The W3C Secession: fixing the Web
Jorn Barger
Secession: HTML Forever"... but I decided that title would need to
include a critique of XML/XHTML, and when I started googling into
those topics, I quickly bogged down in technicalities that held no
interest for me.
So this time I want to start 'one step back' with the more general
question, "How is the Web broken, and is the W3C really qualified
to fix it?"
The bottom line of my critique is that the websites created by
W3C-standards-enthusiasts are generally awful-- unreadable--
because *communication* is actually a low priority for them, and
their highest priority is new technology for its own sake (XML,
stylesheets, etc).
I simply don't believe that poor communication can be improved
by technological fixes-- adding layers of complexity just makes
it worse. And W3C-tech always seems to demand that authors
toe the line and obey the new rules, before any benefit is
even possible-- but this is absolutely contrary to good human-
factors (and to human nature). If a design-solution is going
to succeed, it needs to fit people's existing work-patterns,
and *add significant value* at negligible cost.
What I _don't_ see emerging from the W3C is any technology that
actually makes any difference in my daily surfing-- the best I
think they can claim is that surfers whose browsers implement
stylesheets 'correctly' see layouts that are closer to the
authors' intentions... _if_ the authors used stylesheets
'correctly'.
But for all its limitations, HTML has always been able to
present simple, clear _communication_ if the author bothered
to understand those limitations and work around them.
The two great challenges: 404s and search
So when I visualise the future direction of the Web, I try to
imagine a world where all pages are basic HTML, made by
authors who are genuinely trying to communicate, and I ask
what new tools might be offered that would really assist them
in this task.
To me it seems the two greatest challenges are 404s and search.
I think Xlink was supposed to be the W3C's fix for 404s, and
XML for search. But I don't see any glimmer of hope that these
techno-fixes will ever work, and I'm not willing to wait!
Whenever anyone does a search, they go thru the same basic
process: try a searchpattern, scan the results, load each page
that looks promising, and maybe refine the searchpattern until
the desired information is found.
From the point of view of information theory, this sorting-
process should reduce the overall 'entropy' of the Web-- but
instead the whole effort is usually thrown away, and only the
final information, or at best a bookmarked URL, is retained
and possibly republished.
So what I propose is that authoring-tools be developed that
(partially) automate the task of _publishing_ this sorting-
process: imagine that every time you do a search, a webpage is
created that remembers what searchpattern(s) you used, what
results the searchengine returned, which results-pages you
checked, and most importantly a brief comment on what you
thought of each of these. (In a previous thread I called this
'action hypertext': [1])
Even at this primitive level, the pages you generate ought to
provide a useful guide to future searchers, at least by
increasing the search-engine ranking for the most useful pages
found.
But skillful author-researchers will be able to refine these
primitive 'action hypertexts' into a detailed overview of
_all_ available web-resources for a given topic-- and such
hand-built overviews should become the backbone of the real,
practical semantic web. [2]
But this will make 404s a much more serious problem, because
where the search-engines themselves are continually weeding
out 404s, the utility of these web-resource overviews will
inevitably (and quickly) be dragged down by the weight of
vanished pages.
But again, rather than trying to solve this by a technological
fix that requires all web-authors to maintain their links in
a different format (eg Xlink), I propose an authoring tool
that leverages its action-hypertext record of the original
searches, and actively monitors for 404ed-links, so that the
task of maintenance becomes almost a matter of a single click:
you tell the authoring-tool you're in the mood to maintain
page X, and it not only lists the expired links, it also
re-runs the original searches and even suggests likely
substitutes.
The third great challenge: html-junk
There's one more way I think the Web is broken, that W3C
standards can't really fix-- the burden of webpages that render
slowly and bewilder the reader with distractions ('html-junk').
I'd also like to see this dealt with by an intelligent tool,
instead of a (W3C-style) new set of authoring-rules.
