The online racing simulator
Test Patch U25: Multiplayer Updates
(353 posts, started )
Quote from Scawen :You should find the TTime in IS_RES now has the qualifying session time. Please let me know if it looks right. I did a quick test with the IS_RES appearing live and when requested with TINY_RES.

I've run through a couple of replays and the live and requested TTime values do seem to be consistent with eachother. The sorting is working so far Smile


One observation, the IS_RES TTime ("session time") doesn't quite match the IS_LAP ETime ("total time")

Result: id: 10 ^0[^7MR^0c] ^7Michal (michal 1279) laps: 0.0 pos: 6 total: 20:21.690 fastest: 01:51.010
Lap Time: 01:51.010 Total Time: 20:18.220 (michal 1279)

If I understand correctly, the IS_RES time is equivalent to the replay time and IS_LAP is equivalent to the time since the lights go green?

I've noticed that some packets have ETime and others are TTime - are they always consistent with the interpretation of the two times above?


Like I said, just an observation and it doesn't affect what I need the IS_RES time for.
-
(vanopaniashvili) DELETED by vanopaniashvili : I often don't engage in threads and don't want any of it appear on my posted threads list
Hi! I've noticed several strange things and bugs in training lessons. At least one of them is related directly to this test patch and I found two others due to me trying to test that new Live Replay-feature in this test patch, meaning that they could be related to this test patch as well.

STRANGE THING #1:
While you can't use most commands in training lessons, for some reason /canrefuel works here even with parameter no when using that command via chat bind. This makes that "Pitstop - Formula BMW" training lesson significantly easier (and yes, I know you can already cheat in that lesson by just simply opening F12-box and making sure that you don't refuel at all, but this makes it even faster). I've attached a screenshot about this.

STRANGE THING #2:
You can cheat in "Chicane Course - Formula BMW"-lesson by simply shortcutting after second corner. You don't have to do those last two corners, you can simply drive between two cones and straight to the finish line. I've attached a screenshot about this.

STRANGE THING #3:
Due to that live replay-function introduced in this test patch, I had to test this in single player as well (I know it doesn't work here!). If I'm doing a training lesson and then decide to open another LFS instance to view that temp_spr-replay, something strange happens:

Viewing temp_spr-replay is not allowed at all and LFS will instead inform something about checkpoints/autoslalom tracks or failed header check. I've attached couple screenshots about this. However:

I found one way to view temp_spr-replay from training lesson when playing. It's really strange how I did it, but doing that will not only allow viewing that temp_spr-replay in another LFS instance, but it's actually a bug that causes training lesson to get stuck (I mean by that it's doesn't automatically end). I will tell more about this next:

Four actual bugs:

BUG #1:
So this is the way to view temp_spr-replays from training lessons: If you use command via chat bind that would normally result replay not getting recorded in single player (like command /wind), LFS itself will never end that training lesson. And you cannot end it by yourself either by pressing 1, 2 or Esc-button. The only way to end it (other than closing LFS entirely) is to press Shift+X or alternatively, use command /end via chat bind. What this means you can now watch that temp_spr-replay from that LFS instance or open another LFS instance to watch that.

What this does result as well is that even if you successfully complete training lesson and then use that command which normally would stop recording single player replay, LFS doesn't let you clear that training lesson.

Way to reproduce this:
1) Make a command that would normally end recording a replay in single player (like /wind 1) to a chat bind
2) Start training lesson. Doesn't matter if you finish it or not
3) Before that training lesson ends automatically, press that button where you assigned that command. That training lesson will never end then.

BUG #2:
If you don't have replay saving function enabled at all and you do that chat bind-bug I just mentioned, it still mentions "Replay not recorded" even though it wasn't recording in first place (or it actually was, check bug #3). This doesn't happen in single player. I've attached a screenshot about this.

BUG #3:
Even if saving single player replays is disabled, LFS still saves automatically replays of training lessons. Way to do this is to open another LFS instance and wait until your training lesson ends automatically in that original LFS instance. Now when in that training lesson you're viewing result of your training lesson, you can actually see in that other LFS instance that it saved that replay automatically. You can't rename it and if you exit that result screen in that training lesson, it will be deleted. To save it, you have to watch that replay at the same time in that other LFS instance when exiting that result screen.

Way to reproduce this:
1) Open LFS-instance. It doesn't matter if your SPR-saving is set to no save, manual or autosave
2) Open another LFS-instance after that.
3) Start training lesson and wait until it automatically finishes. Now, don't exit that screen where you view your result of that training lesson.
4) In that another LFS instance, go to Replays-page. That replay can be found here even if you don't have SPR-saving enabled at all. And if you're watching that replay at the same time you exit that "training lesson result"-screen in that another LFS instance, that replay will be saved. After that, you can rename it.

