Right. He compared both browsers out of the box (the way most people actually use them, fwiw).
You can check whether the extensions are the issue by running Firefox in safe mode; I'd be interested in knowing how that affects the startup performance.
The other thing is that Taras was comparing development versions of both browsers. Firefox 3.6 has a bunch of startup improvements over 3.5, but even more have been made since then.
The problem is that "go fater" is very subjective (not for benchmarks but for things like "snappy", "fast pageload", etc). Most users don't look at benchmarks. They just use whatever "feels faster"....
Well, for one thing Firefox made the mistake of listening to what the Linux users and distros were telling it and actually integrated with the GTK theme system, fontconfig, offloads graphics ops to the X server, etc. Chrome does none of that, rolling its own instead.
Unfortunately, the GTK theme system and fontconfig are slow, and the X server version of pixman tends to be several years behind the one shipping in cairo (which Firefox also includes, of course... but doesn't use because the Linux folks all claimed that using the X server is the Right Thing(tm)).
Just not trying to do the whole native theming business through GTK dramatically speeds up the Firefox UI.
> If it is WOFF, what prevents one from decompressing and installing it locally?
Nothing, just like nothing prevents you from recording songs off the radio.
The key is that it makes it impossible to say you didn't know you had the font on your system, or that it was accidentally dragged from your cache folder to your fonts folder or whatnot. The compression is not meant as DRM but as a way to make the font smaller, from the UA point of view. From the foundry point of view it makes the "my browser just put this decompressed font on my system" defense not fly: if it's there and decompressed, you decompressed it or got it from someone who did.
If we're talking about Safari on iPhone/iPad (as GP was), then in fact Apple has the power to prevent people changing browsers without going through a great deal of pain (or switching to totally different hardware).
Of the browsers which come close, IE8 is arguably closest. Something to do with getting to implement _after_ all the spec changes that have happened over the last 5-6 years, not as they're happening, for one thing.
OPEC is a cartel of governments, not oil companies.
Notably, BP is NOT an OPEC member and has no official say in OPEC decisionmaking. Unless it's become a sovereign country and joined when I wasn't looking?
You should read the rest of the law text, which also talks about _who_ has to encrypt the information under the law. Unless you're selling goods or services to yourself, you're clear!
> Actually, I read the article that was referenced in the summary, and the article that was > referenced in that article. Neither one said anything like what you just posted.
As usual on most topics, the articles are more or less complete bullshit. The text of the law (all 4 pages of it) is at http://www.mass.gov/Eoca/docs/idtheft/201CMR1700reg.pdf and the definition you want is on page 2 under "Personal information" in the alphabetical list of definitions.
What I find scary, really, is that any time I see an article on a topic I know something about it's pretty bogus. Do I really have any indication that the press does better on topics I _don't_ know about?:(
Chrome's marketing is a lot more aggressive than that of any other browser. I haven't seen any of the others taking out huge billboard ads all over the London subway, for example.
Because they're comparing browsers on the same hardware and OS; in their case on Windows.
And Safari on Windows has a quite different rendering pipeline from Safari on Mac; it can't leverage the same set of system frameworks, etc (and the general drawing model is quite different too). So on benchmarks which are gated on rendering I would expect significant differences between Safari on Windows and on Mac. I would expect similarly significant differences between Firefox on Windows and Mac and Linux (especially Linux; high-performance graphics without just using OpenGL is a huge pain on Linux).
This is why, for example, webkit only supports 3d transforms on Mac: they just offload the work to a system library that exists on Mac but not elsewhere.
> Shareholders sue corporations all the time for not maximizing shareholder value.
Sure. Do they win? Both of your articles there are about suits being _filed_, not being _won_. Filing a suit is easy!
The Guth_v._Loft_Inc reference is much more interesting. Thank you. That puts a much firmer basis under the "corporate officers are required to maximize shareholder value" claim (though with the unspoken assumption of a publicly held company; I would not be surprised if the vast majority of corporations in the US, unweighted by revenue, are privately held).
