@Sparr Yea, the point is for later entries to see to it that they do mesh well I suppose. Otherwise, their teammates won't be very happy with them.

I'm going to post something on meta later.

I can easily convert this to a 1v1 tournament-style koth if I have to, but I'd really like to see it team-based.

sure, but why does anyone care if their teammates are happy? hell, this contest would incentivize a sock puppet account for submitting a bad bot to the other team

or, slightly less maliciously, just convincing a bad bot author to switch to the other team :)

@Geobits now I kinda regret those two answer downvotes from earlier :D

@EoinCampbell how are the parallel trials going?

@MartinBüttner Parallelism is baaaad m'kay - gave up for the evening. branched the repo and kicked off another run (60 players / 1800 matches) on the synchro version.

ah okay, too bad

there's some bug in the PR i got where messages get stuck in parallel enumerator (the look like they've started, but they never get marked as complete) so the thing just keeps spinning. Will try sort it at the weekend.

anyone have tips for using files to store data when a contest controller might run multiple different matches with my entry at once?

specifically I'm interested in caveman duels

but others as well

write to a file called <GUID>.txt from each instance to avoid write concurrency problems.... read from *.txt if you need to get information from a bunch of previously run instances ???

caveman duels runs one instance for every round of a challenge

finding the right file to continue from the previous round would be challenging with that system

ah.

whoops I missed part of the spec for the date thing, so my score wasn't quite that impressive. but it was wrong before, so the absolute savings (of 32 characters) were actually correct.

Barely made it! :P

golfy :p

@Doorknob did he delete it right away?

Uhh

Looks like it

oh yeah, and could you do that superping on Lembik for me? that would be great :)

Ah, right, was checking my notifications and I just got to that one :P

running off for the evening. I hope someone submits a non-java entry for dogfight!

@lembik ^

thanks

@Lembik, when you see this: please undelete your challenge. The problem you posed was very interesting and the only problem with it was to determine how people could compete on it once the puzzle itself is solved. That's easy enough to fix, so don't be disheartened just because you immediately get 2 close votes. Closed questions can be fixed and are called "on hold" for a reason. ;)

@Doorknob are pings case insensitive?

@MartinBüttner Yes, but they don't work if the user hasn't been in the room for some amount of time

unless you're a site moderator, right?

hence, super-ping

@Doorknob yeah I know... what's confusing is that (I think) the autocompletion stops before the ping stops working

so there's no way to tell if the ping will work

@MartinBüttner Yeah, autocompletion is figured out locally in your browser, but whether to ping or not happens on the server. So it's a bit confusing.

ah okay

damn, I just can't figure out this 2048 challenge

I just always end up with an extra tile floating around that gets in the way

AH I have an idea

not sure if I can put that in code though

4-bit parity check solved in 32minutes now

why am I continuing this futile endeavor

:D

I'm still working on the 2048

so don't worry ^^

going to bed now though

see you

ciao