Mozilla Looking to Forge Alliances (MozillaZine) [LWN.net]

Interesting thread

Posted Apr 8, 2004 18:38 UTC (Thu) by aseigo (guest, #18394) [Link]

> it is a file manager with ftp/webdav integration.

like Konqueror? =) yes, i know it isn't XUL based, but at least it exists
today.

Interesting thread

Posted Apr 8, 2004 18:57 UTC (Thu) by libra (guest, #2515) [Link]

Does it run on mswindows too?... you get my point...

Interesting thread

Posted Apr 8, 2004 19:35 UTC (Thu) by rjamestaylor (guest, #339) [Link]

Konqueror runs well on my XP Pro Dell 5150. Cygwin + KDE3 (rootless, too). Only has a problem with dual-headed monitors, but otherwise is impressively "cross-platform," as rootless X on Mac OS X on my Mac G4 Titanium was in 2002, anyway...

KDE on mswindows is pointless

Posted Apr 8, 2004 20:27 UTC (Thu) by libra (guest, #2515) [Link]

It is pointless, I will not install cygwin + KDE on mswindows only for all great apps of KDE (I do not even consider using KDE on GNU/Linux anyway).

Those kind of arguments only prove my point further that better arch is needed, to be both easily cross-platform and native on some (ex: GNOME on GNU/Linux if project goes well).

I understand that people like their OS, their window manager, their apps, but it does not solve the real problems of inner architecture. I think that some ideas presented in the thread are going in the right direction for architecture even if certainly not perfect.

It is at least better than pretending that KDE is everything while it looks as closed as mswindows from my point of view (well, I'm a bit harsh since I know it interacts well with Gnome in fact).

What we need is good underlying layers that allow better cross-platform, independent yet consistent programming. The project of integrating XUL with Gnome is a step in that direction, that's why it may be a corner stone for the future, and why I would support it.

Supporting KDE on cygwin on mswindows on the other hand is just pointless and counter productive to the final goal of better free architecture. Note that Gnome or Nextstep or whatever on cygwin on mswindows would be as pointless as well.

Please do not mix global and deep problems with immediate features and quick hacks.

Thanks.

KDE on mswindows is pointless

Posted Apr 9, 2004 7:56 UTC (Fri) by aseigo (guest, #18394) [Link]

> It is at least better than pretending that KDE
> is everything while it looks as closed as mswindows

KDE is made up of 100% Free Software and works with a LOT of other
projects on issues of sharing technologies and interoperability. KHTML ,as
just ONE example, has found its way onto a number of platforms, and
Freedesktop.org has tremendous support and participation from the KDE
project.

i don't agree with much of the rest of your posting, but while one might
chalk that up to a matter of opinion and vision, the quoted statement
above is just plain disingenuous and a lie.

you speak of furthering Free Software, and then stab it in the back with
your words. way to go.

KDE on mswindows is pointless

Posted Apr 9, 2004 11:41 UTC (Fri) by libra (guest, #2515) [Link]

The point is that KDE has made its own little bubble of things with its navigator, its office suite its graphic tools etc... In that way it is not better than mswindows, that is it works fine, but only in its own world, and I'm not interested in leaving a monopoly for an other one. But I also know that things are evolving, so while KDE still looks a bit bad for me, I know it is no longer as bad as it was from that point of view.

I'm quite ready to consider using KDE apps by the way, as long as they work as well on Gnome (without being encumbered by too much specific KDE library) or Windows and they avoid to all start with a K :) (even Microsoft apps would be more tolerable if they just avoided to have all their names starting with Microsoft).

What interests me is to have same tools on any platform I use, by platform I mean Operating Systems, but also Window Managers. Integrating XUL in Gnome, while keeping the ability to run XUL apps on mswindows or Mac is therefore a step in the right direction for me.

By the way I'm sorry that thread took that turn.

KDE on mswindows is pointless

Posted Apr 10, 2004 1:00 UTC (Sat) by aseigo (guest, #18394) [Link]

well, we (the computer using public) need applications and they need to be
written with SOMEthing, and KDE happens to be there as a free, very
portable and comfortable to use set of technologies. and that's why/where
all the KDE software came/comes from. it isn't some master plan to
implement everything as a KDE app (cue maniacal laughter ;) but rather
those who wanted a program that did X (for many instances of X) scratched
that itch using KDE technology.

btw, as a XUL proponent you will likely be happy to know that there is a
project called kaxul in KDE's CVS that takes a XUL interface and renders
it using KDE widgets. it supports quite a bit of XUL already, though there
is still more to go.

Interesting thread

Posted Apr 9, 2004 7:49 UTC (Fri) by aseigo (guest, #18394) [Link]

no, i don't get your point because i don't really care if it runs natively
on MS Windows or not. it runs on every OS i care to use, and having it on
Windows, which already HAS a decent file manager, is not even remotely
relevant to Free / Open Source software.

you said you wanted a file manager with FTP / webdav support. well, we
have that right now on UNIX/Linux, as does Windows and MacOS X. on
UNIX/Linux it happens to be in the form of Konqueror (among others, i'm
sure). how does having a file manager that runs on Windows and Linux (to
simplify it to two OSes) help ANYthing?

right. it doesn't.

