Slashdot Mirror
Future Directions Proposed For Mozilla
Ars-Fartsica writes "MozillaZine is now featuring a set of slides regarding future directions for Mozilla that were detailed at the recent Mozilla developers meeting. SVG and integration with programming languages are among the directions discussed."
Here's a direct link to the slideshow itself.
Type n, right-arrow, down-arrow, or space to advance a slide. Type p, left-arrow, or up-arrow to go back one slide. Type t to go the the first (title) slide.
Instructions taken from here
Maybe "Integration with operating system" would help.
FireHydrant is a great OS - If only someone would write a web browser for it.
The Mozilla logo may be so so, but the Mozilla Firefox logo is probably the nicest logo I have ever seen.
Buckethead
Can SVG be expected to take off now if all the developers use flash instead?
What if any SVG based graphic tools are there?
What other benefit besides native browser support will SVG have to use against Flash?
There's plenty to keep them busy for the forseeable future. Lemme see, there's :
Fire - fly
Fire - storm
Fire - engine
Fire - hydrant
Fire - alarm (add-on for the calendar module)
Fire - bird (doh! no already had that one)
Fire - at will
Fire - in the hole
Fire - those responsible
Fire - those who did the firing
Fire - ooh oh oh I'll take you to burn
Come on now, join in everyone ...?
I agree, Mozilla is a bit bloated. However, Mozilla Firefox and Mozilla Thunderbird are meant to tackle problems like that.
The design of Mozilla has been to make it easily embeddable so other developers can use its rendering technology and make their own interface and use a different widget set. Many projects already do, e.g. Galeon in GNOME and K-Meleon (using MFC) for Windows.
My operat~1 system unders~1 long filena~1 , does yours?
Here is the road map to the future of Firefox:
1. Rename Firefox to Foxfire.
2. Add better support for XHTML and CSS 2.
3. Rename Foxfire to Foxxy Brown.
4. Change the XML parsing engine to support new DTMLs.
5. Rename Foxxy Brown to Thunderbird (#2).
6. Put in a proactive pop-up blocker that DoS attacks websites that have pop ups.
7. Rename Thunderbird (#2) to Internet Explorer Jr.
8. Rename IE Jr. to Underpants.
9. Collect Underpants.
10. ????
11. Profit.
Step 10 is going to be the hardest.
You mean, like ActiveX? Er,.....
Tokyo... Check... Going across Pacific... Check... Stomping on Seattle... NYI (not yet implemented) MS should change their browser's name to King Kong, then we would have some fun, eh?
could -one- of you browser whippersnappers please add a 'save browser state/restore browser state' function to whatever the browser de jour happens to be?
...
...)
i want a browser that will remember its state between sessions. if i close the 15 windows i've got open, i want them all back again, same site, same position, when i re-open it again!
sheesh. 15 years of web-browsing, and we're still begging for the most rudimentary, fundamental, web-browsing-workflow features to be implemented, while the rest of the 'web scientists' go off into RFC and NIH land
(apologies if there is actually a 'browser' thats capable of maintaining state information between sessions. please inform me if it'll run on OSX
; -- the corruption of government starts with its secrets. a truly free people keep no secrets. --
Too bad the [Firefox] logo is non-free and will never be checked into public CVS.
Hey, guess what? My signature, my slashdot username and password, and my likeness (i.e., picture),are also non-free and will never be checked into CVS. You can use the ideas in my Slashdot comments, but you can't sign them "orthogonal".
I may grant you a license to use my code -- or other ideas --, but I'm never going to grant you any license yo go around and sign my name to your work. And that's the whole issue here: the Firefox logo is not crucial to the compilation of Firefox code; nothing in the code reads any secret checksums steganographed into the logo.
But the logo is an essential imprimatur that declares a particular build to be an official build, with all that connotes -- such as a well founded belief that it represents the actual work of the official development team and is not likely to be a trojan exploit.
All that not having the logo in CVS deprives you of is bragging rights that aren't yours to claim.
Opinions on the Twiddler2 hand-held keyboard?
gone away?
