Wither TeX? (was Re:Last typewriter factory in the world shuts its doors)

[Evan Leibovitch]

On 12 May 2011 07:29, Stewart C. Russell <scruss-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org> wrote:

> On 11-05-12 00:59 , Evan Leibovitch wrote:
> >
> > Reflowable text means something quite different from that,
> something  that TeX is even worse at.
>


> Well, you can hardly blame Knuth for not adding a feature that didn't exist
> until 20 years after TeX was conceived.


The "feature" was a philosophy that the user had no right to suggest -- let
alone have the ability to alter -- the presentation. Given the rise of the
Internet and its associated shifting control to the consumer, Knuth's
feature is in increasing disrepute,

What has replaced TeX -- XML together with its sub-genres, MathML, HTML, etc
as well as CSS -- could have evolved using TeX syntax instead of what
eventually was used; that is

\documentclass{foo}
\begin{bar}
The quick brown fox doesn't jump over foxes anymore
\end{foo}

could have become the norm instead of

<!DOCTYPE foo>
<bar>
The quick brown fox doesn't jump over foxes anymore
</bar>

There was nothing wrong or lacking in the syntax style (except perhaps an
anglo-centric focus). It was the associated 'screw the user' baggage behind
TeX -- and the features that grew out of that -- that have prevented TeX
from realizing its potential.



> TeX is a killer H&J engine; it optimizes line and page breaks about as well
> as it is possible to do without skilled manual intervention.


Of course, in the digital realm the very concept of the page is
backwards-compatibility hack that may itself be replaced by other navigation
tools, just as the traditional index in books has already been made
redundant by the user's ability to do full-text searches.

 At least in the word processors I know, page breaking can be configured to
be done fairly intelligently (browsers need no concept of physical page
limits, so page breaks are manually added anyway). Same with justification.
Though I would agree that TeX still likely does hyphenation better, even the
demand for that drops as full-line justification appears to go out of
style.

I'm guessing that the e-readers don't do proper visual space balancing in
> paragraph flows
>

Ah, there's the rub... what's "proper"?

In an ereader, the end-user gets to choose what spacing is most comfortable
for them, one size need not fit all. In TeX, the content provider makes a
guess on what's right and then enforces it for everyone. TeX could certainly
enforce the rules better than any other tool of its kind, but that
enforcement isn't needed the way it once was.

As for kerning, that's now done more in the font, which is IMO where it
belongs because it then becomes display-vehicle independent.


> (certainly, no web browser I've seen does more than the most trivial line
> breaking), and very probably don't know how to deal with widows and orphans
> appropriately.
>


This kind of thing is easy to verify -- the HTML/CSS spec certainly allows
for orphans <http://xhtml.com/en/css/reference/orphans/> and
widows<http://xhtml.com/en/css/reference/widows/>.
So if a content provider cares to want them, the browser (or ereader) will
obey.


TeX isn't a display typography engine. TeX isn't for reflowable text. Don't
> try to make it be something it isn't.
>


That's my point.It isn't -- and it can't be -- what is needed in textual
content delivery going forward.

- Evan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gtalug.org/pipermail/legacy/attachments/20110512/318159ac/attachment.html>