W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > October to December 2000

Netscape 6 problems

From: Charles F. Munat <chas@munat.com>
Date: Thu, 30 Nov 2000 16:21:25 -0800
To: "WAI Interest Group \(E-mail\)" <w3c-wai-ig@w3.org>
Message-ID: <002a01c05b2c$a47259a0$0100a8c0@aries>
Hello,

I've been examining the new Netscape 6 browser and looking to see how well
it complies with standards. I admit that so far I've been surprised: they
really cleaned it up a lot since PR2. I was nearly apoplectic after the
first preview release showed none of NS4's myriad bugs had been fixed.
Yesterday I was ready to proclaim NS6 a victory for standards compliance,
but then today I found this:

DOM (Core) Level 1 provides a DOMImplementation interface:

"The DOMImplementation interface provides a number of methods for performing
operations that are independent of any particular instance of the document
object model."

I read this to mean that these methods are above any given document. In DOM
1, there is only one method: hasFeature. Here is the IDL definition:

interface DOMImplementation {
  boolean                   hasFeature(in DOMString feature,
                                       in DOMString version);
};


As I read this, I should be able to use

DOMImplementation.hasFeature("XML", "1.0");

to check to see if the XML core features of DOM 1 have been implemented. I
tried this in Netscape 6 (which claims to implement these features) and got
this response: "Error: DOMImplementation is not defined."

Hmmm. So I went searching through the numerous pages at netscape.com
bragging about standards compliance. There I found this method:

document.implementation.hasFeature();

I tried it, and sure enough it told me that XML Core of DOM 1 was
implemented.

Note also that the DOM uses Document, not document. I tried using the
capitalized version and NS6 didn't recognize it. So I assume that ALL of
NS6's DOM implementation is wrong.

So, am I reading this wrong, or have they screwed it up again? What is the
point of "standards compliance" if you rewrite the standard your own way?
How difficult would it have been to simply do it right? But most
importantly, why? why? WHY??? I cannot believe that this is sheer ignorance:
the DOM recommendation seems clear enough to me. So is it willful? Did
someone just decide that it wasn't important to follow the DOM in this
instance? It seems to sabotage the entire effort. Why bother if developers
are just going to have to make duplicate versions of all their code to allow
for the spelling differences between versions? Or am I reading it wrong and
Mozilla got it right?

If anyone on this list can enlighten me, I'd appreciate it. And if Mozilla
HAS screwed it up again, then I hope that everyone in the developer
community who is concerned with standards compliance will let them know
LOUDLY AND CLEARLY that this sort of thing doesn't cut it. If they want to
brag about how "standards compliant" their browser is, then make it
STANDARDS COMPLIANT. Not "almost".

(BTW, I note that their web site claims that NS6 is the "most" standards
compliant browser. That's marketing talk for "we didn't really make the
effort to do it right, but we want the credit anyway.")

Sincerely,
Charles F. Munat
Seattle, Washington
Received on Thursday, 30 November 2000 19:17:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:13:50 GMT