Author Topic: "stuck in pit lane"  (Read 13463 times)

0 Members and 1 Guest are viewing this topic.

MrFlippant

  • Here a Real Lot!
  • *****
  • Posts: 104
  • ¯\_(ツ)_/¯
    • View Profile
Re: "stuck in pit lane"
« Reply #15 on: December 10, 2014, 06:20:58 PM »
As a person using an older, underpowered laptop (cheapie from Wal-Mart that came with Vista and crapware, downgraded to XPXP3), I very much appreciate all the work you do to keep the software running smoothly on such computers. I also agree that it would be unfair to withhold features just because some people might not be able to use them without slowing things down too much. I try to make it clear that whenever I suggest a feature, I mean it should be optional, but if I forget, please assume that I mean for any feature to be optional, if possible. :)

I don't think you need to worry much about what other cars are doing in the pit lane. As Grunz said, if you see a car cross the sensor when it's still in the pit menu, start a timer (user configurable) and any car detected passing the sensor before that timer expires will get a penalty.

For what to do when that happens, if possible, make them optional, but if that's too complicated, I'd be fine with:
Fuel: stop at current tank level
Tire: Can it be treated like a tire wear level? If switching to different kind of tire, just give it to them. I would consider this part of the penalty to the driver that pushes the car. They just gave the other can an advantage by pushing them out of the pit lane. As pit lane strategy can mean a lot in a race, it rarely behooves the pushing driver to give the other car fresh tires without the necessary delay.
Repair: stop at current level of repair based on elapsed time. Maybe like Tires, give them full damage points back.
Stop&Go: cleared. Just like Tires, the pushing car is put at a disadvantage because he was impatient.

IMHO, giving the pushed car the benefit in such a situation will be a strong deterrent to those who would consider pushing them out of the way. While they might get out of the pit lane sooner, they'll just have given the other car the upper hand with no penalties or damage to clear, fresh tires, and heck, maybe even a full tank of fuel. :)

But, that's me. ;)

GRUNZ

  • Here a Lot
  • ****
  • Posts: 89
    • View Profile
Re: "stuck in pit lane"
« Reply #16 on: December 11, 2014, 02:01:51 AM »
Martin
I am not saying that is easy, but I do not see what  the two pit lanes have to do with it. In the end all the cars go over one lane where the pit out sensor is. I have tried to put in pseudo code to try to clarify that.
But anyway, I think that this is a minor feature that would require a lot of coding. Still, I think that the closing of the pit menu when a car goes over the pit out should be implemented especially in tracks with single pit lanes.
« Last Edit: December 11, 2014, 08:38:25 PM by GRUNZ »

MrFlippant

  • Here a Real Lot!
  • *****
  • Posts: 104
  • ¯\_(ツ)_/¯
    • View Profile
Re: "stuck in pit lane"
« Reply #17 on: December 11, 2014, 02:34:00 AM »
Yes. The penalty for pushing stuff is a "would be cool" but definitely low priority.
Getting cars to kick back into racing mode when pushed out of the pit is the priority. Even if they get everything stopped and have to start from scratch after another lap.

Grunz, what are your thoughts on what happens to pit-related activity for the car that gets pushed?

martin

  • Administrator
  • Here a Real Lot!
  • *****
  • Posts: 276
  • RCS64 Programmer
    • View Profile
Re: "stuck in pit lane"
« Reply #18 on: December 11, 2014, 05:51:48 AM »
When using a single lane pit-lane you know the order off the car entering into the pit, then I know what car pushing..
With a two lane pit, I have to make a timer. Today there is a timer running every 100 ms to handle all kinf off things like pitmenu functions, so when you press a butoon. but I am not sure 100 ms is fast enough for this so I have to turn down on this update rate, but then again it affect the gui update and also the throttle feelings. driver update rate. So I actually need to make another parallel loop for this running faster than 100ms. but if it was a single lane pit I know the order then i dont have to have another loop running..

Does it make any seense?

MrFlippant

  • Here a Real Lot!
  • *****
  • Posts: 104
  • ¯\_(ツ)_/¯
    • View Profile
Re: "stuck in pit lane"
« Reply #19 on: December 11, 2014, 06:55:34 AM »
I forget, is 100ms equal to 1 second, or one TENTH of a second? If .10, then that's plenty fast enough.
1s would be a little long, but I wonder how a car could be pushed past the exit sensor at ALL if not by the car immediately following it at any point thereafter. Even if the pushing car stops before the sensor, he'll be the NEXT one over the sensor no matter what.