BUG #4:
If you use command /aiset or /aiset_all via chat binds, you can cheat in overtaking lessons. So what you can do, is for example to put command /aiset PRO 1 or /aiset_all 1 in F1 chat bind. If you then press that button during that overtaking lesson, it just changes that AI-driver's skill level to that skill level which makes these lessons easier (or harder) for you. I've attached a screenshot about this.

Also worth noting in overtaking lessons your skill level when completing those is based on how many AI-drivers you overtook. So if for some reason OK-wins and you beat both PRO- and QUICK-drivers, your skill level is QUICK, not PRO, even though that you won that PRO-skilled AI-driver.

Way to reproduce this:
1) Make a command /aiset PRO/QUICK/OK 1-5 or /aiset_all 1-5 to a chat bind
2) Start any training lesson about overtaking.
3) At any point during that training lesson before it finishes, press that button where you assigned that command. This way you can change that skill level for your AI-drivers to make that overtaking lesson easier or harder for you.

POSSIBLE BUG #5:
Is it on purpose that if I restart overtaking lesson, AI's will have that minimum fuel load at start instead of 50%? Check out that screenshot I attached about wrong skill levels, it also displays AI-drivers only having 2% fuel load at start instead of 50%.
Attached images
Bug2.jpg
Bug4.jpg
StrangeThing1.jpg
StrangeThing2.jpg
StrangeThing3A.jpg
StrangeThing3B.jpg
observation: If you are willing to cheat training sessions then you do not understand the purpose of training sessions
Quote from Degats :One observation, the IS_RES TTime ("session time") doesn't quite match the IS_LAP ETime ("total time")
...
If I understand correctly, the IS_RES time is equivalent to the replay time and IS_LAP is equivalent to the time since the lights go green?

I've noticed that some packets have ETime and others are TTime - are they always consistent with the interpretation of the two times above?

Thanks for the test.

Yes, the IS_RES TTime in practice mode is the time since the start of the session until... and here is the dirty secret... the time the packet is processed at the server (a short and unpredictable time after the packet is sent - in this case 0.47 seconds). And, as you said, the IS_LAP ETime is the time since the lights go green (3 seconds after the start of the session) until the time the car crossed the finish line. So that's how to get the 3.47 seconds difference.

Explanation: The IS_LAP is triggered by an internal packet that has a proper time stamp (from the driver's computer) of total time as that is very important for accuracy. But the internal result packet that triggers the IS_RES packet in qualifying mode doesn't have that accurate time stamp as it was never needed before. So the server goes with the time the packet is processed for this new use of TTime in IS_RES in qualifying. I think for the purposes you are describing, that additional slight delay won't be a problem. I think it could only be a problem if two qualifiers cross the finish line at almost exactly the same time, with exactly the same lap time, then the tie-break between these two identical laps might not get the right order. But that might never happen! Smile And if it did happen I think that would match the order generated by LFS...

EDIT: Sorry, this isn't quite right as they are both triggered by the same internal packet. So I think it would be possible to make the IS_RES TTime in qualifying equal to the associated IS_LAP ETime if needed, by passing some extra parameters to a couple of functions. But then the order might not match the order generated by LFS in that very rare case described.

Quote from vanopaniashvili :Increasing the range of steering wheel rotation is welcome news and I got a question about it. What was the reason behind changing all street cars to 900 degrees of rotation?

I'm concerned about the non-wheel users of LFS. It's certainly gonna affect their understanding of vehicle handling characteristics which they are already familiar with and force them to rethink everything. Obviously, steering ratios play a huge role in letting you know about the handling character of a vehicle and LFS was always familiar with 720 degrees of rotation, which gave a certain steering ratio output while being paired with default 36 degrees of steering lock. It gave street cars a certain driving characteristic, which is also what most of the users are familiar with. Changing LFS's signature steering ratio, might cause a confusion in LFS users.

I always knew the steering wheel turns in LFS were too low, compared with real road cars. I had already changed most of the road cars in the development version to the new value. It is an old request to update the steering angles, maybe ever since the higher range controller wheels were available. So now as I was dealing with wheel turns on the racing cars I started to look into steering ratios.

Side note for anyone who doesn't know: steering ratio is the degrees turned by the steering wheel, compared with the degrees turned by the front wheels. https://en.wikipedia.org/wiki/Steering_ratio

Normally in road cars this is between around 12:1 and 18:1 - usually closer to around 15:1 - but in LFS previous versions:
- FWD road cars had a ratio of 12:1 (720 deg steering wheel for 60 deg road wheel)
- RWD road cars had a ratio of 10:1 (720 deg steering wheel for 72 deg road wheel)

So the previous version of LFS was outside of the values ever seen in real road cars. With the new adjustment, the FWD cars have a ratio of 900:60 which is 15:1 and the RWD have a ratio of 900:72 which is 12.5:1 and this is closer to reality (though still a bit low in the RWD case).

I know this will make the cars feel different, while having the same underlying physics, but I'm hoping that as it's closer to reality that shouldn't be a bad thing.

