Very annoying mpz_get_str function Problem !
codrin at flynet.iasi.rdsnet.ro
Wed Nov 26 22:14:19 CET 2003
Maybe I was a little angry but I have spent a lot of time trying to debug
that behavior. I'll do my best to find a minimum example to prove my
complaints. As I found until now it seems that the function is making
problems with bignumbers that are the power of 2, for example generating
random numbers between 0 - 2^1024.
On Wed, 26 Nov 2003 DTAshley at aol.com wrote:
> TG is right. You failed to provide a minimal test case demonstrating the potential defect for anyone else to look at. Additionally, I concur with TG's logic that the probabilities favor the bug being external to the GMP.
> Try calling Microsoft technical support for one of their products (have a credit card handy), and unless it is a known bug they are going to want the same thing: some way for them to reproduce the behavior. There was nothing unreasonable about TG's observation.
> Those are the rules of the game.
> "Sucks" doesn't make the grade. It has to be "sucks", combined with "and here is a way for you to get the same behavior".
> In a message dated 11/26/2003 2:03:27 PM Eastern Standard Time, The Codrinus <codrin at flynet.iasi.rdsnet.ro> writes:
> >YES ! And I will not use the buggy mpz_get_str function or GMP in the
> >future !
> >sorry about bothering you,
> >On 26 Nov 2003, Torbjorn Granlund wrote:
> >> Please don't bother us with vague complaints like this. The bug
> >> is almost certainly in your code. Debug your code, and then if
> >> you do have a test case for a real bug in GMP, send it in.
> >> Until then, we don't want to hear any "sucks" from you.
> >> What do you think is more likely, that your newly developed code
> >> has a bug, or that the carefully tested mpz_get_str (or its
> >> descendants) has a bug?
> >> --
> >> Torbjörn
More information about the gmp-discuss