[Top][All Lists]

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

Re: Bibliography & ToC placement

From: Valeriy E. Ushakov
Subject: Re: Bibliography & ToC placement
Date: Wed, 22 Apr 1998 14:51:19 +0400

On Tue, Apr 21, 1998 at 12:45:44PM +0200, Thomas Steffen wrote:

> >>>>> "Paul" == Paul Prescod <address@hidden> writes:

Well, this should be "Dan" in fact, as you quote a material that Paul
just quoted from


by Dan Connolly.  Note the quotation marks and the URL in his original
message.  Also, this quote was cited in particular conext:

| address@hidden wrote:
| > Sure would be nice to have a WYSIWYG editor for someone like my wife that
| > stored in lout format.
| What you want is a word processor that *works*. Word's file format has
| almost nothing to do with its problems nad I think that lout format
| would actually be a step backwards. Working with a programmable format
| in a WYSIWYG editor would be more prone to bugs and crashing than
| working with the (mostly) non-programmable formats that Word accepts.
| http://www.w3.org/MarkUp/html-spec/html-essay.html
| "My experience with document interchange led me to classify [.....]

> declarative or algorithmic is very much a question of viewpoint, and
> not of the format itself.

You can interpret any sequence of bits whatever you like.  An SGML
document might be a valid VAX binary.  "Format" is not only some
pattern on how bits are stored but a "viewpoint" on how to interpret
that pattern of bits.

>     Paul> The reason that this distinction is essential with respect
>     Paul> to document interchange is that extracting information from
>     Paul> documents in "programmable" document formats is equivalent
>     Paul> to the halting problem. That is, it is arbitrarily difficult
>     Paul> and cannot be automated in a general fashion.
> which is wrong. converting a algorithmic format to a declarative is
> basically the same as processing it. so it is possible, but not
> necessarily easy. which is exactly because algorithmic formats can solve
> arbitrarily complex problems.

How can you conclude that that paragraph of text set in bold 14pt
sans-serif font is a section heading?  Programmable formats are
invariably presentation markup, though Lout and LaTeX demonstrate that
one can *mimics* a declarative markup with a programmable presentation
markup system (read: text formatter) - and it seems that this often
misleads people.

>     Paul> For example, I conjecture that it is impossible to write a
>     Paul> program that will extract the third word from a TeX
>     Paul> document. 
> TeX can.

*Only* TeX can.  Anything else will be a reimplementation of TeX and
thus, effectively, the TeX.

Note the particular topic that was discussed: WYSIWYG editorthat
stores in Lout (programmable) format.

SY, Uwe
address@hidden                         |       Zu Grunde kommen
http://www.ptc.spbu.ru/~uwe/            |       Ist zu Grunde gehen

reply via email to

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