On 11/24/17 5:26 PM, Jean-Christophe Deschamps wrote:
At 22:38 24/11/2017, you wrote:
One proof of the falsehood of your assertion is that we CAN fill a
database with some data using UIDs, and we will almost certainly not
get a collision, while you assertion we will.
This is an attempt at "proof by example". Keith is perfectly right
--mathematically speaking-- and your "proof" doesn't hold water, I
mean as a formal proof. The best proof that your "proof" isn't a
proof is that you feel obliged to add "almost certainly".
DISproof by example is a perfectly valid method. If someone makes a
claim that something is ALWAYS true, ONE counter example IS a disproof.
I said almost certainly as the chance of a collision isn't 0 (to be able
to say with certainty) but is most defintely less than the 100% claimed.
If SQLite was coded so that "SELECT 3.1415926" would "almost
certainly" return the expected value you just wouldn't use it. Same
thing can be applied to, say, a the classical Hello world! program and
a C compiler, but in this case experience teaches us to be even much
more careful!
I agree that you or anyone else _may_ consider the odds of UUID
collision(s) rare enough to ignore the issue and this is examplified
in practice by a huge number of systems using UUIDs or similar things.
But this doesn't make a proof of anything. That's the difference
between theory and practice. Search for a good quote in this list
about theory and practice ;-)
Considering that for a 'reasonable' number of records, (maybe even into
billions), the odds of a collision are probably on the order of once in
the life of the universe (I would need to run the math to get the exact
number, but it is minuscule in the period of the lifetime of a typical
computer), it seems to be a reasonable assumption. A big part might be
the consequences of a collision.
JcD
_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
--
Richard Damon
_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
