The online racing simulator
Spinoff : DirectX and Windows version discussion
(291 posts, started )
Quote from Flame CZE :Nothing will change visually anyway, so why all the fuss

i have my reasons ...
Hi, the DX9 version seems good but there are a couple of minor things to fix (easy) and also I should have a look at my list of things to do. There isn't really anything else new in this version so it's not a very exciting test patch so far! But do let me know if it is important to release the DX9 test more quickly.
Why did we need a DX9 version again?
Some graphics cards recent drivers for XP have a bug in them where LFS in DX8 goes wrong. But using a DX8 to DX9 converter the bug disappears. Of course this extra software should not be needed, so it's a good idea to fix LFS. DX9 is much more used than DX8 so it's going to be well supported by everyone for some time to come, whereas lack of testing might introduce bugs in DX8 support. It's better than DX8 and has more possibilities (don't ask me what) so that could help with future graphical developments. We do not want to go further than DX9 any time soon because that would eliminate the possibility of LFS running on XP or Linux.
-
(Bmxtwins) DELETED by Scawen : spam
-
(CodeLyoko1) DELETED by Scawen : reply to spam
Quote from Scawen :The other problem with Win7 is the removal of the ability to select dual monitor as a full screen option, using a single device. That was possible in XP and not since. The nvidia multiple monitor thing is 3 screen only.

That must be an nVidia driver limitation. I've had no problem with my last couple of a ATI cards, on both Windows 7 and 8.
Quote from Bob Smith :That must be an nVidia driver limitation. I've had no problem with my last couple of a ATI cards, on both Windows 7 and 8.

In Win XP you can set two monitors behave like single surface, independently of graphics drivers... and that was droped since Windows 7. You can set it using software for controlling graphics, but it isn't so native like in XP.


I guess DX9 should bring better performance to Windows 7 and later OSes
Quote from Scawen :Hi, the DX9 version seems good but there are a couple of minor things to fix (easy) and also I should have a look at my list of things to do. There isn't really anything else new in this version so it's not a very exciting test patch so far! But do let me know if it is important to release the DX9 test more quickly.

Quote from Bmxtwins :Does dx9 make tire physics easier to implement?

The thing that DX9 support aids greatly in (other than driver bugs) is with recording/streaming of LFS. FRAPS seems to be the only application that can record LFS in DX8 mode. Most streaming applications won't operate in "Game" mode (which is more performant) with DX8 applications, so required a DX9 proxy DLL or to run it in Windowed mode (which is slow).

Your snark about the tyre physics is (as per your posting norm) counter productive and quite useless! Especially because we already know that it will be done when it's done.

Quote from Bob Smith :That must be an nVidia driver limitation. I've had no problem with my last couple of a ATI cards, on both Windows 7 and 8.

The amusing part is that the reason it's "broken" in Windows Vista/7/8 is because they improved their handling of multiple monitors. XP treats multiple monitors simply as 1 gigantic surface, where Vista and newer actually recognize the display boundarys.

The nVidia drivers do have "nVidia Surround" which does allow you to create 1 giant surface. Just like ATi cards have "Eyefinity".
Quote from dawesdust_12 :The nVidia drivers do have "nVidia Surround" which does allow you to create 1 giant surface. Just like ATi cards have "Eyefinity".

However Surround requires 3 screens, Eyefinity only needs 2.

I think they're both fairly terrible for desktop work unless you run everything in small windows... and hope that Windows (OS) remembers their location and size, which it usually doesn't.
Quote from dawesdust_12 :The thing that DX9 support aids greatly in (other than driver bugs) is with recording/streaming of LFS. FRAPS seems to be the only application that can record LFS in DX8 mode. Most streaming applications won't operate in "Game" mode (which is more performant) with DX8 applications, so required a DX9 proxy DLL or to run it in Windowed mode (which is slow).

ShadowPlay records fullscreen vanilla LFS. ShadowPlay can be used to save locally and/or to stream to Twitch.
Quote from Matrixi :However Surround requires 3 screens, Eyefinity only needs 2.

I think they're both fairly terrible for desktop work unless you run everything in small windows... and hope that Windows (OS) remembers their location and size, which it usually doesn't.

only thing it doesnt remember is the task bar. I disable my eyefinity when not playing lfs.
Quote from amp88 :ShadowPlay records fullscreen vanilla LFS. ShadowPlay can be used to save locally and/or to stream to Twitch.

Have you actually tried to stream to twitch with ShadowPlay? I can't seem to get it to stream for more than 1 minute. It's terrible
Quote from dawesdust_12 :Have you actually tried to stream to twitch with ShadowPlay? I can't seem to get it to stream for more than 1 minute. It's terrible

I don't have a Twitch account.
That means Victor is behind it then? I doubt Scawen tweets.

