[ntp:questions] Re: Busted link on http://www.eecis.udel.edu/~mills/ntp/html/index.html

David Woolley david at djwhome.demon.co.uk
Mon Oct 13 06:46:36 UTC 2003


In article <3F8A249D.958F1E05 at udel.edu>, David L. Mills <mills at udel.edu> wrote:

> Yours is a breath of fresh air and hopefully has fixed the problem. I
> was using the # on both the reference and target, since that was what I

I wondered about this (although I found it difficult to believe), which
is why I wrote the following in a followup (on March 22nd) to the bug
report on this:

    Just for clarification.  The hrefs are correct; it is the id that is wrong
    for those hrefs.

> came away from the documentation with. My IE/NS tests used only two of

Which documentation?  The HTML specification even has worked examples,
even though, strictly speaking, URLs are outside of its scope.

> I have this expensive 1000-page "profesional HTML" reference book, but
> it didn't have the answers either.

Most popular computing text books are sources of bad advice.  They seem
to be written by people who are good at selling themselves to publishers
but have little understanding of the source documents and, I suspect, in
some cases have never read them.  For HTML, they may have limited value
in identifying things that are correctly and consistently implemented
in real world products, but they often advise techniques that are
in direct conflict with the design aims of HTML.  Although I haven't
checked it in detail, the only HTML text book I would consider likely
to be a reliable source is Dave Raggett's, because he wrote a lot of
the actual specification.

In particular, from a popular text book's point of view, fragments are
irrelevant to typical attempts to produce visual home page designs, so
they will be unlikely to have received any real consideration from the
authors.  The books are generally written from the point of view of
the wants of the prospective purchasers, which, for HTML often violently
conflicts with that of the language designers (i.e. visual v
structural markup).

This, though is a good example as to why it is pointless to tell people
to read RFC 1305.  Like you with the HTML specficication, most will not
bother reading a technical source document.  As you can see from the
HTML text books, any mass market documentation needs to be controlled by
the originators of the protocol; otherwise its contents will be largely
urban mythology.  If people use the product, mass market documentation
will be produced, with or without central editorial control.

You're also demonstrating the mass market's desire for cookbook
documentation, which says what to do, rather than what it means, even
though the latter would allow them to generalize to many more cases.



More information about the questions mailing list