W3C home > Mailing lists > Public > www-dom@w3.org > July to September 2000

Re: Usage of null

From: Joseph Kesselman/Watson/IBM <keshlam@us.ibm.com>
Date: Mon, 18 Sep 2000 13:25:38 -0400
To: www-dom@w3c.org
Message-ID: <OF92B98D89.22C5BED3-ON8525695E.005EE817@pok.ibm.com>

I believe the only places where the DOM uses null values are when an
attribute is expected to be an object, so it should always be possible for
each of the DOM's language bindings to provide a suitable "null object".
And I believe we do provide an indication of how nulls and empty strings
are interpreted; in general, they are NOT considered interchangable. If you
can cite specific places where those practices have been violated, we
should definitely take another look at them. Otherwise, I think this is
implementable as it stands.

The "abstract" IDL shown in the body of the DOM spec is not guaranteed to
be fully compliant OMG IDL. We know there are some keyword clashes, for
example. The OMG seems to be aware of this and has not complained; instead,
their response has been to try to develop a separate IDL binding for the
DOM, just as there are bindings for other languages. I believe that binding
does attempt to take advantage of valuetypes.

______________________________________
Joe Kesselman  / IBM Research
Received on Monday, 18 September 2000 13:25:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 22 June 2012 06:13:47 GMT