Thank you for the citation! That does seem like a pretty clear-cut case of "you're not using the money to increase shareholder value". The hard case to win would be one where you reduce dividends and do something to develop your business with the money (something that was _not_ going on here according to the Wikipedia article, which also explicitly points out that this case is often incorrectly used in support of the "must maximize shareholder value" viewpoint).
But at least now I have some idea of where the meme is coming from.
I assume GP's point was that the money could be better spent employing people in Afghanistan, so that they have something better to do than join the Taliban and whatnot. It's not clear to me that GP is right, mind you.;)
Really? Citation, please? This idea gets thrown around a lot, but I have yet to see evidence.
One can argue that unless the articles of incorporation suggest otherwise the fiduciary duty of the management is to maximize shareholder value (NOT to be confused with instantaneous profit) and that shareholders could sue the management in civil court over perceived failures to do so. It'd be a pretty hard suit to win, I bet.
But in general, the purpose of a corporation will be outlined in its articles of incorporation. It might have nothing whatsoever to do with shareholder value, much less profit; plenty of non-profit incorporated entities out there.
Another major fallacy here is considering the war in Vietnam as a war of "the west" against "the communists".
If we view the war as part of a global political struggle without assuming that countries were monolithically on one side or the other of that struggle, then the winners of the war in Vietnam (which would include quite a number of left-leaning intellectuals in the US and their followers) _are_ in fact writing the history books.
Slashdot Mirror
Yeah, I'm not saying older Firefox builds didn't start slowly. They did. We'll see how Fx4 does.
> To be fair, I run the chrome nightly build (available as a ubuntu package).
Ah. Well, ok. Comparing apples to oranges... ;)
> Do you know of any data showing how many extensions the average user uses?
Depends on how you define "average user". http://blog.mozilla.com/addons/2009/08/11/how-many-firefox-users-use-add-ons/ says that somewhere between 33% and 53% of users had at least one extension installed. So the average user uses 0-1, is my guess.
> He used no extensions.
Right. He compared both browsers out of the box (the way most people actually use them, fwiw).
You can check whether the extensions are the issue by running Firefox in safe mode; I'd be interested in knowing how that affects the startup performance.
The other thing is that Taras was comparing development versions of both browsers. Firefox 3.6 has a bunch of startup improvements over 3.5, but even more have been made since then.
Or you use open tabs as your todo list.
Curious. What version of Firefox? And on which OS?
You may be interested in http://blog.mozilla.com/tglek/2010/01/19/chromium-vs-minefield-cold-startup-performance-comparison/
The problem is that "go fater" is very subjective (not for benchmarks but for things like "snappy", "fast pageload", etc). Most users don't look at benchmarks. They just use whatever "feels faster"....
Well, for one thing Firefox made the mistake of listening to what the Linux users and distros were telling it and actually integrated with the GTK theme system, fontconfig, offloads graphics ops to the X server, etc. Chrome does none of that, rolling its own instead.
Unfortunately, the GTK theme system and fontconfig are slow, and the X server version of pixman tends to be several years behind the one shipping in cairo (which Firefox also includes, of course... but doesn't use because the Linux folks all claimed that using the X server is the Right Thing(tm)).
Just not trying to do the whole native theming business through GTK dramatically speeds up the Firefox UI.
Opera mini does not include a JS engine.
> If it is WOFF, what prevents one from decompressing and installing it locally?
Nothing, just like nothing prevents you from recording songs off the radio.
The key is that it makes it impossible to say you didn't know you had the font on your system, or that it was accidentally dragged from your cache folder to your fonts folder or whatnot. The compression is not meant as DRM but as a way to make the font smaller, from the UA point of view. From the foundry point of view it makes the "my browser just put this decompressed font on my system" defense not fly: if it's there and decompressed, you decompressed it or got it from someone who did.
You assume most people will consider the inability to change browsers lame. History doesn't necessarily support that view...
If we're talking about Safari on iPhone/iPad (as GP was), then in fact Apple has the power to prevent people changing browsers without going through a great deal of pain (or switching to totally different hardware).
No one has quirk-less CSS2.1 support.
Of the browsers which come close, IE8 is arguably closest. Something to do with getting to implement _after_ all the spec changes that have happened over the last 5-6 years, not as they're happening, for one thing.
