Uh... I suggest you do some reading about how JITs actually work and then try to understand how the differenet levels of a JIT interact with each other. Macromedia/Adobe put a bunch of work into nanojit (the one Adobe bit Tracemonkey uses), but then Mozilla did as well. And Mozilla built all the code that has to generate the LIR, which is where a lot of the optimization work happens.
Similar for Jaegermonkey: the assembler was borrowed from JSC, but the compiler code on top of it was written by Mozilla.
It's really worth learning about things before making false claims about them.....
A lot of this "feature creep" is web platform stuff. Audio+video support. More SVG support. SMIL support. CSS Transitions, CSS Animations. And so forth. If you just compare the size of the core rendering library across those releases you will see similar if not faster growth. Providing these things through "plugins" is not really desirable unless you mean something preinstalled. And that would actually _increase_ bloat due to the need to handle both SMIL being available and SMIL not being available.
Some more of the size increase, by the way, is a move toward compile options that optimize for performance at the expense of code size...
Chrome still has H.264 support last I checked. They announced that they were planning to take it out, but they never said when, and they haven't done it yet as far as I know.
The site was just broken in non-Chrome browsers in various ways. For example, for a while it was serving up both H.264 and Theora video, with Chrome using the H.264 and Firefox using the Theora. Except the Theora url was 404, so it didn't work in Firefox.
Similar things elsewhere, where the site takes different codepaths in different browsers and some of them are just not tested very well.
As of 2009, Mozilla had $120 million in net assets. Expenses in 2009 were $61 million. Revenues were $104 million. They were hiring as fast as they could find good people, and earning more money than they could spend. They had 2 years worth of operating costs in the bank. All of this is public data, as it is for any other nonprofit.
So if trends continued in that revenue and expenses grew at the same percentage rate, and if you assume that Google is still 85% of their revenue stream (the data on that doesn't seem to be available), what would happen if Google pulled out is that Mozilla would have about 2.3 years to find funding sources to replace that revenue. Assuming they kept spending as much as they do now in the meantime instead of trying to stretch the money out.
On the other hand, you also have to wonder what the bottom line for Google would be from 20-30% of internet users not having Google as the default search engine anymore, say. And if that were a possibility, why Google would want to risk that.
I think you have the tail and the dog confused here on the bond-buying issue. China buying fewer US bonds is something the US has been asking for for years. It's also known as allowing the yuan to appreciate against the dollar.
This would be a very good thing for the US economy in general, I suspect.
> so I guess the idea is that it vanishes if you don't > need it?
No, you're just looking at a variety of screenshots of various brainstormed proposals. Stuff along the lines of "we should think about this and see whether it makes sense", as opposed to "the next Firefox UI" like the BS summary says.
Dunno. I used an addon like this at some point, then stopped because I found better ways to deal with the problem. I don't recall what it was called at this point.
> He's talking about redenominating and then paying > off your debt through just-as-inflationary-as-you >-were-doing-before actions
You don't have to redenominate to do the "paying off debt through inflation" thing, though. Sounds like the whole redenomination part is either a red herring, or a fundamental misunderstanding on the part of the original poster there about how these things work....
> To be fair, the Marks at that time were already > worthless due to hyperinflation
Yes, that's always the case when countries redenominate their currency.
> So, it's kind of difficult to hold that as the same > example,
It's _exactly_ the same as other redenomination examples. The point is always either to put a stop to the inflation or to make the numbers look sane again after stopping inflation by other means.
Typically, the implication of adding "that isn't western europe" would be that countries in western europe don't do that.
If you insist on taking the statement at face value strictly logically, in that it's making a logical statement about non-Western-European countries whose economy is cratering only, without making any statements about any other countries, then it may be true (though I suspect it's not; it's hard to tell because then you have to rigirously define "cratered" as well as "country" and a few other things to get something actually well-defined). Or, of course, you could take the statement the way it was meant. Your choice.
Yes, precisely. Grandparent claimed that all countries not in Western Europe do currency redenomination every 60 years or so. Canada is not in Western Europe and has not had such a redenomination to my knowledge, so is a counterexample, just like Western European countries that _have_ had redenominations are counterexamples.
France last did this in 1960. That's 51 years ago.
Germany last did this in 1923. Closer to 88 years.
Italy did this when they went to the Euro, as far as I can tell.
