Slashdot Mirror
Firefox's Effect On Other Browsers
An anonymous reader points out an interview with Mozilla's "evangelist," Christopher Blizzard, regarding the future of Firefox and how it affects other browsers. It's an Austrian site, so forgive the comma abuse. From derStandard:
"It's sort of interesting though, part of our strategy is to make sure, that we continue making change and the indirect effect of this is that Microsoft continues to have to do releases, because if we get so far ahead that we're able to drive the platform they are not able to keep up and keep their users. I mean, we have this joke which says 'Internet Explorer 7 is the best release we ever did,' because they would not have done it, if we would have not built Firefox. And the same is true for Apple, they are doing a lot to keep up with us. Safari 3.1 is a good example, as far as we see it, the only reason they did this release was that Firefox 3 would come out and have Javascript speed which would be twice as fast as theirs, cause that's how it was before. So by pushing other people to make releases we can go on our mission to make sure the web stays healthy."
What astonishes me is more that this latest release has gotten even my totally non-tech-savvy friends to download it and acknowledge its superiority to Internet Explorer 7. The Firefox team has not only improved the browser for those of us who already used it, but managed to convert another large segment of the market. It's sort of like the Nintendo Wii effect -- they realized it made more sense not to enlarge their slice of the tech-savvy pie, but to expand the pie to include casual users as well. Or at least that's how I see it, feel free to correct me with your own interpretation.
Or maybe they did it because they were pushing javascript apps for the iPhone, and working on the javascript-based SproutCore frameworks and the associated MobileMe apps.
Not everything revolves around Firefox.
Apple did not release Safari because of Firefox. After all, Firefox was on Apple. They released it because they wanted to be in control of their future. As it was, MS had announced that they were going to pull MSIE from them. What amazes me, is that Apple has not pushed OO to be on there. They would be smart to add a few coders to the project just to ensure that it can compete against Office on their platform.
I prefer the "u" in honour as it seems to be missing these days.
Comment removed based on user account deletion
I always maintained that Win2K was such a good OS specifically because of the competition Microsoft was getting from open source, they didn't want to be caught napping and wake up to find Linux as a good desktop solution. This theory kind of fell apart with Vista, I have no idea what that steaming pile is in response to.
Kwisatz Haderach
Sell the spice to CHOAM
This Mahdi took Shaddam's Throne
Safari is not trailing Firefox as it is being developed in all ways, especially JavaScript performance. I actually prefer to use Firefox 3 on the Mac (much better array of plug-ins, and better security), but the latest WebKit nightlies, on http://www.webkit.org/ since the implementation of Squirelfish (see blog there) are quite a bit faster in JavaScript performance than Firefox. If anything, Firefox is going to have some catching up to do in that department.
And Opera is feeling so pressured by Firefox that it is systematically forced to copy Firefox's features months and even years before Firefox releases them... ^_^
What will Firefox copy next? (what? troll?)
middle-click-to-close on tabs comes to mind
It's hard to tell between a left-click, middle-click, and right-click on a one button mouse...
While Firefox may have inspired the release of IE 7 and pushed Apple to jump into the fray with a Windows release of Safari, it is also true that FF 2 was not all that it should have been and just maybe IE 7 and Safari pushed Mozilla hard enough to really ace FF 3 which it seems that they have done.
As a software developer who once loathed the idea of having to code for multiple browsers, I have now accepted that there will be differences and have learned to deal with it and promise to stop whining.
I applaud the browser race and hope that they continue to leapfrog each other for a long time to come.
is that Firefox has been driven (to a large extent) by Opera.
Credit where credit is due, please.
"National Security is the chief cause of national insecurity." - Celine's First Law
I never knew, that German, was quite so, comma-happy.
Colin Dean Go a year without DRM
Apple does very little of the core work for Safari. They just take the open-source WebKit engine and slap their own UI over it
You are incredibly misinformed. A quick glance at recent WebKit changes readily shows how blatantly false your claim is.
They just take the open-source WebKit engine and slap their own UI over it
WebKit was developed by Apple, originally as a fork of KHTML for their Safari browser. Apple open-sourced WebKit and it was so good that many of its improvements were copied back into KHTML. It's also being used by a number of mobile phones because of its strengths relative to e.g, Gecko, including Android.
