IntlHarvester's Journal: "We had no choice but to implement XUL" Is The Big Lie

Why AC trolling can sometimes be valuable ...

----------------------------------------------------
Re:If...(Score:0, Troll)
by Anonymous Coward on 07:41 AM July 16th, 2003 (#6452427)

Netscape a.k.a Mozilla got flushed because Mozilla sucks. Check the Register for what really killed Mozilla (http://www.theregister.co.uk/content/4/31765.html ) [theregister.co.uk]

Long live Opera!

----------------------------------------------------

Re:If...(Score:1, Interesting)
by Anonymous Coward on 07:50 AM July 16th, 2003 (#6452528)

I actually agree with The Registers analysis. When Netscape and AOL needed a great browser to battle with Microsoft, the Mozilla developers gave them an entire "application framework" that they didn't need, and a bug tracking system that could simply have been done with a commercial offering or even a few PHP scripts & a MySQL database. Mozilla developers were trying to be a "platform" instead of a damn browser; if they had worked on the portable Gecko completely and forgotten (Or at the very least, pushed right back) things like XUL and skined interfaces, they could have written a handful of application shells for their supported platforms and dropped in an excelent browser engine. They could have done it, from scratch, in two years. Instead we got Mozilla, the framework no one wants. Gecko is great and if it had been earlier, could have held Microsoft in check.

JZW was right, in a way. Starting from scratch is not always the best thing to do. With Gecko, they were probably right to do it. Did they need to throw away their existing Netscape applications & write XUL though? Did they really need to re-write Netscape Mail? Probably not.

----------------------------------------------------

Re:If...(Score:5, Insightful)
by Gerv (15179) on 07:58 AM July 16th, 2003 (#6452624) (http://www.gerv.net/)

if they had worked on the portable Gecko completely and forgotten (Or at the very least, pushed right back) things like XUL and skined interfaces, they could have written a handful of application shells for their supported platforms and dropped in an excelent browser engine.

So, Mr. Know-It-All Anonymous Coward, pontificating from on high, here's a pop quiz. If you have to implement an entire widget set in your browser to have any hope of supporting styleable form controls etc. (as outlined in CSS2 and above), is it better to:

a) Write one user interface for all platforms using those same controls, and use that UI as another testbed for them
b) Write five or more separate user interfaces, and have to keep them all up to date and in sync?

Without XUL, there would have been no Netscape help in doing Mozilla for Linux, Mac, BSD etc. because there would have been no incentive to chase such a small part of the browser market.

Gerv
([e-mail])

----------------------------------------------------

ask a stupid quesiton...(Score:5, Insightful)
by Doktor Memory (237313) on 11:56 AM July 16th, 2003 (#6454939)

So, Mr. Know-It-All Anonymous Coward, pontificating from on high, here's a pop quiz. If you have to implement an entire widget set in your browser to have any hope of supporting styleable form controls etc. (as outlined in CSS2 and above), is it better to:

a) Write one user interface for all platforms using those same controls, and use that UI as another testbed for them
b) Write five or more separate user interfaces, and have to keep them all up to date and in sync?

Guess what, hotshot? The answer to that question is: Whichever one will not take 4+ years to ship in a working form while the world's largest and most predatory corporation is working overtime to dig your grave.

Please notice that despite the nonstop handwaving from the Mozilla team about how maintaining seperate native interfaces for the assorted Gecko frontends was supposed to be some sort of impossible herculean task that no reasonable person could be expected to tackle, in the time that it took to produce ONE semi-functional version of Mozilla, Opera Software, a company with not even a tenth of AOLNSCP's resources, produced multiple versions of a fully functional web browser, for all of Mozilla's major target platforms. Not only did they produce, maintain and upgrade native Windows, MacOS and Linux versions of Opera, but they increased their market share, and made money doing it.

"We had no choice but to implement XUL/XPFE" is the Big Lie of the entire Netscape saga. The fact that mozilla team members are still stating it with cultish earnestness suggests not that you all came to a reasoned engineering decision, but that your project management was not merely incompetant, but downright pathological. If 1% market share and the firing of your entire development team isn't enough to convince you that somewhere, somehow, you made the wrong decision, you are simply delusional.

Hopefully, some of the core Mozilla developers and managers will use some of their newly acquired free time to read Fred Brooks' "The Mythical Man-Month." When Brooks talks about the Second-System Effect, he's talking about you.

