W3C home > Mailing lists > Public > public-html@w3.org > November 2011

Re: noted 3 issues re: time/data (was Re: minutes for HTML WG f2f, 2011-11-04, part 1)

From: Ian Devlin <ian@iandevlin.com>
Date: Tue, 15 Nov 2011 17:08:42 +0000
Message-ID: <CAOYOhSu3dcHL_kZ25DUHh3170gWCKZEuvUdgG9UWABmBZzx8CQ@mail.gmail.com>
To: "Tab Atkins Jr." <jackalmage@gmail.com>
Cc: Sam Ruby <rubys@intertwingly.net>, Peter Winnberg <peter.winnberg@gmail.com>, Tantek Çelik <tantek@cs.stanford.edu>, public-html@w3.org, Maciej Stachowiak <mjs@apple.com>, Paul Cotton <Paul.Cotton@microsoft.com>
Hi,

My example was more to illustrate a scenario where a page may have multiple
<time> elements and anything that's traversing that page, for example a
search engine spider, couldn't possible know which one of the <time>
elements on the page in question was representing the publish date, and
therefore which one is the relevant one (in this case).

> Note, though, the conflict between "publication date" and "last modified
date".  The former never changes, the latter does, and the latter is always
the same or following the former.

Very true, and it is an issue. I would propose either:
1. pubdate should always represent the most recent date for the page,
whether it be when it was published or when it was modifed.
2. Add another attribute such as moddate which represents the modified date
which then allows both sets of data to be retrieved.

Regards,

Ian

On 15 November 2011 15:56, Tab Atkins Jr. <jackalmage@gmail.com> wrote:

> On Tue, Nov 15, 2011 at 12:55 AM, Ian Devlin <ian@iandevlin.com> wrote:
> > Hi,
> >
> > I'd like to oppose the intention to drop the pubdate attribute from the
> time
> > element. Whilst like some other of the HTML5 elements and attributes,
> it's
> > probably currently not being used for anything useful just yet, I can see
> > situations where it would be useful to automatically know which time
> element
> > on a particular page is the published date.
> >
> > Say for example there was a page with a list of tour or event dates. This
> > could of course consist of any number of time elements, each containing
> the
> > specific date and time of a concert or event. In amongst all of these
> time
> > elements, is also the publish time/date of the page itself and surely it
> > would be useful to be able to distinguish that one, quickly, from amongst
> > all the others?
>
> This doesn't seem useful - it appears that you're talking about a
> human reading the page here, in which case you don't want a @pubdate,
> you want some actual text that says "Last Updated: 1 April 2011".
>
>
> > Similarly when returning search results would it not be more appropriate
> for
> > the most recent (and therefore relevant) results to be returned, and the
> > pubdate attribute could contribute to this as it could be used to
> determine
> > which of the pages are the more recent. Google touch on this usefulness
> > themselves on their blog:
> >
> http://googleblog.blogspot.com/2011/11/giving-you-fresher-more-recent-search.html
>
> This seems potentially useful as a signal.
>
> Note, though, the conflict between "publication date" and "last
> modified date".  The former never changes, the latter does, and the
> latter is always the same or following the former.
>
> ~TJ
>



-- 
ian devlin
w: www.iandevlin.com
e: ian@iandevlin.com
t: iandevlin <http://www.twitter.com/iandevlin>
skype: idevlin
Received on Tuesday, 15 November 2011 17:09:17 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:17:41 GMT