Tuesday, August 07, 2007

how high can it go?

this is troubling:

      i've been checking the proof of a particular claim,
in order to assure that it is a theorem.
it involves Lipschitz (continuous) functions.

      every time i edit the writeup,
i spot an error, but that's not the trouble;

      so far, i can always fix them ..
      .. but the Lipschitz constant gets bigger, every time. ):


EDIT (9 august 2007): never mind.
i found an error that i can't resolve.

nature: 1,   janus: 0.

2 comments:

Saara said...

Let's just hope that there are finitely many errors?

janus said...

as it happens, i finally found the big, proof-killing error.

my argument doesn't work. i assumed too much, and the special case which is true is something that people would laugh at.

so sometimes, a finite number of errors won't do. ):