MozillaZine

Netscape Cannot Win

Monday November 6th, 2000

I have recently become convinced that Netscape is fighting a losing battle against third-party micro-managing in their attempt to produce a browser for market.

The first conflict was with the raving loonies of the WSP. The WSP whines and bitches about standards compliance. They create a petition to convince Netscape to switch to development of their "Gecko" rendering engine technology, which promises greater standards compliance. They take credit for the Mozilla project moving to this new technology.

Then, they decide that the WSP is much more concerned about dictating project deadlines than standards compliance, and demand that Netscape release a product before the end of the year 2000, and stop any more work on Netscape 4.x -- including security fixes, apparently. (And read below to see what their fearless leader had to say today).

Now, a guy named Dave Flanagan has determined that Netscape 6 is too non-compliant to be released, although any reasonable, objective assessment of the application would prove that it is much more compliant than any browser currently on the market (yet you don't hear Mr. Flanagan petitioning against the release of iCab or IE6 or Konqueror).

And guess what? Mr. Flanagan has a petition you can sign, telling Netscape to hold off release of Netscape 6! But you can also use his petition to log your dissent.

[Mr. O'Reilly, I'm saddened that you signed an assent to Mr. Flanagan's petition without making an assessment of Netscape 6's standards compliance.

And Mr. Zeldman, I was dismayed, but not surprised, to read your statement in which you backtrack on the WSP's call for product release by the end of the year, when you obviously have not done an honest assessment of Netscape's standards compliance. "Releasing a close-to-perfect standards-compliant browser would be fine. (No software is perfect.) But releasing a browser with seriously buggy, incomplete standards support will not serve Netscape, will not serve developers, and will not serve the cause of web standards."]

I have made it known to a few people that I will no longer be fielding Netscape news on my site. You may see news of product releases, but you will no longer see links to reviews, complaints, rants, or any other information having to do with the Netscape product.

This site is called MozillaZine for a reason. I started it to support the Mozilla Open Source project. Along the way, I tried to be a voice of support for Netscape as it navigated the murky waters of being a commercial contributor to an open project. But what I am confronted with is the fact that if I were to continue posting Netscape news I would be forced to continue posting links to this garbage, giving it more credibility than it deserves. Look what happened when news of Mr. Flanagan's petition hit Slashdot. The petition now overflows with rants against Netscape, and I bet not one of these ravers has done even a modest assessment of Netscape 6's standards compliance. Is that the kind of support you wanted Mr. Flanagan? If you can get lots of people to raise their voices, does the din drown out their ignorance?

In any case, this crap will no longer will find a home in MozillaZine. So, take a good long look. Gaze into Mr. Flanagan's eyes. Because he's the last guy with an axe to grind against Netscape that you will see in these pages. The last armchair-marketer to get a say on this site. You'll have to go elsewhere to get your fix. (However, I allow myself an exception to clobber the hell out of the WSP if they continue to act like pussies.)

Netscape is in the unenviable position of choosing between bug-fixes and product release - the Scylla and Charybdis of software development. They've set their course. Maybe they will stop and reassess. That's up to the Netscape managers and the PDT team. But the waters grow increasingly insipid, and they might just end up deciding that it's not worth the trouble and pack it all up and call it a day. Maybe they can put in the past this peanut-gallery micro-management Hell that they've fallen into.

I'll leave you with a very thoughtful post to the n.p.m.layout newsgroup by Dylan Schiemann.


#34 We should really use their input, but...

by Arondylos <malte@cornils.net>

Tuesday November 7th, 2000 7:11 PM

You are replying to this message

Hi, I am not a mozilla developers although I have submitted a few bugs here and there and use the product daily. Still, where have those people been hiding all along? A lot of them are web designers and even coders, and the mozilla code has been out in open source form for years now, there is a voting system in place in bugzilla so you can vote which bug is important for you. Why did you not *use* this functionality if your daily job involves handling non-standards-compliant browsers? If it is that important to your company (and you are oh so angry that mozilla does not support DOM level 5 subrevison 4711) why did you not fix it way earlier? The browser was really usable *for evaluation* since about M15, and that was way before the now-strict deadline and fixing policy. I would love to see the bugs mentioned in the article fixed too, but they are not as important as the RTM bugs, and even in the mostly-automated build process, a single incorrectly applied typo patch can hose the system (though not likely). What do you think the Netscape and external contributors do right now, they are not twiddling their thumbs.

Still, let us see what Flanagan proposes: #57634 is about the browser not being lax when encountering invalid HTML. Which forces web developers to use the standard. Is that standard *compliance* a bug to be fixed? Doubtful.

The problem with the others is mostly that good and reliable testcases to make sure the proposed fix does not break the really important stuff have not been made, or that the issues are not important enough - I say that because the web and its browsers have always been evolving rapidly, and you can build a mozilla-readable and standard-compliant site much more easily than before. DOM functions which no other browser supports yet either and similar ECMAScript functions should be fixed, sure. But if I recall correctly, there has been a vote on mozillazine what should be worked on before release. The winner was definitely stability and performance. That is what the people who *care* about mozilla wanted, and it was done.

And the open source *mozilla* project will incorporate all these little fixes, even if NS6 will not.

What do you expect from a, from any commercial software release? It is not possible to do otherwise in a world with deadlines and monetary pressure.

It is a .0 release. They cannot be perfect. Name any project larger than Hello world which shipped a release with no missing, but important functionality.

Still, a few points are valid. There should be a README in the source which points to the great architectural overviews on mozilla.org for the developers. The learning curve is steep (I would have to learn C well first) even for experienced developers, same goes for Linux kernel/Wine and quite possibly for the Windows kernel (this is a polite understatement, but I am not among the (un)lucky few who read that ;)), but this should be remedied if possible. This is being done for mozilla 1.0, which the developers should work on, not NS6.

And a few bugs mentioned are important, and the PDT *will* react and push them for inclusion (the fixes, not the bugs ;)) and hopefully put out a nice press release telling the world that they do listen. Just for marketing reasons, even if they should not fix these bugs. Unfortunately, this would make most of the people who commented on the Flanagan article that much more happy. Even if we know we were right :-)

Have fun

Yours Arondylos #8-)

(and thanks for a great, amazing product!!!)