Interesting thread

Posted Apr 9, 2004 11:53 UTC (Fri) by libra (guest, #2515) [Link]

In fact it is cornerstone to users migration from one platform to another. The other point is that I don't consider explorer on mswindows to be descent, it handles webdav badly (even if it certainly one of the best for that matter), and a lot of other things too (even when you spend time configuring it to behave better).

The fact that for you the current situation is satisfying doesn't mean it can not be improved or is satisfying for everyone else. But maybe I'm wrong, and it would in fact be quite relieving, because even with the battle for better infrastructure on the desktop is lost there would be less bad consequence if I'm wrong. But if I'm right, losing that battle can damage our hopes of freedom and innovation greatly.

Writing applications with XUL - not easy!

Posted Apr 15, 2004 12:13 UTC (Thu) by rwmj (guest, #5474) [Link]

I'd dearly like to use XUL for writing applications, but when I tried I came across a number of roadblocks which basically make it infeasible:

(1) It's undocumented. The book [Creating Applications with Mozilla], as you described, is confusing and dull. Writing anything more complex than a trivial window is entering unknown territory where the only documentation is Mozilla's enormous codebase. On top of this the terminology used for all the different "technologies" is confusing, so there's one way to do static pages, and another technology you need to use to display data pulled from a dynamic source. Ugh!

(2) You need to install an application as root and go through hoops to register it before you can run it, and restart Mozilla. This makes the edit / compile / debug cycle unusually stressful and lengthy. "Debug" is a relative word here: the debugging tool seems to consist of the Javascript console, which is waay inadequate for serious work.

(3) You need to write applications in Javascript. Javascript isn't really a suitable language for writing applications. There were (are) rumours of Perl bindings. There ought to be bindings across multiple languages.

(4) Rate of change: the environment is different for each different version of Mozilla, and it changes at a rapid rate.

(5) You need to distribute Mozilla with your application (and the right version of Mozilla at that, because XUL keeps changing). The Mozilla distribution is huge, and I'm not actually sure if you can have parallel versions installed together.

So this was about a year ago when I tried to write a Wiki-like editor in Mozilla. I gave up - easier to just do it in Gtk directly.

Rich.

Writing applications with XUL - not easy!

Posted Apr 15, 2004 13:50 UTC (Thu) by gerv (subscriber, #3376) [Link]

The situation may have been partly as you describe a year ago, but:

1) "Rapid Application Development with Mozilla" rocks.

2) The edit/compile/debug cycle is actually really short - particularly as there's no "compile" step if you are just using JS/CSS/XUL/XBL. With the caches turned off, it's as quick as editing HTML and pressing Ctrl-R.

3) JavaScript is a highly suitable language for what the Mozilla Framework uses it for - binding front-ends to back-ends. This is not your web-page scrolling-text document.forms JavaScript; JavaScript is an excellent OO, weakly-typed language with some handy features (e.g. prototypes) you won't find in any other mainstream language.

4) The XUL and XBL syntaxes have been fixed since Mozilla 1.0. And you don't have to worry about change if you pick a stable version (1.4 or 1.7) and develop on that.

5) The platform is currently designed to be shipped with the app. See ActiveState Komodo for an example. We do have plans for a Runtime Environment usable by multiple apps, but that's a way off yet.

No platform is a cakewalk; all have a learning curve. But I think with the publication of RAD With Mozilla, we are in a position to compete.

Gerv

Writing applications with XUL - not easy!

Posted Apr 15, 2004 13:57 UTC (Thu) by rwmj (guest, #5474) [Link]

Thanks for clarifying these points. As I said I'd dearly like to be able to write applications in XUL. It looks like I need to get the new book: RAD with Mozilla.

Rich.

Writing applications with XUL - not easy!

Posted Apr 15, 2004 14:05 UTC (Thu) by rwmj (guest, #5474) [Link]

Before you go and spring 20 quid for this book, it's available for free download here:

http://www.informit.com/content/downloads/perens/0131423436_pdf.zip

(I would give you the actual page which links to this content, but ironically that page doesn't render properly in Mozilla :-)

Writing applications with XUL - not easy!

Posted Apr 21, 2004 17:30 UTC (Wed) by angdraug (subscriber, #7487) [Link]

3) JavaScript is a highly suitable language for what the Mozilla Framework uses it for - binding front-ends to back-ends. This is not your web-page scrolling-text document.forms JavaScript; JavaScript is an excellent OO, weakly-typed language with some handy features (e.g. prototypes) you won't find in any other mainstream language.

Having learned tens of programming languages, I have finally found the one that has everything I need and is fun to program. And no, it is not JavaScript. Need I continue?

Here is yet another angle: I have a fairly large project that is written in a language that is neither C/C++, nor JavaScript. To use XUL in new GUI for this project, I'd have to add advanced JavaScript to the list of skills of my developers.

Mozilla Looking to Forge Alliances (MozillaZine)

Posted Apr 15, 2004 13:42 UTC (Thu) by gerv (subscriber, #3376) [Link]

However the book, Creating Applications with Mozilla, is so unreadable and long-winded that I have stopped at 100 pages twice.

I own both this and "Rapid Application Development with Mozilla", and can tell you that the later outshines the former as the sun outshines the moon. RAD is longer and more detailed (I don't know if you'd call it "long-winded"), but contains everything you need to know to write serious apps using the Mozilla Framework.

Gerv