POSIX close(2) is broken

Re:POSIX close(2) is broken by JoeAltmaier on Saturday December 17, 2011 @08:06AM (#3364173)

Its time to stop reusing identifiers in kernels, libraries etc. There is no good reason not to use a uuid pretty much universally?

[ Parent | on HN ]

Re:POSIX close(2) is broken by ajross on Saturday December 17, 2011 @10:34AM (#3364380)

Other than compatibility with 30 years of software and the huge annoyance of using an identifier that doesn't fit in a machine word. The integer file descriptor is here to stay.

[ Parent | on HN ]

Re:POSIX close(2) is broken by JoeAltmaier on Sunday December 18, 2011 @08:55AM (#3366619)

New application environments are created all the time.

A new api that exports the true internal (uuid-based) identifier need not obsolete existing (broken) applications.

[ Parent | on HN ]

Re:POSIX close(2) is broken by ars on Saturday December 17, 2011 @08:01PM (#3365547)

uuid? Talk about severe overkill. It would be enough to simply not reuse fds, and just keep assigning them sequentially.

A uuid is a 128 bit number - you could assign them sequentially forever and not run out before the universe does.

Even 64 bits would be enough to last longer than the computer will no matter how fast you assigned them.

[ Parent | on HN ]

Re:POSIX close(2) is broken by JoeAltmaier on Sunday December 18, 2011 @08:56AM (#3366621)

Yes; but uuids are network-sharable, never expire (never get reused), can be conjured up without cooperation with anybody. You can merge two handle spaces without any issues, ever.

Its time to stop prematurely optimizing handles.

[ Parent | on HN ]