Re: [Aide-devel] [Aide] aide version 0.12-rc1 released

[Richard v. d. B. <ri...@vd...>]

Marc Haber wrote:
> Not being able to reproduce a heisenbug does not mean that it is not
> present.

I totally agree, but I'd rather know you cannot reproduce it in 0.12-rc2
than having -V255 output from 0.11. If you could come up with a
reproducible test case (either 0.11 or 0.12-rc2) that would be best of
course. :-)

The real fix for this issue (and I'm sure others) is to get rid of all
the calls to abort(). That will ensure the program flow is
pre-determined and consistent, and cannot change at random like it seems
to happen now. That's not something I'll have time to work on soon though.

Sincerely,

Richard van den Berg