[uf-discuss] invisible metadata

[Chris Messina]

I think the practice of defining hidden meta falls into the
pornography assessment rule: I know when I [don't] see it.

I'm just going to propose a metric that might be useful here... when
embedding meta data using microformats in a way that is usually opaque
in the viewport, I think that it's ok if having that invisible data
helps to clarify a relationship or remove ambiguity. For example, with
the <abbr> tag used on times in hCard -- we use the value of the title
attribute to remove the ambiguity raised by colloquial language like
"half past six".  In this case, it's not hidden meta data; it's a
means to clarify the data that is already visible.

In the XFN example, we're starting to border on invisible metadata.
However, I would argue that this "data" can be exposed in a useful way
via CSS. For example:

a[rel=friend] { color: green;}

In that way, we are able to use that data to disambiguate or clarify
the visible data. It's also the case with vote-for and vote-against,
which you could style, adding semantic *visual* data to yoru content.

In any case, a good example of truly invisible (and therefore what you
*should not* do) would be hiding the address content of an hCard
because you only want to list a person's name. Just because you could
include their full vcard as display:none microformatted content
actually defeats the principle of distributed design. You shouldn't
need to encode the entire vcard everytime you list the person's
name... encode it once, and then mark up your usual content in
semantic ways -- let the machine pull the data together.

Anyway, that's my take on this topic because it is an extremely
important one. Microformats should be designed for humans first,
machines second.

Chris

On 11/19/05, Brian Suda <brian.suda at gmail.com> wrote:
> I think the difference in your XFN example and in the hCard example is
> that you are making  the VALUES invisible. I know i'm guilty of this,
> but when your data is out-of-site, it is out-of-mind. Things have a
> better change of drifting away if you are not staring them in the face
> each day.
>
> If i have a falling out with someone and want to remove them as a
> 'friend', with the XFN example their name is VISUALLY on my page, with
> hidden data that i do not see each day, i am more apt to forget about
> it.
>
> The VISIBLE data we are refering to are the VALUES.(i'll probably
> catch some heat for that last comment, so i'll mention that i am fully
> aware that our ABBR encoding sort of contradicts what i just said, but
> it is a representation of the Value, so it is still visible) In the
> XFN example that would be "Mr Bloggs"
>
> -brian
>
> On 11/19/05, Andy Hume <andyhume at thedredge.org> wrote:
> > Hi
> >
> > >>> As said before, microformats encourage the use of *visible data*,
> > >>> not
> > >>> invisible (meta)data, and thus microformats agrees with the
> > >>> position that
> > >>> you should NOT use CSS to hide microformatted content.
> > >>
> > >> This is a concept I hadn't heard voiced before, and I guess I agree
> > >> in principle.
> > >
> > > Yes, this is a *very* important aspect of microformats.
> > >
> > >> However, microformats could be powerful tool for marking up invisible
> > >> meta data.
> > >
> > > The problem is that invisible metadata is nearly worthless.
> >
> > It's not quite as simple as that, surely? Invisible metadata
> > shouldn't be worthless, although I do understand the arguments
> > against hiding it.
> >
> > <a href="http://myfriend.com" rel="friend met colleague">Mr Bloggs</a>
> >
> > These XFN values are essentially invisible to humans when rendered by
> > a browser. Explain how this is different to hiding something with CSS.
> >
> > Apologies if I'm going over old ground here.
> >
> > > See previous
> > > writings on this re: meta keywords etc.
> >
> > Very interesting, and I understand the parallels with meta keyword
> > values, etc. I *have* taken this on board, however my question above
> > still stands.
> >
> > Cheers,
> >
> > Andy.
> >
> >
> >
> > >
> > > _______________________________________________
> > > microformats-discuss mailing list
> > > microformats-discuss at microformats.org
> > > http://microformats.org/mailman/listinfo/microformats-discuss
> > >
> > >
> >
> >
> > ______________________________________________________________________
> > This email has been scanned by the MessageLabs Email Security System.
> > For more information please visit http://www.messagelabs.com/email
> > ______________________________________________________________________
> > _______________________________________________
> > microformats-discuss mailing list
> > microformats-discuss at microformats.org
> > http://microformats.org/mailman/listinfo/microformats-discuss
> >
>
>
> --
> brian suda
> http://suda.co.uk
> _______________________________________________
> microformats-discuss mailing list
> microformats-discuss at microformats.org
> http://microformats.org/mailman/listinfo/microformats-discuss
>