Re: mc-4.6.1 release?




On Sat, 23 Oct 2004, Leonard den Ottolander wrote:

> Hi Roland,
>
> On Sat, 2004-10-23 at 14:55, Roland Illig wrote:
> > I don't think the issue is that important in this case. A conversion to
> > an unsigned long should be enough in all practical cases. An unsigned
> > long has at least 32 bits, so the only case in which there could be name
> > clashes are when two distinct users on a system dont have a username and
> > when their (uid & 0xffffffff) are the same.
>
> Do I understand correctly that you want to introduce a cast because
> there's a bug in Cygwin? If so, I don't think that is a correct
> approach.

There is no bug on Cygwin as there is no bug on Linux. As you can
easily see for yourself both define uid_t as `unsigned' type. With
extremely large values greater that INT_MAX you will get wrong result
with the format type specifier %d. I quoted SUSv3 as a reference -
from the text it is clear that it is up to the vendor whether uid_t
is signed or not since it is not explicitly specified.




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]