O'Reilly Network: Netscape Navigator 6.0 to Fail Standards Compliance: An Update

Published on O'Reilly (http://oreilly.com/)
See this if you're having trouble printing code examples

Netscape Navigator 6.0 to Fail Standards Compliance: An Update

by David Flanagan
11/08/2000

I've been shocked by the number of responses to my article and petition to the Netscape product development team. I've also been very pleased by the quality and thoughtfulness of most of the responses. Recently, however, many of the reponses sound more like flames or rants. Before things get out of hand, I'd like to take this opportunity to summarize reader response so far, and to clarify and restate my position.

Repeated themes that emerge from the responses include:

The browser wars are over. There is no longer a race to release new versions, so Netscape should not worry about slowing down and getting this release right.
We've already waited this long for a new browser from Netscape and we're willing to continue waiting a bit longer.
Netscape is on the right track; they've got a good product, but please make 6 a great product!
This is Netscape's last chance to get it right.

Many of the responses expressed a lot of anger, which surprised me. I've obviously touched a raw nerve. I wrote my article from the theoretical perspective of someone who only does enough Web development to write books about it. Readers who live and breathe Web development wrote from their in-the-trenches perspective, and many of them are angry, frustrated, and dismayed. This is obviously not a representative sample since most readers who disagreed with me did not choose to respond, but nevertheless, it would behoove the Netscape marketing and PR folks to try to understand where this reaction is coming from! It is worth noting, however, that a lot of the angry comments seem to concern Navigator 4.x, and so are not really relevant to my point about Navigator 6.0, which is based on entirely new code.

It appears that some readers have taken my article to mean that Navigator 6.0 will not comply with standards at all, or that Navigator 6.0 will be as noncompliant as Navigator 4.x. That is not the case. Netscape and Mozilla engineers deserve tremendous credit for creating a browser that has very good standards compliance. In fact, according to many who have studied it more throughly than I, Mozilla and Netscape 6 are more standards compliant than the competition. See, for example, Netscape Standards Challenge. I regret that I did not make this more explicit and give more credit to the Mozilla and Netscape engineers in my original article.

Allow me to restate, then, what I consider to be my main points:

First, although Navigator 6.0 has good standards support overall, there are several egregious known bugs, which can be easily fixed if the Netscape PDT would allow the schedule to slip. Consider the omission of Date.toDateString() and Date.toTimeString() ECMAScript methods. Neither the folks who implemented JavaScript 1.5, nor the folks who wrote the test cases, bothered to read the standard carefully enough to discover that these methods needed to be implemented! That kind of sloppiness does not say "ready for RTM" to me. Also consider the <DL> tag that cannot be nested within a <DD> tag. These tags have been part of HTML since the beginning. To break them now and think it is okay to simply mention the fact in the release notes is just unacceptable.
Netscape has not yet released a credible beta version of Navigator 6.0. There have been three "preview releases," but in my lexicon "preview release" is a code word for alpha release. Maybe I'm just out of touch with this, but my personal experience with PR3 on Linux bears it out; PR3 displayed an "Activation" screen that locked up, continuously spewed debugging statements to my console, and then crashed after a short period of use. When it crashed, the talkback feature was unable to successfully talk back to Netscape and report the crash. So I gave up on PR3 and decided to wait for the beta release, which I imagined must surely be forthcoming. If my experience is at all typical, then PR3 may have gotten much less external "beta testing" than Netscape realizes.

These are the reasons why I believe that Netscape should rebrand the upcoming release as a beta. Give us a good stable-enough-to-test beta release now, distribute it widely, and let the Web developer community really bang on it while you allow your engineers to get down to work on the bugs that they know need to be fixed. I, and everyone who has responded to my petition, want Netscape and Navigator to succeed, but we worry that it will not if left in its current state.

For an alternative viewpoint of these issues from inside Netscape, see Eric Krock's essay on the trade-offs of waiting for perfection versus getting a product to market.

There is also a discussion of my article on slashdot.org, although much of that discussion appears to have veered off topic to a debate about Netscape versus Internet Explorer and open source versus Microsoft.

--David Flanagan

Read David Flanagan's original article about the compliance problems of Netscape 6, titled Netscape Navigator 6.0 to Fail Standards Compliance.

David Flanagan is the author of several best-selling O'Reilly books, including JavaScript: The Definitive Guide, Java in a Nutshell, and Java Examples in a Nutshell.