69513 - mozilla internal URI schemes should be "moz-"

Status: VERIFIED FIXED

(Core :: Networking, defect)

Description

[Dan Mosedale (:dmosedale, :dmose)]

In <news:[e-mail]>, there's a reference to a proposal at
<http://www.mozilla.org/directory/addrbook-refactoring-proposal.html> which
includes adding some new mozilla-internal URI schemes.  It would be nice if we
could do this in such a way that there was some assurance that we're not
breaking fowards-compatibility.  RFC 2396, at least, doesn't seem to provide any
guidance about how the URI scheme namespace is or should be managed.

Prefixing new URI schemes with moz- has been suggested.  Is this the reasonable
/ correct thing?

Comment 1

[Darin Fisher]

The suggested moz- prefix sounds good to me (x- may also be a good choice,
given that it's already in use for non-standard MIME types).

Comment 2

[J Luh]

IANA keeps track of registered scheme names:

ftp://ftp.isi.edu/in-notes/iana/assignments/url-schemes/

There is a Best Current Practice document titled "Registration Procedures for
URL Scheme Names":

http://www.ietf.org/rfc/rfc2717.txt

Comment 3

[Hixie (not reading bugmail)]

http://www.ietf.org/rfc/rfc2717.txt would seem to suggest that we would be best
off using a "moz-" prefix on our schemes. 

We should actually register the "moz-" prefix, but that document is unclear as
to how to go about doing that.

Reassigning to <[e-mail]>. If you can find out how to register
new trees, please do tell us, otherwise feel free to mark this bug FIXED, as
you found out the answer. Cheers!

Comment 4

[J Luh]

I don't know of any existing registration procedures for the prefixes. However, 
this (expired) Internet-Draft proposes the form vnd.moz.foo:

http://www.ietf.org/internet-drafts/draft-king-vnd-urlscheme-02.txt

Comment 5

[J Luh]

Using moz-foo: probably isn't perfect, but it seems to be the best course in the
absence of any further guidance/administration from IANA and the IETF.

Comment 6

[benc]

VERIFIED:
I think everyone is in agreement here, so I've changed the summary to reflect that.

We keep a list of badly named schemes in bug 130089.