On the other hand, Canada hasn't done this. So I'm not sure what your "Western Europe" schtick is about.
Also note that if $1 "fake" has the same buying power as $1 billion "real" (which is how currency reforms of the sort you describe usually work), then it's not like the creditors lose out. They lost out during the inflationary period, not the redenomination.
One other comment. Our currency is "strong" because we've had it a policy that it be so (at the expense of domestic employment), and because other countries have policies of propping up the value of the dollar to improve their domestic employment situation.
Basically, the US trade deficits first appeared when oil prices shot op in the 70s, though at this point that's only part of the trade deficit of course.
> The US has always consumed more than it produces
Where by "always" you mean "in the last 40 years", right?
Before 1970 the US was a net exporter.
I'm not sure what linking to a graph that only shows the last 2 years of data is supposed to say about "always".
Re:This site works best with...
on
OK Go Goes HTML5
·
· Score: 1
There are issues with both the "open" and the "standard" part of a lot of what Google calls "open standards". Creating an implementation behind closed doors and then throwing it over the wall (even with the source open), then refusing to make any changes to it when people have legitimate criticism about it not being implementable in other settings due to fundamental flaws is neither "open" nor "standard". It's a bit better than what Microsoft was doing with IE in the late 90s, but only barely.
> It's just an implementation of a standard, or a > proposed standard
Except in many cases there is no standard involved, proposed or otherwise. Again, CSS Animations, which is more or less abandonware in standards terms (none of the feedback that has been submitted about the draft has actually resulted in any changes to the draft that I can see) is on the better end of some of the things Google is implementing in Chrome.
Again, I have no problem with them writing code. I have a problem with them calling random parts of the code they write "open standards".
Re:This site works best with...
on
OK Go Goes HTML5
·
· Score: 1
A _standard_ is something that defines how something should work. A standard that cannot be understood is not useful. A standard that doesn't actually define behavior is not useful. Even if you'd like to hide your head in the sand and pretend that is is.
That doesn't mean I can't work with it via reverse engineering, but it does mean I'd appreciate you not trying to pretend that it's a "standard".
This is _exactly_ the criticism people had for OOXML, and it applies just as much to organizations you happen to like as the ones you happen to dislike.
Re:This site works best with...
on
OK Go Goes HTML5
·
· Score: 1
> Both wrappers would usually be installed at the same > time.
This happens to not be true for most OEM installs.
Chrome does indeed bundle a (somewhat hacked, actually) version of Flash with the browser, and update it using their updater.
Re:This site works best with...
on
OK Go Goes HTML5
·
· Score: 1
Actually, just black-box reverse engineering is simpler in this case than trying to sort through their code.
And to be clear, I think that the evidence for climate change is pretty solid. The evidence for anthropomorphic causes, at least in part, is fairly decent. That's based on a small dose of checking things myself and a large dose of trusting; as you point out the only way to avoid the latter is to go and specialize in the field myself.
What I'm not as convinced on, because as far as I can see the data really doesn't support them and they rely on projections which are not that rooted in reality, are some of the proposed courses of action. That includes concerns about both efficacy and necessity. And honestly, I'm more worried about the former than the latter...
> Are you really suggesting that all the climatologists > the world over are in some big conspiracy?
I'm suggesting that "scientific consensus" tends to be self-perpetuating. People who may disagree with it don't get funded to actually do the checking of whether it's right. This is not necessarily a bad thing, but it does lead to the "science advances when the old scientists die" effect.
> firefox doesn't steal my passwords and send them > off to the Mozilla foundation
I _have_ checked the code, and I'm pretty sure that you're right as far as that goes
What I can't guarantee to you is that your particular copy of Firefox isn't stealing your passwords and sending them to the author of the virus infecting your computer. In fact, I can guarantee that there are probably many people in the world right now for whom just that is happening and that there are definitely people for whom it's happened in the past.
This "objectively obtain knowledge" thing is only true to the extent that the funding agencies disburse funds that way. Scientists working for BP et al are clearly biased. Scientists working for an NSF committee (which is more or less the other alternative) are likewise biased to work on things the NSF committee will fund and to produce results that will result in more grants.
Now the NSF is a lot more interested in objectivity than BP, but that's a pretty low bar. From what I've seen, they're not really _that_ interested in objectivity in many cases...
