Re: HTML+

lamport@src.dec.com
Thu, 15 Sep 94 12:14:14 -0700


Message-Id: <9409151914.AA07703@lilac.pa.dec.com>
To: tkn@VNET.IBM.COM
Cc: <www-html@www0.cern.ch>
Subject: Re: HTML+
In-Reply-To: Message of Thu, 15 Sep 1994 19:45:38 +0200
    <9409151636.AA09426@dxmint.cern.ch>
Date: Thu, 15 Sep 94 12:14:14 -0700
From: lamport@src.dec.com


   What I expect from a WWW browser is, that paragraphs are formatted
   according to my personal preferences of font, color and window size.
   Therefore the _viewer_ must be able to re-format paragraphs on the
   fly, and this is a point at which TeX is very bad suited.

I agree completely.

   Therefore, if I now adopt a system built on top of TeX with
   pre-formatting, one day our mind will change and we will want to have
   on-line reformatting.  

Correct--except it won't be a change of mind; we already know that's
what we want.

   And this functionality will put us at the same stage in development,
   where we are today -- only a few years later.

It will delay us by the time needed to hack up a TeX-based system,
which should be a few months, not a few years.  Since I estimate that
it will take at least 3 years to implement the system we want, this
will increase the time from about 3 years to perhaps 3.5 years.

So, what do we gain by this extra 1/2-year of work?  We gain the
possibility of actually succeeding.  If we undertake a 3-year
development program, by the time we have something working, there will
already be a dozen different systems in use.  The window of
opportunity will have passed, and we will have no chance of getting
everyone to adopt a single, sensible standard.

Leslie Lamport