Re: DSSSL style editing (was: RE: Positioning...)

Jon Bosak wrote:
> 
> [Mike Wexler:]
> 
> | With HTML I can create a page in Emacs post and post it to my web
> | server.  Then John can download it with Netscape Gold, edit it, and
> | post it back to the web server. Then Steve can download it and edit it
> | with PageMill and post it back. Then Andrea can download it, modify it
> | with FrontPage and post it back.

I've done it, and regularly do it, with Emacs, HoTMetaL, AOLPress and
Amaya, no problem. (All on Unix, of course.) Amaya is still not the
editor I use most often, but is has some unique features, such as the
ability to transpose a table. (Gives some "interesting" effects with
certain pages on the Web :-) )

(Mike, when can I add FrameMaker to this list?)

> 
> I will believe in your ability to freely mix and match HTML+CSS
> documents when I see it.

It shouldn't be so hard to reach the level of interoperability that RTF
offers currently under Windows (which is not a lot, once you leave
Windows :-( )

When you create a page in MS Word, you expect somebody to be able to
read that page and edit it, not just the text, but also its style. Now
assume the interchange format is not RTF, but HTML (or XML) + CSS. Any
reason why that shouldn't work the same way?

> 
> | I expect that eventually, we will be able to do the same thing with
> | HTML+CSS or even XML+CSS. I don't think this will ever be possible
> | with XML+DSSSL.
> 
> Other people are less pessimistic than I am about the ultimate ability
> to exchange DSSSL stylesheets among different stylesheet editors;
> we'll see.

A style sheet should not be a black box that magically produces some
output. The style sheet itself is information, and should be reusable.
We would have defined a "CSS bytecode" if we believed otherwise.

> 
> | I think this ability to for users on different platforms with
> | different software to share not only viewing of documents, but editing
> | of documents is one of the key reasons for the success of the web. I
> | would bet a large percentage of the pages posted by individuals are
> | modified versions of somebody else's pages, often created with
> | different tools.
> 
> Perhaps so, but "a large percentage of the pages posted by
> individuals" is not "a large percentage of the pages posted".  I think
> that HTML and CSS do just fine for individuals.  That's not the
> problem space that I personally am interested in.  I'm interested in
> the problems relating to large-scale commercial publishing and data
> distribution efforts.  XML+DSSSL may or may not be useful to
> individuals; that's up to other people to decide.  But the level of
> functionality they offer is essential to solving the kind of
> large-scale problems that I'm interested in.

Maybe, but I doubt that the needs of commercial publishing are so
different from those of non-commercial institutions and of individuals.
The only difference is that the former may have the money to reach their
goals through a brute force approach.

> 
> I will not attempt to prove, but will for the moment merely assert,
> that ten years from now the total number of pages on the Web posted by
> commercial content providers will be much larger than the number of
> pages posted by individuals.

I think (and hope) that you are wrong. Look what happened to DTP:
nowadays every village and every street has its own newsletter. With the
Web, publishing will be even easier. OK, a text by <insert favourite
author> will be read by millions while the Harbour-Street-Herald will be
read by a few dozen, but there are millions of the latter and only one
of the former.

>  Whether I'm right about that or not,
> that's the problem space in which I'm working.  The kinds of
> publishing problems that I alluded to earlier arise from this area.
> CSS cannot effectively address these problems; DSSSL can.  That's why
> I'm interested in it.

But so can Perl, or Awk, or Java, or... why use CSS?

I'm a strong believer in the "right tool for the job." CSS is not a text
manipulation language, it is not a search engine, it is not a
statistical analysis tool, it is not an indexer, it is not a sorting
program, it is not... That's not weakness, that's strength. DSSSL can do
all these things, but I happen to believe that Perl, Awk and Java are
better at it than DSSSL.

But let's not get lost in religious wars. I want to tap Jon's knowledge
of large-volume e-publishing to help design the Web (that is the *whole*
of the Web, not just CSS). These discussions help, but there may be more
efficient ways.


Bert

Received on Wednesday, 5 February 1997 16:51:52 UTC