I've proposed a browser called the 'DrawBack project' that
would allow surfers to easily reformat others' page-designs,
via 'smart local stylesheets'. [3] This could make web-
browsing more like reading netnews in trn: you'd get offered
a menu of new 'postings' and select the ones you like, then
almost instantaneously page thru them (in trn this could be
done mostly using only the spacebar).
[1] http://groups.google.com/groups?th=dc41049d34060194
[2] http://www.robotwisdom.com/web/
[3] http://www.robotwisdom.com/web/drawback.html
ian glendinning
correctly.
ie That "source style" be filtered out to leave the "parsed source
semantic content" and then a local "receiver" style and transformation
re-applied at the browser to suit the purpose and interpretation of
the user, not the originator.
This is the holy-grail as far as I am concerned, and worth striving
for. I'd be very interested to see if your "drawback" idea gets us
towards this goal. Apart from the usual issues about true semantics
and finite vocabularies of schema tags, there are several purely
pragmatic obstacles to progress I guess.
Thinking out loud .... Most people generating content still, and for
some time, will be using fairly simple HTML tools, and will therefore
add their own style and features in a noisy way that gets jumbled up
with the essential "communication". I know, I'm one of them. Beauty
and the message are of course then in the eye of the beholder.
I know it's not what you are suggesting literally, but it would be
nice to turn the clocks back, or stop time for a period, so that the
components of an ideal solution could be applied in the right order.
However I see it as a fact that the XML / W3C technology is advancing
faster than the re-invented ideas can grasp what the basic underlying
communication issues already were. Similarly, despite the advance on
the techie front, there is still far short of a critical mass in
people actually using intelligent XML tools, keeping content, schema
content, transformations and styles strictly separate anyway. There is
something Darwinian / memetic about the way ideas about technology and
useage are pervading the web. I'm almost as interested in how
solutions arise from that state of affairs, as I am in what a good
solution actually is.
I'm interested in seeing other responses to your ideas Jorn. In the
meantime I'll see if I can make use of the stuff on your "drawback"
page.
Ian Glendinning
++++++++++++++++++++++++++++++++++++++++++++++++
jo...@enteract.com (Jorn Barger) wrote in message news:<16e613ec.02070...@posting.google.com>...
Gordon Joly
> Secession: HTML Forever"... but I decided that title would need to
> include a critique of XML/XHTML, and when I started googling into
> those topics, I quickly bogged down in technicalities that held no
> interest for me.
>
> So this time I want to start 'one step back' with the more general
> question, "How is the Web broken, and is the W3C really qualified
> to fix it?"
Well, maybe they are at a loss. After all, the legacy of HTML is huge.
>[...]
>
> The two great challenges: 404s and search
>
> So when I visualise the future direction of the Web, I try to
> imagine a world where all pages are basic HTML, made by
> authors who are genuinely trying to communicate, and I ask
> what new tools might be offered that would really assist them
> in this task.
>
> To me it seems the two greatest challenges are 404s and search.
>
> I think Xlink was supposed to be the W3C's fix for 404s, and
> XML for search. But I don't see any glimmer of hope that these
> techno-fixes will ever work, and I'm not willing to wait!
> [...]
XML is not for search. RDF, etc and of course the "Semantic Web" are
the next generation of search techniques, to supplement our own
research skills (which used to be based in library indexes citations
etc etc
The Semantic Web is still being "held back" by W3C, IHMO.
Seems that they "want to get t right first time"....
Ho hum.
Gordo
Gordon Joly
Gordon Joly <gordo...@pobox.com> wrote:
>
>XML is not for search. RDF, etc and of course the "Semantic Web" are
>the next generation of search techniques, to supplement our own
>research skills (which used to be based in library indexes citations
>etc etc
>
>The Semantic Web is still being "held back" by W3C, IHMO.
>
>Seems that they "want to get t right first time"....
>
>Ho hum.
>
>Gordo
Seems that they "want to get it right first time"....
(!!)
Gordo