Re: sort -R: a more-conservative approach

[Bob Proulx]

Frederik Eaton wrote:
> Well, I guess it's only a factor of two, but there's a difference
> between bugs that cause a segfault in a tool that many people use, and
> bugs that might cause a new feature with 0 existing users to not be
> 100% cryptographically secure. That's just my two cents though

Your statement does not quite parse as I read it.

Sort is definitely a tool that many people use.  So agreed that
introducing bugs in a tool that many people use is a bad thing.  Also
agreed that this is a new feature and as such there are no previous
users and so pushing the inclusion of the feature out for a while
won't affect any existing users.  So that is safe.  Of what good is a
new feature that core dumps when run?  If a segmentation fault is
being triggered then it is a bug and adding the feature would be
adding a bug too.  That's bad.

If you are saying that since it only core dumps on 64-bit hardware
then I can only think that marginalizes everyone who uses 64-bit
hardware.  Such as myself since my normal equipment is a 64-bit
system.  But that does not change the fact that it also adds a bug
because programs should never segfault.

Bob