The TORCS Racing Board
Username: Password: Remember Me?
Lost Password Register
Author: firechief | Created: 2017-08-22 16:27:16
Subject: A potential solution for Tiger/Mouse slowness
I realised that in my earlier (hurried) test, Mouse was fast whereas Tiger was slow - but I'd also botched the install of the new Mouse sourcecode for this race. When I copied that in properly, both robots were slow.

So that gave me the idea of trying it again with the sourcecode from the Hidden Valley race. Bingo! Both bots are fast.

If another solution doesn't present itself, perhaps running these robots with the old code might be ok in this instance? I can only speak for myself obviously, but I'd rather see every bot running at its best & I know its no fun for the author when things like this happen.
Last Edited: 2017-08-22 16:27:16 by firechief
    Author: timfoden | Created: 2017-08-22 16:54:09
    Subject: Re: A potential solution for Tiger/Mouse slowness
    That's a possible solution for sure... I haven't changed too much... but I think I've found the problem anyway..

    On line 2755 of MyRobot.cpp:

    #if defined(DEV) && !defined(BRAKING_TEST)

    should be:

    #if !(defined(DEV) && defined(BRAKING_TEST))

    .. so that the code lines 2756 through 2761 are included in the build.

    Cheers, Tim.
    Last Edited: 2017-08-22 16:54:09 by timfoden
      Author: firechief | Created: 2017-08-22 17:26:23
      Subject: Re: A potential solution for Tiger/Mouse slowness
      Excellent!
      Last Edited: 2017-08-22 17:26:23 by firechief
        Author: firechief | Created: 2017-08-23 05:03:24
        Subject: Re: A potential solution for Tiger/Mouse slowness
        Btw I wish I'd video-recorded this race instead of taking screenshots. There was FAR too much going on for a pictorial diary to do it justice.
        Last Edited: 2017-08-23 05:03:24 by firechief
          Author: dummy | Created: 2017-08-23 09:38:48
          Subject: Re: A potential solution for Tiger/Mouse slowness
          In my race one of my bots suddenly got confused and went into Crazy Ivan state, as if there wasn't enough mutual destruction before :)
          Last Edited: 2017-08-23 09:38:48 by dummy
            Author: timfoden | Created: 2017-08-23 11:50:50
            Subject: Re: A potential solution for Tiger/Mouse slowness
            And in mine at one point I think I saw a Klin going pretty fast in the wrong direction along the grass on the outside of the track. Not sure how it got into that state or what it was trying to do! :)
            Last Edited: 2017-08-23 11:50:50 by timfoden
              Author: firechief | Created: 2017-08-23 12:30:11
              Subject: Re: A potential solution for Tiger/Mouse slowness
              It had gone into super-Klin mode, and was trying to drive backwards as fast as possible in order to reverse time & undo all the damage...
              Last Edited: 2017-08-23 12:30:11 by firechief
              Author: dummy | Created: 2017-08-23 14:49:59
              Subject: Re: A potential solution for Tiger/Mouse slowness
              I had to laugh more in this race alone than the whole year before, it was this funny.
              Last Edited: 2017-08-23 14:49:59 by dummy
              Author: phi | Created: 2017-08-23 21:17:07
              Subject: Re: A potential solution for Tiger/Mouse slowness
              You guys are in a good mood :)

              Day after day I forget to fix that bug: Klin has a "blind angle" that makes him shoot in the opposite direction and stops only when he hits something; this can happen when he tries to recover from an accident.

              --------------------

              I think my race was not as drastic at the level of isolated events, but... we'll see what it will reveal...

              EDIT: I remember however that in my first race (though canceled), there were a lot of collisions and various confusions, and one of the DummyDrivers was in great trouble.
              Last Edited: 2017-08-23 21:55:35 by phi
      Author: dummy | Created: 2017-08-22 18:42:59
      Subject: Re: A potential solution for Tiger/Mouse slowness
      Thanks Tim. That's much better :)

      BTW fa1_2017 is about as slow in qualifying as you were before the fix. Can you confirm this?
      Last Edited: 2017-08-22 18:42:59 by dummy
        Author: timfoden | Created: 2017-08-22 19:39:04
        Subject: Re: A potential solution for Tiger/Mouse slowness
        Almost but not quite as slow, yes.
        Last Edited: 2017-08-22 19:39:04 by timfoden
      Author: phi | Created: 2017-08-22 19:10:55
      Subject: Re: A potential solution for Tiger/Mouse slowness
      I will run with these changes.
      Last Edited: 2017-08-22 19:10:55 by phi