Re: Fwd: In-Band Information Considered Harmful

[Art Pollard <pollarda@xxxxxxxxxx>]

On Mon, 26 Oct 1998, Rich Pasco wrote:

> Why harmful?  Looking to see your article.

Why harmful?  Simple.  Styles, markups change.  Before HTML/XML people
used a myriad of different markups so people had to write a million of
different interpreters.  The same thing holds though for XML.  Do you
really think that XML will be used in 500 years?  If so, then use it.  If
not then perhaps we ought to rethink our use of XML.

> If not embedded then where?

On the side like a salad. :-)  Separately from teh text though.

> Until I read your article, I find myself inclined to disagree.

You should agree.  It is much better. 

> I would much rather have an ASCII file with embedded markup than a file
> with binary markup (e.g. MS-Word, WordPerfect, etc.).  At least I know
> I'll be able to read it 25 years from now.

Howabout rather than having an ASCII file with embeded markup, wouldn't it
be better to have an ASCII file with no markup at all?  You will be better
able to read in in 25 years than either embeded markup or binary markup as
you put it.  Simply keep all your markup separately then put it back in
the ASCII file when / if you need it at run time.

> What about an ASCII file with a separate file for markup, giving byte
> offsets into the ASCII file?  This would then require a special editor
> to keep the pointers straight, whereas HTML, etc., can be edited with
> any vanilla text editor.

Yep.  Now you have the idea.  This is what Ted has been advocating for a
long time.  (Though he doesn't say "separate file" just separate.)  Sure,
you need a special editor but, it is better to write a new editor for a
TON more power.  Then when a new "style" or markup comes which is better
than whatever we are using now (surely there is something better than XML)
it would simply be a matter of creating a new style file than having to
redo the text itself.

-Art

Art Pollard <PollardA@xxxxxxxxxx>
Moderator for Comp.Theory.Info-Retrieval
List Maintainer for the Hyper-Theory (Hypertext Theory) mailing list.