W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > January 1997

Re: Relationship Taxonomy Questions

From: Jon Bosak <bosak@atlantic-83.Eng.Sun.COM>
Date: Fri, 24 Jan 1997 11:44:13 -0800
Message-Id: <199701241944.LAA20581@boethius.eng.sun.com>
To: w3c-sgml-wg@www10.w3.org
CC: bosak@atlantic-83.Eng.Sun.COM
[David Durand:]

| At 2:34 PM 1/23/97, Jon Bosak wrote:
| >1. Specify the behavior without telling you what I mean, or
| My only problem with this is that this is something new, just for linking.
| IN XML I do _not_ have the ability to say how something is displayed, but
| not what it means.

I had in mind something like the rainbow DTD.  Or XML with embedded
CSS style attributes dropped into some future version of Internet
Explorer.  (No, that's not what we intend, but that's the behavior you
will likely get if you try it.  If anyone wants to start an
impassioned discussion on this point, please do it under some other
subject heading so that I don't have to read it.)  Or bitmaps wrapped
in <GRAPHIC> tags.  Or PostScript files in CDATA sections.

If we make DSSSL the default stylesheet language, then using nothing
but GIs like 001, 002, and 003 and pointing to a stylesheet
effectively accomplishes this; I'm telling you what to do, but not
what I mean by it.  Note that this does not guarantee that I will get
what I want; see below.  But it's close enough for my purposes.

| >2. Say what I mean and let you figure out what to do with it, or
| This is easy, and is supported for everything but is likely to be unpopular.

Yes, except among people designing content for the widest possible
range of display devices, which so far hasn't occurred to most HTML
authors.  Let them see how their hardcoded presentations are rendered
on the LCD of their local gas pump and they might start to think about
this differently.  ("Supported for everything" implies a shared
vocabulary of some kind; I'm assuming that you have something like
HTML in mind.  Without *some* common understandings you can never
figure out what to do, because then you don't know what I mean at all.
The shared vocabulary may be no more than identifying which elements
are supposed to be titles, but I think that there has to be something.
I can imagine heuristics so clever that they could tell by an
examination of the content itself what things are supposed to mean,
but those heuristics would be a good deal smarter than I am.)

| >3. Specify the behavior and tell you what I mean, but require that you
| >follow the specified behavior regardless of what you think about it,
| Again, how does one propose to enforce this. We don't offer the option in
| markup for text display, because we can't. Why do we _need_ to offer it for
| linking.

I didn't say that we needed it.  See below.

| I agree that it would be nice, for some applications but it's a requirement
| that is made almost unenforceable by generic markup. It's even
| unenforceable in Acrobat, if you're willing to pick through the file and
| try to mangle that. Though acrobat has such a small level of abstraction,
| that it is _very unlikely._

What I said was

   As the content provider, I want all of these options.

I didn't say that I expected to *get* all of these options.  I agree
strongly with Terry that I want to be able to tell the application to
follow the specified behavior, but I also agree strongly with you that
I'm not likely to get any guarantee that the application will in fact
do so.

My hunch is that Terry's intellectual property requirements will be
met through enforcement mechanisms outside of the publishing process,
just as they always have been.

| >Note that what I have just said applies to both the linking and the
| >non-linking parts of the document and for the same reasons.
| If this is true, then I have radically mis-understood what we can (and
| intend to) offer in the non-linking display of documents. How could we
| enforce a display discipline on top of a generic XML parser? I can see how
| we can enable publishers to write XML-using software that dopes such
| things, but it seems otherwise opposed to the whole purpose of using XML.

Sorry, I was being sloppy.  What I meant was that the same principles
that apply to the separation of meaning and behavior in the nonlinking
parts apply to the linking parts equally well.

And I was saying that I knew what I wanted, not that I knew what I
could get.  Figuring out what we can get seems to be a basic reason
for having this discussion.  I don't believe we can get everything
that we want, but I would like to start with what we want and then see
what it is we can't get and why it is we can't get it.

Received on Friday, 24 January 1997 14:45:07 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:25:07 UTC