Re: [Uri-review] Request for review of "ab:" URI scheme

[Barry Leiba <[e-mail]>]

>    The following requests IANA to register a new URI scheme according to
>    the IANA registration template specified in [RFC4395]:
>
> I see RFC 4395 is normative reference here.  I do not think reading this
> document should be compulsory for understanding your document.  Maybe
> Informative reference is more appropriate here.

Clearly not; see below.

> I understand this is defined with ABNF.  Your document in Section 1.1
> mentions this, but I think mentioning this issue once more in the template
> wouldn't be harmful, as I propose below.

RFC 4935, along with RFC 3986 (which 4935 references) makes it clear
that URI schemes are defined with ABNF.  I see no reason to add
anything further here.

> You can also add something like "The "ab" URI takes the form of <paburi>
> rule below, defined using ABNF [RFC5234]."  I also do not understand why the
> scheme name is "ab" whereas the rule is "paburi".  If you decide to change
> the scheme name per comments from Björn, this rule should also be corrected.

The names of the ABNF elements aren't related to the text strings that
they produce.  There's no reason for any change here.

>    URI scheme semantics:  "ab" URIs are used for designating references
...
> This section should also have some words about the role of <addrbook> and
> <extensions> parts in the URI.

You're right; I thought that was already in there, but it's not.  I'll
add something.

>    Interoperability considerations:  Applications are only REQUIRED to
>        support "ab:default".
>
> I cannot find any distinctive description of this reserved URI, information
> about what it is used for.  Having at least sentence dealing with this issue
> would be OK, I think.

It's in the second paragraph under "Intended usage:".

>    Security considerations:  Applications SHOULD ensure appropriate
>        restrictions are in place to protect sensitive information that
>        might be revealed by "ab" URIs from access or modification by
>        untrusted sources.
>
> I do not see anything regarding the URI scheme in your section 3.  While it
> concerns the proposed extension, some words specifically about the URI
> scheme won't be redundant, IMO.

Yes, that's reasonable.  I'll add.

>    Contact:  Sieve mailing list <[e-mail]>
>
> Shouldn't this be "Sieve WG <[e-mail]>".  Maybe the same is below - it
> should be just IESG, shouldn't it?

Huh?  I don't understand what you're getting at.  The contact
information is correct, and isn't related to who has change control
(the IESG).

Barry