Re: Date/time tag??

• Mail folder: WWW Talk Jul-Oct 1993
• Next message: David Martland: "Re: yet another tag"
• Previous message: frans van hoesel: "yet another tag"

From: Dave_Raggett <dsr@hplb.hpl.hp.com>
Message-id: <9308021335.AA17646@manuel.hpl.hp.com>
Subject: Re: Date/time tag??
To: p.lister@cranfield.ac.uk
Date: Mon, 2 Aug 93 14:35:46 BST
Cc: www-talk@nxoc01.cern.ch
Mailer: Elm [revision: 66.36.1.1]
Status: RO

Thank you Peter, for raising the issue of representation for Date/time.

So far the HTML+ spec only mentions date/time in regard to the INPUT tag.
You can restrict the input field with the TYPE attribute:

    The following types of field are supported:

    ...
    date    restricted to a recognised date format
    ...

The spec for HTTP/1.0 mentions a "Date:" field in the RFC822 format for
the date the document was last modified. This is useful for checking if
a document has been updated, e.g. for automatic updating of indexes.

This leaves open:

    a)  identifying the date/time of last update for documents
        retrieved by protocols other than HTTP

    b)  specifying the "recognised date formats" for input fields

    c)  canonical representation for date/time information within documents

(a) could be addressed by supporting an update element in the header section.

One solution for (b) would be to restrict support to the RFC 822 format.
Another idea is to prohibit giving the month as a number. This can be avoided
if authors state the desired format in the document as part of the normal
text.

(c) could be handled by making "date" a recognised role for logical emphasis.
Browsers would then have the option of displaying the date as given in the
document or reformatting it according to local preferences. Alternatively,
a new "date" tag could be introduced for the same purpose.

My proposal for HTML+
---------------------

    o   To ignore (a) and recommend that HTTP servers all return
        the "Date:" header for the date/time last updated. One can
        also get at the date/time for FTP access in most cases.

    o   To ignore (b) and recommend that authors disambiguate the
        desired format as part of the document's normal text.

    o   To add "date" to the list of recommended roles for logical
        emphasis and suggest that it is used for RFC 822 dates only.

What do other people suggest?

Dave Raggett