Re: Adobe's PDF

Steve Heaney (
Wed, 21 Jul 1993 11:41:59 +0200

> From Tue Jul 20 23:06:02 1993
> Date: Tue, 20 Jul 93 15:54:50 -0500
> From: (Marc Andreessen)
> To: (Roy Smith)
> Cc:
> Subject: Re: Adobe's PDF
> X-Md4-Signature: 71d00992f2d205981eda3c5eded11ebd
> Content-Length: 487
> Roy Smith writes:
> > >There are many other concerns related to document interchange than those
> > >being addresss by the WWW community. The presentation of the document
> > >is paramount to many publishers and they would not consider the use of
> > >something like HTML/+ to distribute their publications.
> >
> > I really do have to agree with this. HTML is the kind of thing that can
> ^^^^
> > only be loved by a computer scientist.
> Yup.

Pthptt !!

I would agree with the statement that many publishers would not consider the
use of HTML or HTML+ - it may not be sophisticated enough for their
requirements in terms of the elements it contains.

But don't confuse that statement with Marcs assertion that this applies
equally to SGML. SGML allows you to define _any_ DTD - HTML is a specific
DTD designed for the WWW project.

SGML was designed specifically for publishers. It allows you as much
flexibility as you want in determining the presentation style of a document.
If necessary, processing instructions can be added to a document, the
NOTATION construct allows embedded data of other types and by abstracting
the information about formatting of elements, you can display it 101
different ways if thats what you want.

It certainly does appeal to computer scientists because of the power of
the idea. In a project I am working on, we are using one source document
coded using SGML to provide online documentation, printed documentation
(several versions - reference manual, quick reference card etc), context
sensitive help for a GUI, and the program the SGML document documents even
reads the SGML itself to check information about input parameters. Do that
with PostScript or PDF :-)

> Marc


Steven Heaney
SQM Group
Schlumberger Geco-Prakla