If not, there is (still) a market for mozilla.
Sometimes I feel like I'm bailing out an ocean, but I'm converting users one at a time. To non-geeks, it's starting to hit home, as to just how bad the crapware is getting. I do a little show and tell. "see this program (points to IE) - BAD!!!", "see this program (points to mozilla) - GOOD!!!". I of course give them a run down (in laymens terms) on how the sneaky stuff gets on their system, and how 99% comes from IE and Outlook Express. After that, all are more then willing to try something different. So on goes Moz!
One thing to remember is that it's very important that you setup Mozilla for them. Make sure the pop-up blocker is enabled. Also set it so that these things are disabled(unchecked):
-move or resize existing windows
-raise or lower windows
-hide status bar
-change status bar text
-change images
Finally. _warn_ _them_ , that Mozilla won't work on every single site. Tell them to fall back to IE on the few sites that don't work(with moz)... But that Mozilla should be first line of defense.
I always found the "users want a standard look across platforms" argument a little ridiculous.
That may have been a justification, but I think that the real reason for Mozilla to have non-native widgets is that it's a lot of work to maintain all the platform-specific codebases. There are already platform-specific issues, but in general someone can add a feature to Mozilla without knowing how to code for every platform under the sun.
I don't know exactly how this will work with native widgets, unless the Moz folks want to take a least-common-denominator approach.
Plus, I wonder if they can rely on sizes of various widgets. Remember that they're integrating widgets with chunks of their laid-out document, when placing, say, a Submit button on the window. With their own widgets, they know exactly how big everything is.
Another issue might be different code structures. For example, the Macintosh Toolbox uses an event loop. GTK uses callbacks. How does one reconcile differently structured widget APIs?
I believe that Netscape Navigator 4.x tried to do this with native widgets back in the day...but the widgets operated different from regular widgets on my classic Mac.
I agree that native widgets would be wonderful from a user standpoint, but there *are* issues with having an extremely cross-platform program with native widgets on each platform. Remember that the MSIE developers only have to worry about one platform...
May we never see th
That's interesting. I've often thought that some bad Acrobat and FireFox interaction is causing problems.
FireFox 0.8 has memory leaks. Load enough instances and tabs, and it will always crash. (This has been verified under Linux and Windows XP.)
When FireFox crashes, it also crashes Windows XP SP1! Windows XP SP1 doesn't show an error message, but the OS becomes unstable, and it is necessary to reboot.
This is shocking to me. The explanation seems to be that the features of Windows XP that most users see run well, but a little below the surface, Windows XP is not a finished operating system. I think a fundamental definition of an operating system is that a real operating system can handle bad behavior of a program without self-destructing. So, after all these years of development, Windows is more a sociological phenomenon than an operating system. It amazes me that Microsoft managers are unable or unwilling to take care of business.
When FireFox crashes under Linux, Linux remains completely stable. (I suppose you could have guessed that.)
I have copies of all the browsers, and in my opinion FireFox is by far the best. Browsers are windows on the world for an increasing number of people, so it is important that the world has an excellent one.
I think FireFox's memory management issues should be fixed before any other work is done. Of course, that is for the FireFox/Mozilla team to decide.
(Posted using FireFox, of course.)
SVG is much different from Flash. Flash is currently primarily used for two things: (1) to provide crummy interfaces (an ugly wart from designers coming from the "multimedia era" when CD-ROMs came out and later the ".com era" when people thought that novelty was what made people keep coming back to websites). (2) To provide an efficient format for vector-based graphic animation.
SVG is lousy at both of the above. I have a friend that looked into the feasibility of SVG as an interface medium, and came back pretty depressed. At one point, I got a bit interested in using SVG for animation, and took a look at the format. I'm reasonably comfortable making the claim that it would be extremely difficult to make an efficient rendering engine for animations using SVG. Furthermore, SVG does not provide functionality for synchronizing audio and phases of an animation (which I believe Flash does).
SVG is good, IMHO, for the following:
1) Tagged diagrams. SVG allows tagging elements with data. This could be a big benefit for CAD and diagram usage.