Without Apple, there would be no WebKit. But don't let reality get in your way.
I have a problem with pop-under windows. They "reappeared" recently, and I'm using FF exclusively. Unfortunately I can't tell if my switch from FF2 to FF3 was the reason, but it was around the time. Is this a known bug? I know I can try to figure out the domains of the sites appearing in those unwanted windows, but I'd be more interested in a general solution. BTW, I have "block pop up windows" activated in the settings, with a few exceptions.
Apple open-sourced WebKit and it was so good that many of its improvements were copied back into KHTML.
Umm, KHTML was licensed as LGPL, which means Apple had to open source their fork if they distributed it. As for improvements being copied back, well that happened to some extent, but the Konquerer team seems to have pretty much given up on KHTML and are contributing to Webkit now.
KDE open-sourced KHTML. Apple didn't have a choice in the matter.
Nonsense. KHTML is LGPL. Apple could have used the libraries without contributing anything back.
Moreover, the DOM is Apple's, not KHTML's. WebCore, the basic component of WebKit, has very little relationship to KHTML.
It was so divergent that the KDE folks pretty much had to accept WebKit as the new KHTML if they wanted to accept the improvements.
That's not at all true. Most of the improvements shared back upstream, including KHTML's ability to pass Acid2, were adapted prior to the merger. KDE adopted WebKit by choice. There was nothing stopping them from continuing development of KHTML separately, nor was their any requirement that the KDE people actually adopt any of Apple's improvements.
Sour grapes that KHTML was largely abandoned in favor of something better doesn't explain why it's WebKit, and not KHTML, that is being adopted by other platforms.
Yes, because holding two fingers on the trackpad and then clicking is so much easier than just clicking the other button...err, wait...
Actually it is. According to the usability tests I've seen, it is faster and has a lower failure rate because to hit the second button you have to either stretch your hand over or use your other hand, neither of which is ideal. For mice, where one hand is already off the keyboard, multiple buttons are a usability win for experts, but for trackpad users it is a loss for novice users and expert users and more usable but less learnable for middle of the road users.
No, you're thinking of GPL. The LGPL would have allowed them to use KHTML libraries without giving anything back.
They can link to it without giving anything back, but the LGPL does not allow them to make changes to it and distribute them without giving the source back. Since Apple had to make significant changes to make it work modularly and the way they wanted, they had to give all those changes back. They don't have to open source the code for Safari, which links to Webkit, and in fact they don't.
WebCore's "improvements" are largely Apple's own doing, apart from those changes which were shared upstream before KDE developers abandoned KHTML.
Apple has done significant work to make Webkit better than KHTML was, but they are certainly building on a lot of work that was done before they entered the game. Apple has played nice with the Konqueror folks and gone out of their way to help them integrate changes and revise the way the shared code base was developed such that improvements from multiple groups including Konqueror, Apple, and Nokia can all be included. That said, to claim Apple had a choice about how Webkit would be licensed or if their changes to it would be open source is simply not true.
the LGPL does not allow them to make changes to it and distribute them without giving the source back. Since Apple had to make significant changes to make it work modularly and the way they wanted, they had to give all those changes back.
You are still being imprecise. The LGPL does allow them to make whatever changes they like, so long as the KHTML libraries they are using are used intact. I do not disagree that any modified libraries had to be shared back upstream, but those changes are portions of WebCore, itself a portion of WebKit. There was no requirement that compartmentalized changes, improvements, and additions be shared if they extended beyond the four corners of the KHTML libraries.
WebCore is much more than rewritten KHTML libraries. WebKit is much more than WebCore.
That said, to claim Apple had a choice about how Webkit would be licensed or if their changes to it would be open source is simply not true.
It absolutely is true. There was no obligation to open-source WebKit. There wasn't even an obligation to open-source the entirety of WebCore and JSCore. There was an obligation to share changes to modified libraries.
What's simply not true is that Apple had no alternative. Apple provided WebKit tactically, not out of obligation to disclose it in its entirety and certainly not out of the goodness of their "hearts".
