W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2009

Re: Multi-server HTTP

From: Sam Johnston <samj@samj.net>
Date: Fri, 18 Sep 2009 01:45:53 +0200
Message-ID: <21606dcf0909171645y48ef817ej47431a563e92a3e8@mail.gmail.com>
To: Anthony Bryan <anthonybryan@gmail.com>
Cc: Mark Nottingham <mnot@mnot.net>, Nicolas Alvarez <nicolas.alvarez@gmail.com>, ietf-http-wg@w3.org
On Fri, Sep 18, 2009 at 12:00 AM, Anthony Bryan <anthonybryan@gmail.com>wrote:

>
> >> 4.2.10.2. The "type" Attribute
> >>
> >>  metalink:metaurl elements MUST have a "type" attribute that indicates
> >>  the MIME type of the metadata available at the IRI.  In the case of
> >>  BitTorrent as specified in [BITTORRENT], the value "torrent" is
> >>  required.  Types without "/" are reserved.  Currently, "torrent" is
> >>  the only reserved value.
> >
> >
> > Overloading type like that is bad; register a media type (or get the
> > appropriate people to do it).
>
> This was from Julian's comment:
> http://www.ietf.org/mail-archive/web/apps-review/current/msg00155.html
>
> BitTorrent uses "application/x-bittorrent". What would I register?
> "application/vnd.bittorrent"?
>

Bittorrent specifies <http://bittorrent.org/beps/bep_0003.html>
application/x-bittorrent
which is ok by RFC 2045 <http://tools.ietf.org/html/rfc2045#section-5.1>:

Private values (starting with "X-") may be defined bilaterally between two
> cooperating agents without outside registration or standardization. Such
> values cannot be registered or standardized.


You don't have the authority to register the type on their behalf anyway
(unless of course you do) and to do so would almost certainly cause
interoperability problems for implementors and users.

I think using a non-standard MIME type rather than overloading "type" is by
far the lesser of two evils...

Sam
Received on Thursday, 17 September 2009 23:46:30 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:10 GMT