EDIT: This change should not affect mouse or keyboard users at all. In their case the keyboard or mouse operates the steering directly so there is only a graphical change. The change in feeling should only affect wheel users and maybe has a slight affect on joystick users.

Quote from tankslacno :Hi! I've noticed several strange things and bugs in training lessons. At least one of them is related directly to this test patch and I found two others due to me trying to test that new Live Replay-feature in this test patch, meaning that they could be related to this test patch as well.

Thanks for the detailed descriptions.

Some of them I won't worry about but I will have a look to see if some of them can be corrected easily.
I was combining the change logs U16 to U23 just now and noticed I had forgotten one addition to InSim:

IS_NPL RWAdj / FWAdj indicate rear / front tyre width reduction
That option didn't turn the view, only adjusted the position a bit as the animation moved. Steer look is a separate thing and that is still available.

The /block command is listed in Commands.txt

/block [0/1/2] - block messages (none / user / all)

/block 1 is like pressing minus
/block 2 is like pressing SHIFT + minus
-
(Scawen) DELETED by Scawen : off topic
-
(redbot_) DELETED by redbot_
-
(Rocket-Racer) DELETED by Scawen : off topic
Are we getting the final patch soon? I can't wait to test it Omg omg omg
you know what i said is not off topic Dead banana
Quote from thegamer147 :Are we getting the final patch soon? I can't wait to test it Omg omg omg

Please test now. There is little point testing after the official version is released!

Quote from MacedoSTI :you know what i said is not off topic Dead banana

Yes, you instructed me to update the driver models, but that is off topic.

I fixed the driver animations to deal with new steering limits. It's not closely related to working on the driver models.
Thank you all for the feedback and testing.

Test Patch U25 is now available. I hope it's the last test patch.

Please have a go with it if you can.

Changes from 0.6U24 to 0.6U25: COMPATIBLE WITH U23 AND U24

Training:

FIX: AI changed to low fuel load if overtaking lesson restarted
FIX: AI skill / admin commands no longer processed during training
FIX: Training lesson did not end if replay saving was interrupted
FIX: Refuelling depended on refuelling allowed in single player
FIX: Logo was visible under title during lesson replay

OutSim:

OutSim packet is documented in docs\OutSimPack.txt
Added steering torque as additional field in new OutSim
All data options can be switched on with OutSim Opts 1ff

Misc:

Removed debug message "Replay name : temp_mpr"
Saving replay name now shown beside option in Options - Game

https://www.lfs.net/forum/thread/95016


I did open U25 and did noticed that driver is standing in the car.
You have to copy few extra move files too from the patch extract data/move

Not just the exe

I made that mistake too before Big grin
I did also noticed that the camera angles are really wierd... i cant even go copckit mode way i want.
ahh.. ok! thanks redbot_!
To install the patch:

1) Move or save the patch into your main LFS folder
2) Double click the patch to extract it to that folder
3) When you see "Confirm File Replace" select "Yes to All"
4) Now you can start LFS in the normal way

If you want to use some other (manual) method of installation, please take care to install all the necessary files.

Thanks for the testing! Thumbs up
Quote from Scawen :Thanks for the testing! Thumbs up

Hmm OK.. I finally would like to know where this appalling lighting (?) is coming from because every time I get confronted with LFS this disturbance also shows up reminding me to stay away from LFS.. Its especially being seen on the track decorations but also on the road surface



what setting do I need to change? It alrready shows up when I am standing still and turning the steering wheel.

This only happens @LFS ... When I am for example moving/looking around @CS.GO everything is fine.
I think it is fixed in the developer version... if I remember correctly this was an issue while developing the day-night cycle. Shrug
I'm not sure what you mean. Are you talking about some shimmering of thin objects and also other things that appear 'thin' such as dark lines on the track in the distance?

There's not much to be done about that in this version and I don't think you are talking about a new issue in the test patch. But you might get slightly better results with 8x AA.
I can't get mine to look that bad on 4x AA, so it might be video compression making it worse - I have to turn down to 2x AA to get it to look like that.
Also check that your GPU driver isn't messing with things, because that can completely override the game's AA setting and even change the AA mode.
wait wait did i miss something... there is no more 60 degree of xrr???
60 degrees is available when you use narrower front wheel width
is possible put slick tyres on XFG for trackdays and drag servers on demo?
No, we can't make changes now. That was discussed before anyway. With the current physics, the road cars can roll with slicks so it's not as good as you might think. We don't want to add things to the demo. Demo is supposed to be a demonstration of LFS, so people can buy a license if they like it. We have bills to pay.

This morning I could fix a serious bug if anyone finds one but other than that I won't be making any changes.
I've just done a quick test of the RWAdj and FWAdj values and they seem to make sense.

Test Patch U25: Multiplayer Updates
(353 posts, started )
FGED GREDG RDFGDR GSFDG