| Conclusion | |
- HTTP 1.1: this is already here, and offers many advantages.
It is well-worthwhile switching browsers and especially servers to ones
that support this version.
- SSL: if security of data on the wire is a concern, it is worth
looking at servers that support SSL.
- HTML 3.2: use all the constructs of HTML 3.2 with the confidence
that nearly all browsers will support them.
- HTML 4.0: this is still in draft form. Nevertheless, many of its
features are already supported by many browsers. Any documents
with a long life could be written using these features. It is
tedious updating the document-type-version information on
each change, though.
- Cascading style sheets: this is the way to go - if you
aren't too concerned about appearance right now!
It will take a while to get stable support for style sheets from
the browsers, but it will make life much easier in the long run.
- XML: this is a major advance, but will probably take a long time
to become widespread. Parts of this need to be completed, and
browsers need to be XML-aware. Writing DTD's for particular
domains will be a specialist task, requiring much planning.
- Meta data: some sites are already loading keywords into meta-data
to take advantage of search engines. Some standards are under
development. Improving the quality of search engines is a
commercially active area, and waiting for a somewhat slow
resolution of the issues will be gazumphed by the search engine
vendors.
- Scripting: there are two competing proprietary systems here.
Although they hold much of the browser market, it is not all,
and some users turn off scripting support. CGI scripts called
out of browsers will still have to perform validation on their
input. Scripting can improve user response times quite dramatically,
though. Fun, but I would wait till DOM comes up with something
serious before investing heavily.
- Java: still in a state of change, which has not been helped by
the release of Netscape 4.0 with its peculiar version of Java.
An applet can replace HTML documents with a more traditional
GUI interface. The security model of HotJava will finally allow
applets to do useful things in intranets. What is missing is
a DOM link between Java and HTML.
| Slide 24 | ©Copyright 1997 | Jan Newmarch |