Future of HTML

[sean at durak.org]

Date: Wed, 14 Apr 1999 20:16:44 -0700 (PDT)

sean at savvysearch.com: You are not subscribed to web4lib at webjunction.org.
Your message is returned to you unprocessed. If you want to subscribe,
send mail to listproc at webjunction.org with the following request:

		subscribe WEB4LIB Your Name

-------------------------------------------------------------------------------
John Creech wrote:
> I've been wondering recently about the "future" of HTML, about upgrading
> our website to HTML 4.0, and other such stuff, and would like to hear
> others' opinions on this.  My crystal ball went south, so I have no way of
> predicting any of this.

join the club :-) - and a very healthy question for discussion here
btw...

standard-compliant html is reasonably forward and backward compatible.
you probably don't have to ``upgrade'' any plain-old static web content
to continue making it available to newer browsers--just agree with your
content authors to adopt the new standard(s) on your web projects at a
convenient cutover point.

> As a very crude experiment, I went to the ARL website, called up the list
> of members, picked a spot in the alpha index at random, and visited 20
> library sites, just to see how the big dogs are handling 4.0/CSS.  Of the
> 20 I checked, 6 had implemented 4.0.  Many, like us here, are still
> declaring 3.2, while some had no DTD's.

you count yourself as big dog numero uno, yes? if you have a new
application that calls for adopting a standard before them, they'll be
turning to you for insight :-). 

CSS: if it helps any, i did checks on CNN, news.com, the major search
engines, and other sites this fall to look for use of CSS, i came away
with a list of popular web sites that have started to employ CSS to help
separate document content and structure from its appearance. 

CSS has an added benefit for high-traffic web sites: by replacing
stylized HTML markup by plain HTML and a single sitewide CSS, you can
realize some measurable (10-20%) bandwidth savings, which translates
into less money paid to the ISP for metered bandwidth. this fall i
checked browser distribution on some high volume web sites (3-15 million
raw hits per month) and found that they all had 85%+ browsers with
(alleged) CSS support. nb browser stats are rarely fair to the lynx
community, who log 1 raw hit for every html document, even when gui
browsers may be  pulling down 15 hits for the same document combined
with inline images.

> What do you all see coming up?  Is there going to ever be a "5.0"?  Are we
> moving to XML?  

brian kelly of UK web focus had a very helpful presentation on this at
ili99 last month -  http://www.ukoln.ac.uk/web-focus/presentations.html
- his big point IMHO is that the earliest adopters of XML will not be
writing for XML browser software -- you will be creating and storing
information in XML because it is a more intelligent format, but you will
still serve the information in HTML format for some time to come, via an
on-the-fly translating server in the middle, or by staging content from
xml back to an html/http server regularly.

> Are the benefits of upping to 4.0 worth the labor?

would not *upgrade* any old html documents unless you need some feature
of html4 in your content. if you implement a local search that can index
and retrieve docs based on dublin core metadata, that might be one cause
to go back and amend older docs (to add DC meta info), but i can't
immediately dream up a benefit to retro-convert a big batch of static
html 3.2 up to 4.0. if someone's got some good examples, even
hypothetical situations that call for this sort of upgrade, please post
them -- i'd rather think about it here than be surprised when it comes
up as an urgent issue with a consulting client :-)

--sean

mailto:sean at savvysearch.com                sean dreilinger, mlis
 http://www.savvysearch.com                http://durak.org/sean