random. Found this in their terms:
"If you violate any of our terms that involve criminal activities, you will forfeit all the privacy and confidentiality privileges that our service provides."

I doubt that any clueful black-hat hacker would pay for VPN services of this kind.

Quote from cargame.nl :

http://www.youtube.com/watch?v=BATk_Yf8D48

This is how stupid it gets with everything here, Scawen writes something and immediately needs to protect / back up his words. And thats why there is no progress report or whatever worth to be putted online.

Always but, I've heard.. No I know better! ... Not.

Firstly, jeez Dave, did you get out of bed on the wrong side this morning??

Secondly, your link shows how a paying customer of a VPN vendor can connect to stuff from lots of different countries. See my comment about "legit account" above? This is not how hackers roll, because the VPN provider could (and in this case probably would - based in the UK, lol) give their details straight to the authorities if a complaint was made. This doesn't really anonymise him at all if what he has in mind is illegal...

Finally, who contradicted Scawen?? Not sure what you mean by that.

Quote from cargame.nl :

Its not called proxy, its called VPN networking.
[...]
I think he is gone anyway. So, time to focus on tire physics in the new year, I guess.

Quote from Degats :

Not all proxies are web proxies.

On the proxying point - there is actual evidence of multiple countries being used? Because while you can do anonymous web proxying from countless servers out there for free, I didn't think the same was true for VPNs or any other methods of proxying. Thus he'd surely have to have actual control of (via a legit account or maybe a trojan) the machines that are acting as middle-men...

But hey, if he has packed up and gone, then great, game over

Quote from DANIEL-CRO :

He is using proxy servers to go online

Huh - LFS works with proxy servers?!? HTTP over proxy - sure. But UDP position updates?

Quote from cargame.nl :

On the other hand I wonder why people demand 2012 stuff while they are at a computer with parts from 2005.

Meh, the issue - whatever it is - is likely there for everyone. It could be that a slower PC highlights the issue/bug/feature... for which we should be grateful, right? Cos then we can squash it

Quote from the_angry_angel :

For instance like hell am I going to continue running a kernel with a known remote code execution in it - even if its firewalled, whats the point in increasing the risk?

Well obviously because the trauma of zeroing your > 1yr uptime is just so devastating!
(PS: can't you just use a kernel debugger to bump the jiffy count right back up there? )

Mmmm, does that imply that a bunch of folk are trying out a (fixed) cheat?