Quote from Scawen :Hi, the DX9 version seems good but there are a couple of minor things to fix (easy) and also I should have a look at my list of things to do. There isn't really anything else new in this version so it's not a very exciting test patch so far! But do let me know if it is important to release the DX9 test more quickly.

I'd like to see it, if for no other reason then "progress". At this point, DX9 is better supported the DX8 as vendors begin to deprecate DX8.
Quote from Dygear :That means Victor is behind it then? I doubt Scawen tweets.

I think it's Scawen who tweets there, because he also has his own twitter profile and tweets there too. And since the latest tweet says "My version of #liveforspeed is running in DirectX 9 at last", that kind of implies that it's Scawen
Quote from dawesdust_12 :Considering the twitter account was created after that quote was stated. Also the LFSRacing twitter account has a follow button on the main LFS.net page.

OK, I guess Victor´s opinion about twitter didnt matter (no offence, Victor, you are great )
Hi Legend. And no, I'm not stalking you. Too much.

And not to be completly offtopic, someone else posted earlier if we can get a test patch of this new DX9 version when its finished, but also if there is going to be multi-threading support in it. I'm really looking forward for that. At the moment LFS is using just one core... other 3 are for porn during racing. I guess its not that bad, but still.
Quote from bogdani.cojocaru :Hi Legend. And no, I'm not stalking you. Too much.

And not to be completly offtopic, someone else posted earlier if we can get a test patch of this new DX9 version when its finished, but also if there is going to be multi-threading support in it. I'm really looking forward for that. At the moment LFS is using just one core... other 3 are for porn during racing. I guess its not that bad, but still.

I doubt that it will happen in this round of test patches.
I was seeing if I could get the a DX9 version out today and have been solving the little things. Only need to do the gathering of screen modes and there will not be any "bugs" left.

But I have to stop for the weekend now, so I'd say it could possibly be released on Monday evening, but while finishing the DX9 support I haven't looked at my lists yet to see if I could do any other quick updates. So I don't want to commit to a Monday release, though it may be a good idea to do that at get the DX9 tested. But if my lists of notes present a few good quick updates, maybe it would be better to spend a couple of days on those and get them tested too.

It looks exactly the same, not a pixel different, so don't be too excited! The good news will be if it just works properly on everyone's computer without any need for a DX8 to DX9 converter, etc...
Quote from Scawen :I was seeing if I could get the a DX9 version out today and have been solving the little things. Only need to do the gathering of screen modes and there will not be any "bugs" left.

But I have to stop for the weekend now, so I'd say it could possibly be released on Monday evening, but while finishing the DX9 support I haven't looked at my lists yet to see if I could do any other quick updates. So I don't want to commit to a Monday release, though it may be a good idea to do that at get the DX9 tested. But if my lists of notes present a few good quick updates, maybe it would be better to spend a couple of days on those and get them tested too.

It looks exactly the same, not a pixel different, so don't be too excited! The good news will be if it just works properly on everyone's computer without any need for a DX8 to DX9 converter, etc...

so it need to rewrite a big part of codes to getting a visual difference, but any chance to getting small fix to render correctly world shadows ? ... because IMO this is the most thing that annoy me with lfs graphics.

thx anyways looking forward for DX test patch.
You mean the shadows of the world objects? The ones that are all spiky and messy? I agree they are a problem but there's no quick solution for that. Really I want to get to the point where the sun can move in the sky and the shadows can move, as part of a changing weather system.

But this is after a tyre physics update. I'm trying to get realistic about the tyre physics update, to make a good solid version that works well and makes LFS cars feel great (because of the force combinations that are mathematically based) with good simulation of tyre load sensitivity but not worry too much about extreme detail as it will never exactly match reality.
Quote from Scawen :You mean the shadows of the world objects? The ones that are all spiky and messy? I agree they are a problem but there's no quick solution for that. Really I want to get to the point where the sun can move in the sky and the shadows can move, as part of a changing weather system.

But this is after a tyre physics update. I'm trying to get realistic about the tyre physics update, to make a good solid version that works well and makes LFS cars feel great (because of the force combinations that are mathematically based) with good simulation of tyre load sensitivity but not worry too much about extreme detail as it will never exactly match reality.

and by the way disable forces view in online racing and hotlapping! it ruins realism.
Isn't forcing views an option in a server?
You can force cockpit view, but AFAIK that still allows you to enable forces view.

Also, I love SETH's continual effort towards the struggle of non-cockpit driving views. It shows a lot of determination. Also is funny.

Spinoff : DirectX and Windows version discussion
(291 posts, started )
FGED GREDG RDFGDR GSFDG