----------------------------------------------------

Re:Opera now has an XPFE though!(Score:4, Insightful)
by Doktor Memory (237313) on 03:39 PM July 16th, 2003 (#6456871)

For the record, I have nothing against the concept of cross-platform development toolkits. They can be great, time-saving things.

But. Priorities. Opera developed a functional product that could be used by the vast majority of their paying customers first. Then they prototyped and shipped versions for secondary platforms. After they started seeing revenue (or the potential for revenue; I'm not privy to their books, merely aware that they're apparently still in business, unlike the Mozilla team), they then wrote the minimum amount of glue to allow them to ship their releases in lockstep. And they did it in what...a quarter of the time it took to build a functional XPFE browser? An eighth?

Second point: XUL was more than just a cross-platform widget set. If that had been all that it was, Moz 1.0 would have shipped in 1999, maybe even 1998. People write cross-platform toolsets all the damn time, and it rarely takes half a decade to do. No, XUL/XUI/XPFE were the logical result of Netscape drinking its own "it's not a web browser, it's an application platform! [suck.com]" kool-aid. It's an API, it's an application framework, it's a development toolkit, it's an XML parser, it's a widget set, it'll walk your dog and it gets your whites whiter!

Just search for comments from users with mozilla.org and netscape.com addresses on slashdot for the past few years: Mozilla wasn't just going to be a better web browser, it was going to be the foundation for an entire industry of "mozilla-based web applications" that someone, somewhere, was sure to write.

See, as far as I can tell, it's the not-so-secret desire of just about every developer who ever lived to write The One Universal Cross-Platform Middleware Library That Everyone Will Use Forever. Therefore, except in the exceedingly rare instances where doing that is the actual stated and understood project plan from the CEO on down (ie: win32, java, .net, openstep), the job of every project manager in the world is to stand behind that developer's back with a rattan cane, and smack them across the shoulders everytime they start to try it. Netscape's management completely failed in this critical task, and Microsoft's near-total control of a market that 5 years ago they were an also-ran in is the entirely predictable result.

----------------------------------------------------

Re:If...(Score:3, Insightful)
by hixie (116369) on 01:27 AM July 17th, 2003 (#6459404) (http://ln.hixie.ch/)

Ok that's it. Gerv, you need to stop talking utter garbage as if you were some kind of authority on the subject. You aren't.

Let's get some facts straight. First of all, CSS (any version) does not require that you style form controls. That is a myth, perpetuated by people like me, who used to want to see that level of control available to authors (As you can tell from the recently released CSS3 UI draft, the CSS working group is in fact moving away from stylable controls altogether).

Secondly, it is quite possible to develop multiple products for different platforms, and in fact, for some platforms it is the best way. In particular, the Mac. The Mac's UI is SO different from other platforms in key, if subtle, ways (menu bar placement, order of menu bar items, the fact that you can have an application running with no windows, etc) that it is significantly EASIER to write an application specifically for that platform rather than try to continually fix XUL to work on the Mac.

Sure, some platforms (Win32, Gnome) are similar enough that you can use one widget set and a few #ifdefs to support both platforms. But that is by no means a requirement.

So please, get some perspective, get your facts right, and stop posting with "@mozilla.org" in your sig as if it meant anything more than "I used to intern at Netscape and they never took away my mail account". The sad fact is you're only on [e-mail] because the rest of staff are too chicken to ask you to leave.

-- Ian Hickson
(Editor of Mozilla's XBL spec, Mozilla's XUL spec, the W3C's CSS2.1 spec, three W3C CSS3 modules; Invited Expert to the W3C; QA contact for a dozen or more Bugzilla components; Mozilla contributor for 4+ years; Intern at Netscape for 4 times longer than Gerv; and currently employed by Opera software. But no fancy e-mail address.)

--------------------------------------

Re:If...(Score:1, Interesting)
by Anonymous Coward on 09:11 AM July 16th, 2003 (#6453265)

b) Write five or more separate user interfaces, and have to keep them all up to date and in sync?

I choose B.

Why? Because it took you guys 5+ years to implement basic features like a customizable toolbar. It took 3 years before you got all 9000 platform-specific key commands and behavors sorted out. It even took a couple years before the fuckin Mac Menu Bar was in the right place. All of that crap is done - a solved problem. No need to do it again.

Meanwhile, I hire one MFC guy and he's hacked together Netscape's UI in about a week. Add a GTK guy and a Mac guy, and I'm set. OS/2, BeOS, etc can fuck themselves.

Maybe you did need your own widgets for web form controls, but there's a big difference between that and an entire application framework that supports a big app like Netscape. The amount of time it took proves it.