On Wednesday 30 January 2008, Keisuke Nakano wrote:
> On 2008/01/30, at 21:57, Kuba Ober wrote:
> > I guess the contest designers have to decide one way or the other
> > on the "it
> > must be a correct program" constraint.
>
> The only constraint for correctness is 'to be accepted by the server'.
> The contest server automatically and immediately judges submitted
> programs
> whether it is correct or not. If the submission is correct, the
> server will
> add the author's name to the ranking.
>
> The error message approach, even if it is palindromic, will be rejected
> because the server does not observe the standard error.

I guess that settles it, then. The approach was nifty, though ;)

Cheers, Kuba