Re: LANG as an attribute

[Chris Lilley:]

| > bearing in mind that the great majority of publishers consider h&j to
| > be a more important capability than bidirectional printing.
| I am sure they do.  The great majority of *readers* on the other hand
| consider reading the text at all to be more important than not reading
| it (safe in the knowledge they are being spared sub-optimal
| hyphenation).  

Your emphasis seems to be on a small subset of the total picture --
getting readable results in a Web environment.  DSSSL (and DSSSL-Lite)
are working in a different problem space.  A basic part of the idea is
to provide a single standardized formatting language that will work
for online AND traditional printed documents, so that people like me
who have to maintain several delivery systems at the same time don't
have to use a number of different languages to do it.

| What I am saying is, it may be more important to provide limited
| functionality than none at all.  And just because an expert says that
| something is impossible does not mean it cannot be done.

Please focus on what is actually being said.  I explicitly stated at
the beginning of my comments that bidi is not impossible.  In fact,
bidirectional rendering is an important feature of the DSSSL standard.
It isn't a question of impossibility but of implementation priorities.
DSSSL-Lite is, by definition, that subset of DSSSL that will be
implemented in the first generation of DSSSL tools.  Which features
get chosen will be dictated by the needs of the market.  The great
majority of publishers who need DSSSL to simultaneously deliver
documents across a variety of heterogeneous output media (including
print media) will consider hyphenation and justification, column sets,
tables, and so on to be far more important than the ability to quote
fragments of right-to-left languages inside left-to-right languages,
or vice versa.  It is that simple marketing fact that lies behind what
seems to be a general vendor consensus that bidi will not be in the
first subset of DSSSL to be implemented.

| You did say that DSSSL-Lite would not support mixed left-to-right and
| right-to-left text *at all*, did you not?  What will it do, then -
| give up?  Refuse to display the document?  Dump core?

It isn't DSSSL that's going to do anything -- it's the application.
If the application is an ambitious Web browser, it could go ahead and
implement the whole DSSSL spec or some appropriate subset that
transcends the basic set of features that we're calling DSSSL-Lite.
That's up to the manufacturer.  If Web browser vendors are like
everyone else, I suspect that they will find bidi falling rather low
on their list of marketing priorities.  But that's up to them and
their customers.


Jon Bosak, Novell Corporate Publishing Services            jb@novell.com
2180 Fortune Drive, San Jose, CA  95131                Fax: 408 577 5020
A sponsor of the Davenport Group      (ftp://ftp.ora.com/pub/davenport/)
   The Library is a sphere whose consummate center is any hexagon, and
   whose circumference is inaccessible.  -- Jorge Luis Borges