Re: SGML tools aren't so great

[Gary Lawrence Murphy <garym@canada.com>]

>>>>> "G" == Gary Preckshot <garrell@inreach.com> writes:

    G> ... I've dealt with commercial publishers, and they
    G> usually want you to use Word with a macro package of their
    G> devising. However, the impetus for doing so is you get paid.

Not if you hold out ;) All the publishers know SGML is their future.
MCP will accept DocBook if you make the same case we've made here.
They want to stay in the dark ages because it is cosy and familiar,
but they also know that if they don't wrap their heads around SGML
today, they will be yesterday's business by tomorrow.

    G> Kevin Turner is the only one who correctly perceived that I was
    G> objecting to having to learn SGML just to write a doc.  He
    G> understood the two points that if you place barriers in the
    G> way, you only get volunteers of a particular sort, and that the
    G> concerns of some posters about the capabilities of SGML
    G> vis-a-vis other formats are best addressed internally.

As with the publishers, my advice to young programmers is to learn
DocBook as a survival skill.  In my younger days, it was ROFF and
then LaTeX, today it is DocBook.  

    G> ... If Stein had had a good GUI SGML editor, it would have
    G> presented him with a choice, and he wouldn't have to ask the
    G> question he did.

Oh, like the way GUI HTML editors intelligently ask you if you want a
<MENU> or a <UL>? :)  When was the last time you saw a <MENU> tag on
a webpage menu?

The parallel with HTML is very apt.  Do we expect to see pure ASCII
webpages or do we pretty much tune out such a page as not credible?

When was the last time you encountered a pure ASCII webpage?  Who does
all that HTML coding?  Did the HTML requirement for the web stop
people from creating the 800 million+ webpages out there?

DocBook is not that much more complicated for simple documents.

Quick: What is the HTML entity for the French c used in francais?
What is the entitity for registered-trade-mark?

    G> If you can't conceive of a way of either accepting input in
    G> other formats, or providing SGML tools that don't require three
    G> Hail Marys and rigorous adherence to an easily violated
    G> instruction book, then SGML isn't so great. 

By this reasoning, the world wide web will never catch on either.

If I check sources of webpages, I find a lot of META tags for keywords
and descriptions.  That's pure meta-information, and as soon as page
authors heard the first rumours that it may make their work easier to
find, even long before that was true, we found meta tags everywhere.

And the situation is the same for GUI tools.  You can get GUI HTML
editors, and a lot of people use them, but their pages are largely
inferior (there are exceptions) and most create pages that rank poorly
in search engines.

    G> Human beings are for doing things computers can't - like
    G> authoring.

800 Million pages of HTML.  Gee, it had me fooled.  Are they all
turned out by the same robot? I didn't realize that machines could
actually pass the Turing test yet.  Now, I know Veronica-2 has indexed
one or two million existing gopher pages so I guess that's where the
real people write ... ooops, no, scratch that: gopher pages have
markup too.  So where do these human beings publish?  Slashdot?

-- 
Gary Lawrence Murphy <garym@linux.ca>: office voice/fax: 01 519 4222723
TCI - Business Innovations through Open Source : http://www.teledyn.com
Canadian Co-ordinators for Bynari International : http://ca.bynari.net/
Free Internet for a Free O/S? - http://www.teledyn.com/products/FreeWWW/



--  
To UNSUBSCRIBE, email to ldp-discuss-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org