Give me a viable, non theoretical use case where the other target attributes and extension parameters would make a difference and I'll reconsider that constraint. On Nov 5, 2013 5:02 PM, "Julian Reschke" <julian.reschke@gmx.de> wrote: > On 2013-10-29 06:59, James M Snell wrote: > >> >> On Oct 28, 2013 10:48 PM, "Mark Nottingham" <mnot@mnot.net >> <mailto:mnot@mnot.net>> wrote: >> > >> > >> > On 24/09/2013, at 5:17 AM, James M Snell <jasnell@gmail.com >> <mailto:jasnell@gmail.com>> wrote: >> > >> > > Just a general FYI... I have submitted iteration -04 of the >> > > LINK/UNLINK draft with a few minor editorial fixes... and, I have >> > > formally requested Last Call status as an Independent Submission on >> > > the Standards Track. >> > > >> > > http://tools.ietf.org/html/draft-snell-link-method-04 >> > >> > In Section 2 of -05: >> > >> > "For any pair of resources, exactly one relationship of any given >> type can exist." >> > >> > That's a new and apparently backwards-incompatible change to the >> model of linking on the Web... e.g., consider "stylesheet". >> > >> >> No, read it again, as a uniqueness constraint on the tuple (resource, >> link relation, resource). That's not new or novel. >> ... >> > > Not convinced. > > Consider the various target attributes, or extension parameters. If they > do not contribute to the uniqueness constraint, you *are* constraining the > linking model. > > Best regards, Julian >
Received on Wednesday, 6 November 2013 01:07:10 UTC