in reply to Re: hash collision DOS
in thread hash collision DOS
I commented on a possible 'fix' for the potential problem with perls current hashing function at Re: Hash Clash on purpose and would be interested on your assessment of the notion.
If that isn't a fix, or cannot be implemented without undue penalty, I wonder if there isn't at least some mileage in having a new warning, along the lines of the 'Deep recursion' warning, that simply notes that a hash is exhibiting anomolous or 'worst case' behaviour at runtime. It wouldn't fix anything, but its presence in the logs could go a long way in trying to track down such behaviour down?
Maybe a corresponding pragma that allowed the 'hash quality' threshold that would trigger the warning to be specified?
Examine what is said, not who speaks.
"Efficiency is intelligent laziness." -David Dunham
"When I'm working on a problem, I never think about beauty. I think only how to solve the problem. But when I have finished, if the solution is not beautiful, I know it is wrong." -Richard Buckminster Fuller