Re: 3 last issues with the URI syntax

Larry Masinter (masinter@parc.xerox.com)
Mon, 13 Jul 1998 00:39:04 PDT


From: "Larry Masinter" <masinter@parc.xerox.com>
To: "Roy T. Fielding" <fielding@kiwi.ics.uci.edu>,
Cc: <uri@Bunyip.Com>
Date: Mon, 13 Jul 1998 00:39:04 PDT
Message-ID: <000e01bdae31$4f5457c0$15d0000d@copper-208.parc.xerox.com>
In-Reply-To: <9807111024.aa28120@paris.ics.uci.edu>
Subject: RE: 3 last issues with the URI syntax

> If we were to do something like the above (which isn't really necessary
> since the layout form does not define syntax, but I understand the
> reasoning), then we would also need to show the other possible forms.
>   
>         <scheme>://<authority><path>
>         <scheme>://<authority>?<query>
>         <scheme>:<path>?<query>
>         <scheme>:<path>
>         <scheme>:<opaque_part>


Sure, I can go for this. Making it clear these are examples.

> >My suggestion is that query should be specified not with *uric, but instead
> >a group of characters which actually is allowed. According to what I can
> >see the rule should be instead:
> >
> >       query         = * (unreserved | escaped)
> 
> That is incorrect.  Reserved does not mean disallowed.  You will note
> that reserved characters are often included in components.  The only
> thing not included in components are the component separators.  Query
> does not have a component separator because it appears at the end of
> the URI string.

There's been some confusion about whether "?" is allowed in query strings,
so pointing out that it -- and other 'reserved' characters -- are allowed,
(but not uniformly implemented) would help.

> >(3) relative-path reference

I think we're converging on dealing with the three issues!

Larry