We're talking about H.264 here, right? You may want to read http://weblogs.mozillazine.org/bz/archives/020400.html
OPEC is a cartel of governments, not oil companies.
Notably, BP is NOT an OPEC member and has no official say in OPEC decisionmaking. Unless it's become a sovereign country and joined when I wasn't looking?
You should read the rest of the law text, which also talks about _who_ has to encrypt the information under the law. Unless you're selling goods or services to yourself, you're clear!
> Actually, I read the article that was referenced in the summary, and the article that was
> referenced in that article. Neither one said anything like what you just posted.
As usual on most topics, the articles are more or less complete bullshit. The text of the law (all 4 pages of it) is at http://www.mass.gov/Eoca/docs/idtheft/201CMR1700reg.pdf and the definition you want is on page 2 under "Personal information" in the alphabetical list of definitions.
What I find scary, really, is that any time I see an article on a topic I know something about it's pretty bogus. Do I really have any indication that the press does better on topics I _don't_ know about? :(
Chrome's marketing is a lot more aggressive than that of any other browser. I haven't seen any of the others taking out huge billboard ads all over the London subway, for example.
> HTML5 will do nothing to replace embeddable TrueType fonts
No, but CSS fonts do.
> vector animation
SVG or canvas depending on whether you want retained mode or immediate. Granted, SVG is a bit of a pain to work with.
> built-in MP3 and MP4 support,
This is tougher, since it has the patent issues we're discussing right now, right?
> etc., etc., etc.
Details?
The fact is, the general web standards push right now seems to be in the "make Flash irrelevant" direction.
Because they're comparing browsers on the same hardware and OS; in their case on Windows.
And Safari on Windows has a quite different rendering pipeline from Safari on Mac; it can't leverage the same set of system frameworks, etc (and the general drawing model is quite different too). So on benchmarks which are gated on rendering I would expect significant differences between Safari on Windows and on Mac. I would expect similarly significant differences between Firefox on Windows and Mac and Linux (especially Linux; high-performance graphics without just using OpenGL is a huge pain on Linux).
This is why, for example, webkit only supports 3d transforms on Mac: they just offload the work to a system library that exists on Mac but not elsewhere.
Firefox doesn't use OpenCL, to my knowledge. Can't tell you anything about Safari.
> Shareholders sue corporations all the time for not maximizing shareholder value.
Sure. Do they win? Both of your articles there are about suits being _filed_, not being _won_. Filing a suit is easy!
The Guth_v._Loft_Inc reference is much more interesting. Thank you. That puts a much firmer basis under the "corporate officers are required to maximize shareholder value" claim (though with the unspoken assumption of a publicly held company; I would not be surprised if the vast majority of corporations in the US, unweighted by revenue, are privately held).
Thank you for the citation! That does seem like a pretty clear-cut case of "you're not using the money to increase shareholder value". The hard case to win would be one where you reduce dividends and do something to develop your business with the money (something that was _not_ going on here according to the Wikipedia article, which also explicitly points out that this case is often incorrectly used in support of the "must maximize shareholder value" viewpoint).
But at least now I have some idea of where the meme is coming from.
I assume GP's point was that the money could be better spent employing people in Afghanistan, so that they have something better to do than join the Taliban and whatnot. It's not clear to me that GP is right, mind you. ;)
Really? Citation, please? This idea gets thrown around a lot, but I have yet to see evidence.
One can argue that unless the articles of incorporation suggest otherwise the fiduciary duty of the management is to maximize shareholder value (NOT to be confused with instantaneous profit) and that shareholders could sue the management in civil court over perceived failures to do so. It'd be a pretty hard suit to win, I bet.
But in general, the purpose of a corporation will be outlined in its articles of incorporation. It might have nothing whatsoever to do with shareholder value, much less profit; plenty of non-profit incorporated entities out there.
Another major fallacy here is considering the war in Vietnam as a war of "the west" against "the communists".
If we view the war as part of a global political struggle without assuming that countries were monolithically on one side or the other of that struggle, then the winners of the war in Vietnam (which would include quite a number of left-leaning intellectuals in the US and their followers) _are_ in fact writing the history books.