Re: LYNX-DEV X-URL header field

[Uzi Paz]

On Wed, 19 Nov 1997, Al Gilman wrote:

> to follow up on what Uzi Paz said:
> > 
> > It is generated both if you use "mailto:"; and then your description is
> > fine, but also if you mail the web-page using the "Print" option, and then
> > it describes the location of the fil, as in the case of
> > "Content-Location:".
> 
> I agree.  On a "print" it would be appropriate to give the URL of
> origin as Content-location.

and I think that in this case X-URL should be replaced with
Content-Location which has a clearer meaning (Perhaps in the next
release?)

> One problem is the temptation to try to make the dictionary of
> field names be a flat partition and not a multilevel taxonomy.

I believe that you have something specific in mind when you say multilevel
taxonomy.

> "URL" as a field name is bad because it is the type of the
> attribute value and not its relationship to the thing for which
> it functions as an attribute.  But it is good because it covers
> the multitude of sins and the relationship is often inferrable
> from context.

If read by human. The usual problem with using non-standard header fields
is that their meaning is not well defined and then if you wish to extend
them or make some order so its possible for programs to undrstand and use
them, it's practically impossible because different practices are already
spread. In such a case, if you wish to register them you have to use a
different name.

The benefit from rfc2076 is the de-facto standartization of non-standard
header fields.

> Are you and/or Jacob doing anything about this currently?  Will
> it be discussed at the December IETF?  I think I am going to
> go.

It looks impossible for me to be in the next meating, although I would
like. 

- Uzi
;
; To UNSUBSCRIBE:  Send a mail message to address@hidden
;                  with "unsubscribe lynx-dev" (without the
;                  quotation marks) on a line by itself.
;