Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1407312

Re: [PATCH v1 1/1] sysctl: introduce uuid_le and uuid_be

From Theodore Ts'o <tytso@mit.edu>
Newsgroups linux.kernel
Subject Re: [PATCH v1 1/1] sysctl: introduce uuid_le and uuid_be
Date 2016-05-26 02:50 +0200
Message-ID <rCRER-7LQ-3@gated-at.bofh.it> (permalink)
References <rCojv-6cC-7@gated-at.bofh.it> <rCqY2-7Zt-23@gated-at.bofh.it> <rCr7H-84g-5@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


I'm also curious what !@#!? idiot came up with the concept of Little
Endian UUID's.  UUID's, and how to transform them from a printed
representation to a binary presentation, were well defined in a very
specific way in RFC-4122, which came from HP's Apollo/Domain OS, and
was adopted by the OSF/DCE, as well as later by Microsoft.  In all
cases, there was never any such thing as little endian versus big
endian UUID's.  Might as well talk about big-endian and little endian
IP addresses.  This way lies madness.

It's also the case that if all you need is a random UUID's, that
*technically* the endianness matters, but in actual practice, it
really won't matter.

					- Ted

P.S.  Let me guess, it was some clueless Intel engineer when they were
drafting the EFI spec?  Sigh....

Back to linux.kernel | Previous | NextPrevious in thread | Find similar | Unroll thread


Thread

[PATCH v1 1/1] sysctl: introduce uuid_le and uuid_be Andy Shevchenko <andriy.shevchenko@linux.intel.com> - 2016-05-24 19:30 +0200
  Re: [PATCH v1 1/1] sysctl: introduce uuid_le and uuid_be Andrew Morton <akpm@linux-foundation.org> - 2016-05-24 22:20 +0200
    Re: [PATCH v1 1/1] sysctl: introduce uuid_le and uuid_be Andy Shevchenko <andriy.shevchenko@linux.intel.com> - 2016-05-24 22:30 +0200
      Re: [PATCH v1 1/1] sysctl: introduce uuid_le and uuid_be Theodore Ts'o <tytso@mit.edu> - 2016-05-26 02:50 +0200

csiph-web