Unless you're proposing that a two lane pit has two exit sensors, one for each lane. Does anyone do that yet? Is it possible yet?

Minardi

  • Administrator
  • Here a Lot
  • *****
  • Posts: 98
    • View Profile
Re: "stuck in pit lane"
« Reply #20 on: December 11, 2014, 08:51:57 AM »
Ok, I understand what is going on here, my thoughts;

- No automatic penalty for the pusher. I believe we can still award this with the  keyboard.Who says the pusher isn't pushing 2 cars ?
And It would avoid the situation that a car on purpose is waiting to be pushed.

-When a car that crosses pit out is in pit menu, give the benefit of the doubt and make the value of the chosen pit activity at the moment of crossing pit out 100%.
Other values remain as they are.

 
I would hate to see put Martin a lot of effort in this while there are so many other (new features!) things we can do first.

MrFlippant

  • Here a Real Lot!
  • *****
  • Posts: 104
  • ¯\_(ツ)_/¯
    • View Profile
Re: "stuck in pit lane"
« Reply #21 on: December 11, 2014, 04:57:04 PM »
That's a good point about a car possibly pushing out two cars, but it would be simple to determine that, because the second car would also still be in the pit menu. So, instead of "the car after the one that crosses in the pit menu" it would be "the first ACTIVE car following any number of cars still in the pit menu".

But, as Grunz and I have both said, it's less important to worry about this penalty stuff. Let's just get pit menu cancellation for any car that crosses the sensor for ANY reason in there. Give them full *whatever they were doing* when they cross, and worry about penalties for the pusher later.

I like that the marshal can apply a penalty, of course, but when the rest of us are racing, having something like that be automatic would be very nice. We almost never have a spare person, so unless someone activated a track call and remembered to indicate that someone gets a penalty for pushing a car in the pit lane, it probably would be forgotten by the next car crash.

GRUNZ

  • Here a Lot
  • ****
  • Posts: 89
    • View Profile
Re: "stuck in pit lane"
« Reply #22 on: December 11, 2014, 09:19:49 PM »
I have thought about the second or third or whatever car being also in Pit Menu mode and being pushed. Basically the offending car is pushing a train of cars.
If you check I have put a condition in my pseudo code example that takes this scenario in to consideration:
 If (PitOut sees ID2) and (ID2.status is not in PitMenuOpen) and (Timer is not Up)
                                       ^here I am checking that the following car is not in PitMenuOpen mode.

I do not know how good LabView works in multi-threading. For sure there is a lot of synchronisation going on here.

As for the two-lane pit lane, it could be the case that there is a XLC in it, so the sequence of car entering might not be the same of the sequence in exit. As an example, let's take the 2-lane pit lane shown below (I hope that is clear enough):

Lane 1____________________________________________________________
Lane 2____________________________________________________________   
Pit Lane 1      \___SS-IN_______1______   ___3_______SS-OUT____/
Pit Lane 2                          \_____2______X___4_____/   
 

The numbers 1,2,3,and 4 represents the pit bays, SS-IN is the smartsensor in (SS-OUT is smartsensor out), and the X is the XLC placed in the middle of the pitlane.

Let assume car with ID 1 enters first and goes to bay 4 and open the pit menu. Car with ID 2 enters the pit lane and goes to bay 3.Car sequence so far is [2,1] (this is not a FIFO as I will show later on). Now, car ID 3 enters and goes to pit bay 2 (ID sequence is [3,2,1]). While car ID 1 is still in pit menu open (and car ID 2 is also still in the pit but does not matter if pit menu open or not), car ID 3 pushes car ID 1 out of the  pit lane. When a car ID goes over the SS-OUT the sequence is updated by removing the car ID from the list: in this case when car id1 goes over the SS-OUT it will be [3,2] and then when car id goes over SS-OUT the sequence will be [2].
As you can see in this simple example, the sequence in which the cars have entered the pit lane does not help for identifying the offending car. And one cannot even assume that the offending car has to enter the pit lane after the pushed car. Here is another example:

Car id 1 enters the pit lane, stops in bay 1 and goes in pit menu open (sequence is [1]). Car id 2 enters, goes to pit bay 3 (perfectly possible by going to pit lane 2 and then using the XCL to go back to pit lane 1), and opens the  pit menu (now the sequence is [2,1]). Now car id 1 close the pit menu and pushes car id 2 out of the pit lane while car ID 2 has still the pit menu open.

As for what happens to the pit activities of car being pushed, I would go for two options: either stops everything as it was when it was pushed over the sensor or give 100% on the activity in course.