Microseconds Or Seconds ?

Till Now, some of us get the same results on the testing maps, and our scripts take the same time which is less than a second.
Is that means, judgement will be on microseconds?
If So, I think it will be not fair, because the same script doesn't take exactly the same time (in Micros) every time it executes. My script on a sample map take from (0.400 ~ 0.750).

yes that's true.
I think some new features must be added to the contest scripts because many of us will have the same script(same algorithm).

some condition like: make coridors so each room is connected and you can go from room[1] to room[50] never walking twice on a box.

-that's easy but ...
it's not much fun when the winner is chosen randomly because of those damn microseconds

you will be judged in microseconds..

but remember, that is the second judging criteria. first one is the cost of your corridors.

so, the short answer is: build a better script..

because this problem is NP-hard, there will not be a perfect script (because of time-constraints), so any algorithm even slightly better than the standard "minimum spanning tree" should put you ahead of all those that use that algorithm..

An example
There are 2 scripts.
In 40 maps they will find exact path lenght, but 1st script always is faster than 2nd.
In 10 maps 2nd will find better solution.

so question:
what script will win? 1st as winner 40 times, or 2nd as it has found better paths in total?

You get a score for each map
you get points for cost and score for time
I think the score will look something like
$score[$i] = 1/($cost+$time)
foreach ($score as $total){
$total += $total;
}
this is just an example. I think there is another algorithm , a bit different ...
so each of those 2 scripts can win

i think it would depend on other scripts, but i think that the 1st script will have a higher score..

again, build a better script. on every map, there will always be a better solution that the standard MST..

i dont think so. i havent found the time yet, to find out, how mst exactly works, but i think its quite similar to what my script does (reinvented the wheel again :/).
and on most maps my script seems to work perfect (as far as i can tell; but i can surely tell for the maps in the example pack!).

in your scenario i think mst will always win.
what i would do:
first judge per map; score only points (so maybe you have 5 scripts on first place and continue scoring with 6th place).
if theres a draw after all maps are done, rank by the sum of the times per map.

and i dont think, that you can wait with the descision until you see the scripts...

with scoring with time per map, i would try to write the fastest mst.
with a scoring like i suggested, i would try to write the script, thet gets the shortest paths (and if it takes 29.9 seconds).
i think the second is more interesting ;-)

look, i will be making the maps used for judging, and i will try to make each map so that it has a better solution than the standard MST.

so, i think it is better to build a script that takes much more time, but always does at least 1% better than the standard MST..

I got a perfect script. Never fails but it parses one map in 25-40 minutes.
Are you sure you want to constraint time to seconds and not minutes?

i doubt that it is "a perfect script", and yes, i am sure about the time limit of 30 seconds..

it tries every combination excepting some "absurd" combinations. So the result will be the minimum cost.
it's not perfect because it takes too much time

so, what would be the fun of the contest if i changed the time limit so that it is possible to write a script that simply tries all combinations?

hey, I was just kidding when I asked to change time. I'm not absurd. I just forgot to add a