[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Groff] Eric Raymond on groff and TeX

From: James K. Lowden
Subject: Re: [Groff] Eric Raymond on groff and TeX
Date: Thu, 3 May 2012 14:15:27 -0400

On Thu, 3 May 2012 08:51:29 -0700 (PDT)
Dorai Sitaram <address@hidden> wrote:

> While it may be possible to mimic the tags of structural markup in a
> presentation-markup language, there is power in completely and firmly
> separating the two aspects: you can then independently develop the
> two aspects; indeed the typographic aspect can come to embrace
> completely new and even unexpected media whereas the
> structural aspect remains the same.  

That's the promise, surely, but it's not borne out in practice.
Especially that bit about "unexpected media".  Markup languages come
and go much more often than media types do.  (Although I've met a few
media types I could do without.)  ;-) 

Consider two pipelines: 

        macro -> troff -> ditroff -> device
        markup -> style -> device

Both allow user input to be semantic rather than presentation, cf.
-mdoc and Docbook.  And both, of necessity, provide device-specific
escapes.  Therefore the claim that they're intrinsically different is

Conversion complexity is similar, too.  Writing troff macros and SGML
stylesheets are both significant undertakings.  XML sought to simplify
styling by using a unified grammar -- it's all XML! -- but the problem
was never syntactical.  

It's interesting to note that SGML, TeX, and troff were all invented to
deal with *print*, the goal being to isolate the user not from "all
media" but from particular printers.  They succeeded because printers
are interchangeable.  

I have never seen a single-source document that looks equally good
as Postscript and as HTML.  If the author had paper in mind, the HTML
suffers from a lack of hyperlinks and from assumptions about
the display.  If the author intended HTML, the printed version lacks
finesse, if indeed there *is* a printed version (because, promise
notwithstanding, it's not that easy to do).  

I suspect the reason is that the document is art, and author artist.
Every author has in mind an intended audience, and that audience has a
media preference that derives from how the document will be "consumed",
as the media types would have it. Those preferences inevitably are
manifested in the document because different media have different
characteristics that are intrinsic to the work the author creates.  


reply via email to

[Prev in Thread] Current Thread [Next in Thread]