On 7/3/05, Andrew Pimlott <[EMAIL PROTECTED]> wrote: > On Sun, Jul 03, 2005 at 01:53:45PM +0200, demerphq wrote: > > On 7/2/05, Andrew Pimlott <[EMAIL PROTECTED]> wrote: > > > On Sat, Jul 02, 2005 at 08:55:34AM +0200, demerphq wrote: > > > > The entire basis of computer science is based around the idea that if > > > > you do the same operation to two items that are the same the end > > > > result is the same. > > > > > > Citing "computer science" as the basis of your position is just too > > > much. The "computer science" answer to the comparison of references is > > > that they are equal if and only if they are the same reference. > > > > We arent talking about reference equivelence, but rather topological > > equivelence or graph equivelence. > > That's fine, I just object to throwing around "observationally > equivalent" or "bisimilar" to justify this.
I used the term structurally equivelent, and i wouldnt say that "observationally equivelent" is that far off. I didn't use "bisimular", Fergal did. > > > Finally, I think that comparing functions (which started this > > > discussion) is insane! > > > > And on the TODO list now. Its actually not that hard to do. > > It may be not that hard to determine whether two functions have the same > implementation, but of course you'll never determine (in general) > whether two functions have the same behavior. So such comparisons will > always be of limited usefulness--say, for testing (de)serialization, as > Michael said. Well I think thats a matter of opinion. But even if it does "just" help testing (de)serialization modules, thats still a good thing. Most of the serialization modules out there have some pretty signifigant flaws, probably because historically the best testing tool for this has been equivelent Data::Dumper dumps and DD itself is flawed. Yves -- perl -Mre=debug -e "/just|another|perl|hacker/"
