Meanwhile PHP will presumably do a similar set of fixes, though I haven't seen the announcement yet. There are workarounds for web apps that involve using extensions to limit the size of POST requests and/or the acceptable number of parameters in a POST request.
This is a problem that hashtables have in general (unless they are randomized).
_ In the case of the Ruby language, the 1.9.x branch is not affected by the predictable collision condition since this version includes a randomization of the hashing function._
So there is some merit to what the commenter is saying, though I doubt he knew the above.
Actual ruby arrays (which are arrays and not hashes) will obviously not exhibit this problem though.
I think for all practical purposes, unless you're doing something really weird, the likelyhood of hash function collisions is rare enough that we don't need to think too much about it.
I think for all practical purposes, unless you're doing
something really weird, the likelyhood of hash function
collisions is rare enough that we don't need to think too
much about it.
(Unlike PHP, though, the Ruby community can head off these particular attacks by releasing a new version of Rack, while waiting for a new 1.8.x release containing a security patch.)
On a side note, Ruby does not appear to have that same problem as similar code executes in .02 seconds. I think it is probably because Ruby's hash function is superior.