>>
>> From: Gary Winiger <gww at eng.sun.com>
>>
>> I feel a lot stronger than Scott on this point. IMO, it is
>> misleading
>> to future maintainers. It also overly complicates the code, again
>> misleading future maintainers
Well I'll ignore your appeal to force and personal taste, and say that
I applaud Bart for actually taking code hygiene seriously; I cannot
see your argument for "confusingness" because it strikes me there is
greater clarity in taking a cheap personal copy of a bit of data that
should (but with modification or on other platforms, might not) remain
invariant - than there is in chasing down a Heisenbug in a few years
time when the code evolves into something else.
Also: clarity is why god invented commenting. :-P
I'm aware of nothing in the coding standards to cause a programmer to
dumb-down data hygiene, and it's not like this routine needs
fastpathing.
So - what's the real beef?
- alec [1]
[1] whom back in 1990 had to hack /bin/login to get round precisely
this sort of "nobody will change the content of the pw_passwd buffer
in this thread" misassumption, because the university i worked for had
to do 1000s of unix groups and I had to kludge an extra getpwnam()
search into a library call...