That's incorrect. Changes made internal to LGPL software must be released, it's only external software that links to LGPL that has the right to stay closed.
For what you're saying to be true KHTML would need to be generic enough to be modified by Apple via linking rather than changing any of the internal KHTML code. The changes Apple have made did involve digging into and changing the guts of KHTML. Again, for what you're saying to be true then KHTML would have to be little more than a canvas for Apple to draw upon.
However we know it's not, and that the changes for progress were internal to the software and therefore Apple did not choose to open source it -- it was open source due to the KHTML license.
This is true, other than that Apple approached it as a fork. They didn't take the time to join the KHTML team and win over the developers with strong arguments and robust debate. It certainly wasn't that kind of software development.
Instead what Apple did was divergent, it was effectively a fork, and KDE chose to go with the fork (probably due to the quality of the code). I personally think that what Apple did was acceptable -- it's permitted by the licence. They could have managed the community in a smarter way but then they like being secretive. It's resulted in some great contributions. Overall I think it was a positive thing.
Tabbed browsing, clean mouse gestures, two-handed browsing, single-click image disabling, single-click user CSS mode.. heck, most of the user-friendly advances have been standard features on Opera for many, many years. And half of the really good stuff *still* isn't stock and standard on any other browser.
But, Opera did open its doors to the free download hype as a result of Firefox. So I owe you that much. :)
But.. catch up already would you?
LGPL isn't the same as a BSD permissive-style license,
No. But neither is it the same as the GPL generally.
The reason it was known as the Library GPL is that it allowed the non-contributory use of GPL'd libraries by other types of software licensed under terms incompatible with the GPL.
The KHTML library changes would have had to be shared per the terms of the licenses. This requirement, however, does not even encompass all of WebCore, let alone WebKit.
As far as I know, however, any changes or improvements made to the LGPL'ed programme itself must be distributed Freely, with source, if it is to be distributed at all.
Any changed or improvements to the LGPL'd software, which it is a complete program or a library. In the case of KHTML, it is a set of libraries. Those libraries were adopted into the codebase for WebCore--and only those libraries derived from the KHTML libraries would need to be shared.
It does not extend to other libraries written by Apple or any other developer, and it does not extend to products merely containing those libraries. Limiting that "wagon-hitching" (widely, and in some ways regrettably, known as "parasitic") effect of the GPL is the reason the LGPL exists in the first place.
A workaround for this is to run Flash inside nspluginwrapper, even if you're on a 32-bit system.
This way, when Flash crashes, it won't bring down the whole browser with it, and all you have to do it reload the page.
This bug is on Ubuntu's bugtracker.
You obviously didn't deal much with Win 3.x. Registry is much better than config files scattered throughout. I wouldn't mind if it were replaced, but it needs to be a step forward, not back. Linux still has config files scattered in a zillion different places. It would be nice if all configs went into an organized hierarchy.
Um. All the configs do go into an "organized hierarchy"! It just happens to be a filesystem hierarchy (/etc) rather than an impenetrable binary file.
XML files located in a couple of standardized locations. As in one location for machine level configs, and one location each for user level configs.
XML sucks for configuration files, to be honest. Trying to hand-edit XML in a 40x80 nano session in single-user mode... no thanks. Not to mention that XML is decidedly grep-unfriendly.
I think I'm going to have to assume that you don't have a clue what you're talking about, I'm afraid. Try harder!
Pirate Party UK
you mean like /home/user/.someprogram/config.conf for user level configs and /etc/someprogram/config.conf for machine level configs?
KHTML provided the HTML and XML parsing engine, the DOM tree exports, the CSS parsing engine, the layout engine.
Source? All indications are that Apple wrote their own DOM, and that their CSS parsing is not KHTML's (which was one of the problems in adapting changes back to KHTML years back). They also certainly wrote the SVG support, which KHTML lacked.
That means that WebCore is a derivative of a pair of LGPL'd products.
No. WebCore does not contain KJS code. That's JSCore. WebCore contains LGPL'd libraries from KHTML, but it contains libraries that are not part of KHTML as well. JSCore contains LGPL'd libraries from KJS.
