Re: [apps-discuss] APPSDIR review of draft-ietf-core-link-format-11

[Julian Reschke <[e-mail]>]

On 2012-02-28 21:13, Carsten Bormann wrote:
> On Feb 28, 2012, at 20:49, Julian Reschke wrote:  A really good, comprehensive review.
>
> Let me grab my implementer hat and pick three of the items.
>
>> I would recommend to allow both token and quoted-string for all new parameters.
>
> I whole-heartedly agree.
> I'd really like to lose the code commencing:
>
>      MUSTBEQUOTED = map_to_true(%w{anchor title rt if})
>
> (The same issue appears to be apply to »anchor« and »title« in RFC 5988.  And I probably even missed some others that another implementer will take for granted, likely to cause some immediate interop issues.  Please raise that erratum now...)

Good to hear I'm not alone with that complaint :-)

>> [...] Which makes it impossible to use "uri" as link attribute. Maybe this should be noted. Alternatively maybe use a format that doesn't require overloading the name.
>
> We could use href as the ersatz attribute name, or something weird that is not allowed by the »parmname« production.  Please also see draft-bormann-core-links-json-00.txt, where I had to solve the same problem (and tried to solve it in a way that it causes the same damage this use in link-format already causes).

Maybe the empty name?

>> I note that hardwiring URI query parameters into the protocol is *not* Restful.
>
> Yes, we are painfully aware of that.  A better way has not emerged yet.

Ack.

> Again, thanks for a very good review.  Let's now work on making the points actionable.

Genau. Schönen Abend noch,

Julian