[uf-discuss] Re: a very early draft proposal hTagcloud

Chris Messina chris.messina at gmail.com
Wed Sep 20 12:22:57 PDT 2006

...Which is why I proposed using <big> and <small> tags: they actually
map to the function of a tagcloud: to visualize relationships between
discreet data.

There's a reason that we don't call them "taglists" but are instead
clouds -- essentially a graph! Hell, you should be able to derive a
pie chart from the data embedded in this microformat! ;)

Inasmuch as IE-blah won't support messing around with styling based on
title values (which seems like a bad idea anyway) more and more we
need to be cognizant of the accessibility on mobile devices as well
(as I write this on my Blackberry).

If again we look at the core utility of a tagcloud, it is to present,
*visually* the relative prominence of one tag in a system to another.
Any microformat for a tagcloud should reflect this aspect, should it
not? Otherwise, we're just talking about a taglist, which is another
thing altogether.

Now, to be clear, let me state that size is not the only way to
express said relationships, but it is the most obvious when it comes
to tagclouds; so long as the core use of the tagcloud is reflected in
the derived microformat, we won't be reducing the functionality when
presentational styling is removed.

Lastly, David, rel=tag is not appropriate for a tagcloud, as Tantek
mentioned, unless the local content is to be tagged with the tags from
the tagcloud. Phew! ;)


On 9/20/06, John Allsopp <john at westciv.com> wrote:
> Drew,
> > On 20/9/2006, "Stephen Paul Weber" <singpolyma at gmail.com> wrote:
> >
> >> Considering the fact that many tagclouds are based on actual numerics
> >> (ie, 50 bookmarks on this tag), it might be nice to have access to
> >> that information further than just a rating.  imho, styling is not
> >> the
> >> microformat's job, the implementor can add extra classes or whatever
> >> to make that work if title isn't good enough for that.
> >
> > It's true that styling is not the job of a microformat, but we must
> > seriously consider how it would be styled. If there's not a practical
> > way to style the tag cloud, no one will use it and we'd all be wasting
> > our time.
> I'd have to concur with that. When considering what I felt was, and
> subsequent discussion demonstrates is a reasonably tricky issue, that
> of how to properly markup the weights, or popularity of tags, I kept
> in mind
> 1. appropriate use of HTML
> 2. what developers are likely to actually adopt, based on an analysis
> of current practice, my anecdotal knowledge of widespread current
> practices, and a straw poll of some developers I know
> One important aspect of the latter is stylability. The title
> attribute of a link if styled with attribute selectors, won't
> actually appear styled in any current released version of IE for
> Windows. So to my mind that excludes using title (at least alone)
> because tagclouds would be effectively useless for the majority of
> todays web users, and so what developer would use title for this?
> It's an interesting issue. In the case of some microformats, for
> example rating is hReview, or XFN, it's not that big a deal. But
> tagclouds are somewhat different - the visual nature of them is a
> strong driver of their appeal.
> This raises a secondary issue, straight from the uf process
> http://microformats.org/wiki/process#Propose_a_Microformat
> If I looked at this microformat in a browser that didn't support CSS
> or had CSS turned off, would it still be human-readable?
> Are this format's elements stylable with CSS?
> In the first instance, in one sense it would be readable  - you could
> read the words. But in a sense, no, as the meaning of the tagcloud is
> lost - we just get an alphabetical list of words with no indication
> of how they compare with one another in terms of say popularity - or
> whatever the cloud represents.
> Are they stylable with CSS? Well, title is stylable thusly
> a[title="vv-popular"] {}
> But as observed elsewhere, as IE 6 and older on windows do't support
> it, is it in reality stylable?
> so, in fact, "styling is not themicroformat's job" does not entirely
> reflect at least the published process of developing a uf
> Thanks Drew and Stephen for the thoughtful comments
> j
> John Allsopp
> style master :: css editor :: http://westciv.com/style_master
> blog :: dog or higher :: http://blogs.westciv.com/dog_or_higher
> WebPatterns :: http://webpatterns.org
> Web Directions Conference :: Sydney September 28-29 :: http://wd06.com
> _______________________________________________
> microformats-discuss mailing list
> microformats-discuss at microformats.org
> http://microformats.org/mailman/listinfo/microformats-discuss

Chris Messina
Citizen Provocateur &
  Open Source Ambassador-at-Large
Work: http://citizenagency.com
Blog: http://factoryjoe.com/blog
Cell: 412 225-1051
Skype: factoryjoe
This email is:   [ ] bloggable    [X] ask first   [ ] private

More information about the microformats-discuss mailing list