Further, since WebKit is, apparently, a derivative of WebCore
No. WebKit is a wrapper, providing API-level access to WebCore and JSCore, as well as integrating the debugging unit (starts with a D...). It is not a derivative work for the purposes of the LGPL.
Note: IIRC, WebKit and WebCore are parallel products - one isn't built on top of the other, but one was forked from the other.
No. WebCore is a component of WebKit.
I totally agree that Linux use of plain text files in the "/etc" directory is a far superior solution. However I'd also like to see all the user level config files that currently go into the various "~/.prog_name" folders collected into something like a "~/etc" directory.
Obviously to hide it during "normal" use you could name it "~/.etc" but I do think that it would be more consistent and far tidier to have all the user level config files in their own subdirectory.
Mind you having said that I'd prefer the directories were called "/settings" and ~/.settings" but I suppose 50 years of *NIX cruft precludes this !
Sky subscribers are morons. They pay to be advertised at !
or even competition. I don't view them that way (I don't pay for any of them) - they're just different choices.
And to all those ignorant mods who called me a troll: Opera has been around in fairly significant numbers since about 2000. Even if it had minimal market share, that is the timeframe in which it became noticed by the web cognoscenti. Firefox came out around the end of 2004 (pre-Mozilla came out around the end of 2002).
At the time Mozilla/Firefox was being formed, IE was pretty static, with no significant feature development occurring (IE6 in 2001, IE7 not until 2006). IE certainly wasn't driving feature development in other browsers. Safari didn't even exist in public until 2003.
In addition to the obvious tabbed browsing (no, they didn't invent tabs, but they did popularize them in browsers), Opera has also set the bar for standards support and rendering speed.
Specifically with reference to the article and Mozilla/Firefox, the three most significant UI features of Mozilla/Firefox, tabbed browsing, easy inline find, and custom shortcuts, all appeared in Opera previously.
Yes, Opera has been a significant factor in driving feature development in other browsers, and it deserves that recognition and respect, even if you choose to use something else.
"National Security is the chief cause of national insecurity." - Celine's First Law
The biggest reason why Opera never got "traction" among Windows and MacOS users was that up until 2005, you couldn't get a full-featured version that was truly "free" (and that meant no ads either!). Meanwhile, IE came as part of Windows since Windows 95 OEM Service Release 2, and of course Firefox has always been free to download (the "free as bheer" thing is really enticing in this case).
Sure, Opera invented a lot of the features we take on IE 7.0 and Firefox 3.0.x for granted, but because of the price issue Opera was never really taken seriously as a competitor to IE and Firefox.
IE7 has the security and reliability. It's also quicker than FF and doesn't leak memory like a sieve
In the classic battle of IE7 vs FF2, he's absolutely right.
I tried FF2 a few years ago when everyone seemed unable to get enough of the kool-aid. While superior to IE6 for its tabbed browsing, once IE7 rolled out, FF2 lost its only edge.
Today, I run FF3 with minimal addons. I don't use NoScript, because it turns normal web browsing into a circus of "allow" clicks, and makes UAC look good.
Still though, I refuse to drink either side's kool-aid. Firefox is not the shining gift from heaven some people think it is, and IE is not the complete trash slashdotters generally insist it is.
Boot Windows, Linux, and ESX over the network for free.
...you mean something like /etc/ and /home/ ?
"Scattered" is just another way of saying fault tolerant.
If all of my eggs aren't in one basket, and a poorly constructed
basket at that, then I don't have to worry about them being all
destroyed by one stupid mistake.
One app doesn't NEED a mechanism by which it can conveniently
destroy everyone else's configuration or the configuration for
every other app.
Once you centralize something you need to start thinking about
how little things like disaster recovery and change management
are going to be handled. "Some XML file" just doesn't cut the
mustard. It's another registry quagmire waiting to happen.
It's this simpleminded "convenience centric" thinking that clobbers Windows.
You've identified no clear problem that needs to be solved with this poorly thought out new mechanism.
A Pirate and a Puritan look the same on a balance sheet.