W3C home > Mailing lists > Public > uri@w3.org > December 2013

Re: reviving the file URI scheme

From: John Cowan <cowan@mercury.ccil.org>
Date: Mon, 16 Dec 2013 02:11:17 -0500
To: Matthew Kerwin <matthew@kerwin.net.au>
Cc: uri@w3.org
Message-ID: <20131216071117.GC14870@mercury.ccil.org>
Matthew Kerwin scripsit:

> What I would really like is your opinions as experts, whether you
> think it's a worthwhile effort, or if my approach is suitable, or any
> specific issues (technical or editorial) with the ID itself.

Specific issues:

Fragments shouldn't be mentioned at all, because they are defined by
the media type of the entity (like text/html), not by the URI scheme.
However, it would be useful to note that media types are typically
inferred from the last component of the URI path: if the component ends in
".txt", then text/plain is inferred, and so on.

For "(some) Macintosh OS versions" read "older MacOS versions".

You should point out that modern MacOS file names are in Normalization
Form D, and consequently, conversion between Normalization Forms C and
D is appropriate when translating from file: URIs to MacOS paths or
vice versa.

> An alternative approach I've considered is creating an Informational
> RFC that "deobsoletes" parts of RFC 1738, since it's a bit unclear
> whether4248 (telnet) and 4266 (gopher) obsolete *all* of it, or just
> those scheme definitions.  If you think that would be a better (or
> worse, or silly) approach, I'd also like to hear so.

I see no benefit to this approach.

-- 
"Why yes, I'm ten percent Jewish on my manager's side."      John Cowan
    --Connie Francis                         http://www.ccil.org/~cowan
Received on Monday, 16 December 2013 07:11:41 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:16 UTC