2) More complex webpage layout. I've never seen it actually done, but it seems that SVG could be used to define arbitrarily-shaped regions in a webpage...up until now, the only regions designers have had to work with, the only thing they could flow text around, was rectangular regions
3) Vector graphics. Plain and simple, it's a standard format for storing vector graphics. This is good for both standalone files and for efficient web-based transmission of graphics.
As for your question about what SVG-based graphic tools are out there -- take a look at sodipodi. It isn't Illustrator (yet), and it isn't going to be for at least a while to come, but it's usable for basic work.
May we never see th
I'm seeing a lot of comments in reply to this article advocating that the mozilla foundation stick to making web browsers, a task that it now admittedly does very well. Follow the Unix philosophy, small programs that do one thing and do it well.
I agree with the philosophy, and agree with what the foundation has done in starting the firefox/thunderbird fork.
But I feel the issue isn't as simple as some fellow /.ers are saying it is, and the longterm prospects are definitely interesting. The key topics mentioned in this slideshow (SVG, XUL, XBL, Eclipse plugin, scripting language integration) are all focussed around the central issue of what the words 'web application' are going to mean in the future.
Think back to several years ago in the dark ages of IE4.0 sheer dominance, when you were hard pressed to find an online banking service that would permit your alternate browser inside without you having to spoof a UA string. Microsoft had defined the standards that the web developers had been using, and we suffered for having a just standards compliant browser set.
We are now at a lull in the web application development market, at least from the client side. Sure on the server side the battle wages ever on, but the front end is pretty sown up. But it won't remain that way. Nothing like that does in this industry.
This is a proposal to start heading the mozilla project in the direction of a web development framework. Extending the front end possibilities, and giving developers the tools to close the gaps between web applications and thin client applications.
Microsoft is heading in this direction. Rumours are that the next major IE that will ship with longhorn will have a framework similar to this idea, with complete integration between the HTML forms and the windows.form components Microsoft is working on. If we stay statically focussed on supporting just the W3C standards, which don't extend to something as encompassing as an application framework, then Microsoft will be allowed to take the iniative again.
At best, this is an attempt to refocus upon what XUL was originally a vision of, just done right this time. At worst, its an attempt to think long term and make sure we aren't taken by surprise when Longhorn ships with a new beast of an IE. We need a framework like this, and I see noone in the opensource world in a better position to do this than the mozilla project.
Mozilla seriously needs more threading. I hate not being able to interact with anything for a few seconds whenever a tab is loading in the background.
For great justice.
Agreed, small chunks are better. Thats why breaking up the original suite was a good idea. But a framework is just a collection of small pieces. Firefox for instance may still just be shipped with what is essentially just a wrapper for the networking and the layout modules. In fact, frameworking like that would probably require factoring the existing code into even smaller discrete chunks. If people want to be able to run a thin client application that uses the mozilla framework, then it could run off and download the relevant XPIs (which you would keep very small) by itself as it needs to. As an example, at the moment MPlayer is undergoing a major redesign led by Arpi in the form of MPlayer G2. It too is much more of a framework than MPlayer is, but in terms of monolithicism and bloatedness, its better in every way.
People frequently ask why Mozilla implements its own widget set rather than just using the widget set available on whatever platform it's running on. This document is an attempt to explain why. Transparency and Z-ordering
Consider this testcase. It's a text field behind an element full of "blah" text. The "blah" element is transparent, so you can see and even edit the text field with the "blah" text overlaid on top. This simply can't be done in with Gtk or Qt widgets (unless this has changed in a very recent version of these toolkits). In Win32 it can only be done in Win2000 or WinXP, and then it is tricky and inefficient. If you don't believe this, try implementing the same effect using your favourite platform toolkit, and email me if you succeed.
