Update: in the middle of page 3, when he integrates by parts, he probably forgets some absolute values in the inequality, namely one associated with (t-t0). That allows him to incorrectly strengthen the estimates. Indirect consultation: Peter Sarnak.
from:http://motls.blogspot.com/2008/01/lev-aizenberg-i-disproved-riemanns.html
Добавлено спустя 2 минуты:
Well the author's mistake is very simple. On page 3, he bounds |I_1| by
something + Integral of (t-t0) phi(t) f(t),
where f(t) is some positive function and phi(t) < C t.
However (t-t0) changes sign on the interval of integration, so it does not
follow, that this integral is bounded by integral of C(t-t0) t f(t).
--
Oleg Eroshkin
olegeroshkin (at) gmail (dot) com
http://groups.google.com/group/sci.math ... 0786834343