Darwinbots Forum
Code center => Bugs and fixes => Bug reports => Topic started by: Testlund on November 11, 2005, 05:18:26 PM
-
When gravity is set to 0 some things still falls down in version 2.37.5!
-
This seems to be fixed in the version [PY] uploaded.
-
This seems to be fixed in the version [PY] uploaded.
2.37.5 ... the version Testlund says he is using.
I find that gravity settings don't always 'take'.
don't know why ...
perhaps after changing it, he didn't hit 'change' to return to the sim.
who can say?
I do know that mutations ... ie ... the 2x or 4x slider ...
is not retained if one restarts ...
and is only there if one clicks 'change' to return to an already running sim.
so there are lots of those little quirks still to be explored.
with a sim that now runs pretty good without runtime errors ...
at least I can now start to experiment ...
to explore some of these bugs ... and report them.
who knows ... in another year or two ...
we might have 2.37.5 actually happening. :D lol
-
I DID click on change, start new, quitting, restarting the program, nothing helps. Perfect if one allways want to run in pondmode, but sometimes one might want to run in the other mode.
-
Can you give some instructions as to how this happened? I'll see if I can recreate it.
-
Well... I f you look at the picture I've added to my post above you can see I've set it to start with Triangulus to the left, veggies all over the screen and Animal minimalis to the right. Then I just move the slider for gravity to 0 and start a new simulation, and as you can see on the picture some things is effected by gravity that isn't suppose to be there. Animal Minimalis and all corpses falls down. The rest stays put.
-
did you go and check that gravity setting after doing the restart?
you may have to first select 'change' ...
and then go back and select start new.
or, after restarting ... check that grav setting again ...
and select 'change' to continue, not restart, the sim ...
and see what happens.
as you have no doubt discovered ...
there are many, many, many ways to screw up ...
I'm still exploring many of them ;)
-
Well as you so often say, garbage in -> garbage out.
Anyway, try as I might I can't reproduce this. My gravity slider starts on 0 anyway, if that's significant.
-
I can only say that I used to run it in pond mode, but I wanted to try it in non-pond mode for once. I loaded up my favorite settings file and unchecked pond mode and changed gravity slider to 0 and started a new simulation, then the above happend. I went into the GUI again, moved the slider back and forth to make sure it was set to 0, then clicked change. No change. I try to quit, restart... Nothing works. But I have given up on 2.37.5 and is running 2.4.9. This version seems to work, finally. Knock on wood!
-
Thanks, it was alot of effort to get it to do so for everyone.
-
Well as you so often say, garbage in -> garbage out.
Anyway, try as I might I can't reproduce this. My gravity slider starts on 0 anyway, if that's significant.
well it doesn't happen for me this way either ...
but then I haven't run pond mode.
just trying to give Testlund a shot at some things to double check ...
as not all setting do stay when you restart ...
but only when you click change.
as new start then resets them.
and ... not all settings are loaded in when saved ...
ad nums says ... most of them are ...
but it only takes one to screw it up ...
but screwed up is screwed up is screwed up ...
when one is trying to repeat consistantly.
-
There is (and always has been) a strange effect of gravity and tied robots.
It seems that the presence of ties negates gravity (and bouyancy for that matter). I still haven't managed to figure out why yet.
I suspect that the gravity settings are effecting everything and that Triangulus is simply exhibiting a reluctance to obey the natural laws due to the presence of permanent ties.
Another one on the "look into" list for the GUI.
-
We could work around this by attributing it to some strange physical relationship between ties and the ways in which gravity warps space time. Might be easier than tracking the bug down. :lol:
-
Probably is related to the way in which accelerations are imparted to other bots in a multibot.
-
That would be my guess. when you accelerate a bot on one end of a tie, the bot on the other end tends to get a shove in the opposite direction.
-
For those interested, this is specifically the reason I redid physics for 2.4. Ties behave weirdly prior to 2.4. There are (uncontrollable) ways to get a pair of bots accelerating across the screen with 0 nrg.
-
Tie phyics make my head spin :wacko:
-
Swimming physics still do.
I may just redo Carlo's old routine for them, inaccurate though it may be.
-
I know it's like doing things over twice but I may end up having to completely re-do the tie physics in 2.37 also
-
You can probably sort of adapt 2.4's source for that. Might be as much work as fixing the problem, I'm not sure.