Getting this right isn't optional. It's a requirement for a correct CSS implementation. Other HTML/CSS functionality
An HTML BUTTON element can contain arbitrary HTML. It's practially impossible to get that to work with any platform button widget. (Note that the HTML inside the button is part of the same document as the button itself.) Printing
On many platforms it's very difficult or impossible to get a native control to print. International languages
When you browse the Web you find content in every language that computers can handle. It is important for the browser to have strong support for uncommon languages. This means it is important for the browser to display form elements containing strange characters and scripts. Many platforms (e.g., older versions of Windows) do not provide good support for locales other than the locale that the operating system itself is installed for. Therefore their widgets aren't good enough for strong browser language support. Performance
On many platforms the per-widget memory and time cost is quite significant. This is OK for most GUI apps because you typically don't have more controls per window than fit on the screen. But in a browser, you sometimes see pages with hundreds or thousands of controls. (Think "a long comments page in Slashdot when you have moderation points".) This has to be fast and not consume too much memory. On some older Windows versions it's simply impossible to create 1000 edit boxes without crashing the system! Event handling
The DOM Events model defines ways for a page to intercept events such as keyboard or mouse input before they are dispatched to the control with focus. It would be very tricky and error-prone to implement this using platform-specific hacks. Arguments For Native Widgets
Here are some arguments for using native widgets, and how we answer them. Native look and feel are critical for usability
Agreed. We have started using platform-specific APIs to render our widgets as if they were native widgets, wherever we can. For GTK, WinXP and MacOSX we actually call theme APIs so that Mozilla picks up whatever theme is currently in force. It really looks like a native app. All of the above advantages are still retained because we're still not using actual native widgets. It also means we automatically "keep up" as the platform look changes, which has been a big problem for "cross platform" UI toolkits in the past.
We're still working on the "native feel" problem. Feel doesn't vary as much as look, it seems, so it's less of a problem, but we have a number of tweaks that vary the feel of our widgets across platform and we'll add more. Native look and feel are critical for accessibilty
We're building in support for platform accessibility APIs in GTK and Win32, so our widgets will be just as accessible as the native widgets. Too much work for developers
Yes, but it's worth it. Too slow, too much footprint
Yes, rolling our own widgets requires some extra code and may not be as well optimized as the platform widgets. But as noted
A browser that kills your machine every hour? I suppose it will be all the kernel level drivers that Mozilla installs that cause that. Oh wait, it doesn't have any. Mr Senior IT Manager for a Corporate, you should know that a userland app should never be able to take down an OS, Windows or not. And you'd know that more often than not, XP is configured not to display a blue screen, but just to reboot. My advice? Check that it is configured to stop on a crash. Apply all the patches. Disable services you don't need. Use Firefox again, and see if it crashes the OS. If it does, make a note of the info on the blue screen, and Google for it. Try swapping the memory/cpu with another similar machine.
But don't go blaming Firefox for crashing your machine.
Get your own free personal location tracker
"Black Screens Of Death" are usually caused by faulty RAM.
I suggest you try some different RAM chips and try Firefox again.
Although, realistically, Black Screens Of Death should occur randomly, not just when using Firefox.
Also various video card drivers are known to screw up your memory and go down with a Black Screen Of Death
The path I walk alone is endlessly long.
30 minutes by bike, 15 by bus.
Yes, and I doubt it's Windows XP on its own crashing it either. Must be some hardware or other software doing it. I'm also having trouble seeing how Firefox, of all software, could take down XP as it very rarely crash due to bad software. Writing outside its allocated memory will for example only force XP to crash the application.
Beware: In C++, your friends can see your privates!
Yeah, i stopped using msn messenger, er.. windows messenger, or whatever it is.. the name changed so many times I just couldn't handle it.. ok.. it was always obvious to me what it was, and ok the icon didn't change too much.. but the name.. the name changed.. my life was ruined.. I broke down at work and needed a months holiday.. my wife divorced me, the dog died.. and the world as I know it came to an end.. all this from the name change.. but finally i just stopped using it, and went to investigate more promising and viable entries in the IM sweepstakes.
It looks like some furred animal is humping the Earth.
Not that there's anything wrong with that...
And before you flame me, I'm a staunch Mozilla fan and this was posted using Firefox.