More precisely, for any sort of non-trivial work, you go to several independent mechanics, and get estimates, then cross-check them against each other. Unless you know a mechanic that you trust personally quite outside his professional capacity. Note that this breaks down if the mechanics have a reason to collude and a way to communicate with each other; if that happens they typically overcharge you for unnecessary work. Of course they may also be honest, but the incentive structure is not conducive to it.
This last case is an interesting analogy for how funding tends to work in science nowadays, especially in the parts of it that involve calling for government action.... Except in the mechanics case a group of mechanics can't prevent another mechanic who won't inflate the estimates from seeing customers, while in science as practiced this is quite possible and actually happens.
So the problem at that point is that I actually do have more faith that car mechanics are not trying to sell me a bill of goods based on what will get them funding than I do for scientists in some fields (including, but not at all limited to climate science) or for that matter doctors (who in fact often say a medicine will help when the data shows that it doesn't help, or even hinders). Now doctors are in fact right in many other cases, but I personally know several people who would likely be dead if they had not double-checked a doctor's orders for sanity, and have at least one relative who is dead for precisely lack of such double-checking. And I'm pretty sure that the vast majority of climate research is completely aboveboard, but for any given claim I just have no way to tell. Especially because the claims that do make it to the news are precisely the ones that people have the most incentive to screw with...
So at heart, the problem is that in many of your examples above it's possible to ask for an independent second opinion (somewhat limited in the architect/engineer case because of common received wisdom), but getting an actual independent second opinion in some science fields is _hard_.
> but if most agree that my computer and OS and > browser aren't stealing my passwords, then I'll > trust them.
This is probably flat-out a bad idea.;) Unless they built your computer from scratch, they have no basis for that judgement. All they can likely offer you is that probably no one cares enough to have messed with your hardware.
Slashdot Mirror
> thanks to Macromedia/Adobe,
Uh... I suggest you do some reading about how JITs actually work and then try to understand how the differenet levels of a JIT interact with each other. Macromedia/Adobe put a bunch of work into nanojit (the one Adobe bit Tracemonkey uses), but then Mozilla did as well. And Mozilla built all the code that has to generate the LIR, which is where a lot of the optimization work happens.
Similar for Jaegermonkey: the assembler was borrowed from JSC, but the compiler code on top of it was written by Mozilla.
It's really worth learning about things before making false claims about them.....
A lot of this "feature creep" is web platform stuff. Audio+video support. More SVG support. SMIL support. CSS Transitions, CSS Animations. And so forth. If you just compare the size of the core rendering library across those releases you will see similar if not faster growth. Providing these things through "plugins" is not really desirable unless you mean something preinstalled. And that would actually _increase_ bloat due to the need to handle both SMIL being available and SMIL not being available.
Some more of the size increase, by the way, is a move toward compile options that optimize for performance at the expense of code size...
Chrome still has H.264 support last I checked. They announced that they were planning to take it out, but they never said when, and they haven't done it yet as far as I know.
Or do you have data showing otherwise?
The site was just broken in non-Chrome browsers in various ways. For example, for a while it was serving up both H.264 and Theora video, with Chrome using the H.264 and Firefox using the Theora. Except the Theora url was 404, so it didn't work in Firefox.
Similar things elsewhere, where the site takes different codepaths in different browsers and some of them are just not tested very well.
> If Google pulls out in favor of Chrome, you have to
> ask what will happen.
You can ask... or you could look up the answers.
The 2010 data is not out yet, but the 2009 numbers are at http://www.mozilla.org/foundation/documents/mf-2009-audited-financial-statement.pdf which means you don't have to worry about citing 2006 numbers.
As of 2009, Mozilla had $120 million in net assets. Expenses in 2009 were $61 million. Revenues were $104 million. They were hiring as fast as they could find good people, and earning more money than they could spend. They had 2 years worth of operating costs in the bank. All of this is public data, as it is for any other nonprofit.
So if trends continued in that revenue and expenses grew at the same percentage rate, and if you assume that Google is still 85% of their revenue stream (the data on that doesn't seem to be available), what would happen if Google pulled out is that Mozilla would have about 2.3 years to find funding sources to replace that revenue. Assuming they kept spending as much as they do now in the meantime instead of trying to stretch the money out.
On the other hand, you also have to wonder what the bottom line for Google would be from 20-30% of internet users not having Google as the default search engine anymore, say. And if that were a possibility, why Google would want to risk that.
I think you have the tail and the dog confused here on the bond-buying issue. China buying fewer US bonds is something the US has been asking for for years. It's also known as allowing the yuan to appreciate against the dollar.
This would be a very good thing for the US economy in general, I suspect.
> so I guess the idea is that it vanishes if you don't
> need it?
No, you're just looking at a variety of screenshots of various brainstormed proposals. Stuff along the lines of "we should think about this and see whether it makes sense", as opposed to "the next Firefox UI" like the BS summary says.
Fwiw, I just use Flashblock for this now. It works fine.
Dunno. I used an addon like this at some point, then stopped because I found better ways to deal with the problem. I don't recall what it was called at this point.
> He's talking about redenominating and then paying
> off your debt through just-as-inflationary-as-you
>-were-doing-before actions
You don't have to redenominate to do the "paying off debt through inflation" thing, though. Sounds like the whole redenomination part is either a red herring, or a fundamental misunderstanding on the part of the original poster there about how these things work....
Addons that do this have been available for Firefox for a long time.
> To be fair, the Marks at that time were already
> worthless due to hyperinflation
Yes, that's always the case when countries redenominate their currency.
> So, it's kind of difficult to hold that as the same
> example,
It's _exactly_ the same as other redenomination examples. The point is always either to put a stop to the inflation or to make the numbers look sane again after stopping inflation by other means.
Typically, the implication of adding "that isn't western europe" would be that countries in western europe don't do that.
If you insist on taking the statement at face value strictly logically, in that it's making a logical statement about non-Western-European countries whose economy is cratering only, without making any statements about any other countries, then it may be true (though I suspect it's not; it's hard to tell because then you have to rigirously define "cratered" as well as "country" and a few other things to get something actually well-defined). Or, of course, you could take the statement the way it was meant. Your choice.
Yes, precisely. Grandparent claimed that all countries not in Western Europe do currency redenomination every 60 years or so. Canada is not in Western Europe and has not had such a redenomination to my knowledge, so is a counterexample, just like Western European countries that _have_ had redenominations are counterexamples.
> that isn't western europe about every 60 years
France last did this in 1960. That's 51 years ago.
Germany last did this in 1923. Closer to 88 years.
Italy did this when they went to the Euro, as far as I can tell.
On the other hand, Canada hasn't done this. So I'm not sure what your "Western Europe" schtick is about.
Also note that if $1 "fake" has the same buying power as $1 billion "real" (which is how currency reforms of the sort you describe usually work), then it's not like the creditors lose out. They lost out during the inflationary period, not the redenomination.
One other comment. Our currency is "strong" because we've had it a policy that it be so (at the expense of domestic employment), and because other countries have policies of propping up the value of the dollar to improve their domestic employment situation.
I'm younger than you are, actually.
Basically, the US trade deficits first appeared when oil prices shot op in the 70s, though at this point that's only part of the trade deficit of course.
> The US has always consumed more than it produces
Where by "always" you mean "in the last 40 years", right?
Before 1970 the US was a net exporter.
I'm not sure what linking to a graph that only shows the last 2 years of data is supposed to say about "always".
There are issues with both the "open" and the "standard" part of a lot of what Google calls "open standards". Creating an implementation behind closed doors and then throwing it over the wall (even with the source open), then refusing to make any changes to it when people have legitimate criticism about it not being implementable in other settings due to fundamental flaws is neither "open" nor "standard". It's a bit better than what Microsoft was doing with IE in the late 90s, but only barely.
> Lastly, WebKit is not owned by Google.
They're working on it. As of a year ago, things looked like this graph: http://neugierig.org/software/chromium/notes/2010/02/webkit-commits.html and since then Google has been hiring more people to work on it. Give them a bit of time and they'll get there. ;)
> It's just an implementation of a standard, or a
> proposed standard
Except in many cases there is no standard involved, proposed or otherwise. Again, CSS Animations, which is more or less abandonware in standards terms (none of the feedback that has been submitted about the draft has actually resulted in any changes to the draft that I can see) is on the better end of some of the things Google is implementing in Chrome.
Again, I have no problem with them writing code. I have a problem with them calling random parts of the code they write "open standards".
A _standard_ is something that defines how something should work. A standard that cannot be understood is not useful. A standard that doesn't actually define behavior is not useful. Even if you'd like to hide your head in the sand and pretend that is is.
That doesn't mean I can't work with it via reverse engineering, but it does mean I'd appreciate you not trying to pretend that it's a "standard".
This is _exactly_ the criticism people had for OOXML, and it applies just as much to organizations you happen to like as the ones you happen to dislike.
> Both wrappers would usually be installed at the same
> time.
This happens to not be true for most OEM installs.
Chrome does indeed bundle a (somewhat hacked, actually) version of Flash with the browser, and update it using their updater.
Actually, just black-box reverse engineering is simpler in this case than trying to sort through their code.
And yes, someone did do it for me. And then I reviewed their work: https://bugzilla.mozilla.org/show_bug.cgi?id=435442
None of which makes CSS Animations an "open standard" as things stand.
And to be clear, I think that the evidence for climate change is pretty solid. The evidence for anthropomorphic causes, at least in part, is fairly decent. That's based on a small dose of checking things myself and a large dose of trusting; as you point out the only way to avoid the latter is to go and specialize in the field myself.
What I'm not as convinced on, because as far as I can see the data really doesn't support them and they rely on projections which are not that rooted in reality, are some of the proposed courses of action. That includes concerns about both efficacy and necessity. And honestly, I'm more worried about the former than the latter...
> Are you really suggesting that all the climatologists
> the world over are in some big conspiracy?
I'm suggesting that "scientific consensus" tends to be self-perpetuating. People who may disagree with it don't get funded to actually do the checking of whether it's right. This is not necessarily a bad thing, but it does lead to the "science advances when the old scientists die" effect.
> firefox doesn't steal my passwords and send them
> off to the Mozilla foundation
I _have_ checked the code, and I'm pretty sure that you're right as far as that goes
What I can't guarantee to you is that your particular copy of Firefox isn't stealing your passwords and sending them to the author of the virus infecting your computer. In fact, I can guarantee that there are probably many people in the world right now for whom just that is happening and that there are definitely people for whom it's happened in the past.
This "objectively obtain knowledge" thing is only true to the extent that the funding agencies disburse funds that way. Scientists working for BP et al are clearly biased. Scientists working for an NSF committee (which is more or less the other alternative) are likewise biased to work on things the NSF committee will fund and to produce results that will result in more grants.
Now the NSF is a lot more interested in objectivity than BP, but that's a pretty low bar. From what I've seen, they're not really _that_ interested in objectivity in many cases...
> When I need my car fixed, I go to a mechanic.
More precisely, for any sort of non-trivial work, you go to several independent mechanics, and get estimates, then cross-check them against each other. Unless you know a mechanic that you trust personally quite outside his professional capacity. Note that this breaks down if the mechanics have a reason to collude and a way to communicate with each other; if that happens they typically overcharge you for unnecessary work. Of course they may also be honest, but the incentive structure is not conducive to it.
This last case is an interesting analogy for how funding tends to work in science nowadays, especially in the parts of it that involve calling for government action.... Except in the mechanics case a group of mechanics can't prevent another mechanic who won't inflate the estimates from seeing customers, while in science as practiced this is quite possible and actually happens.
So the problem at that point is that I actually do have more faith that car mechanics are not trying to sell me a bill of goods based on what will get them funding than I do for scientists in some fields (including, but not at all limited to climate science) or for that matter doctors (who in fact often say a medicine will help when the data shows that it doesn't help, or even hinders). Now doctors are in fact right in many other cases, but I personally know several people who would likely be dead if they had not double-checked a doctor's orders for sanity, and have at least one relative who is dead for precisely lack of such double-checking. And I'm pretty sure that the vast majority of climate research is completely aboveboard, but for any given claim I just have no way to tell. Especially because the claims that do make it to the news are precisely the ones that people have the most incentive to screw with...
So at heart, the problem is that in many of your examples above it's possible to ask for an independent second opinion (somewhat limited in the architect/engineer case because of common received wisdom), but getting an actual independent second opinion in some science fields is _hard_.
> but if most agree that my computer and OS and
> browser aren't stealing my passwords, then I'll
> trust them.
This is probably flat-out a bad idea. ;) Unless they built your computer from scratch, they have no basis for that judgement. All they can likely offer you is that probably no one cares enough to have messed with your hardware.