1) Not quite. Opera actually took their no-H.264 stand first. 2) Apple took a no-Theora stand at the same time as Mozilla took a no-H.264 stand, iirc. That discussion took course over just a few days over email, so the timing of who was "first" is more related to when people checked their inbox than anything else.
Of course Apple's no-WebM stand postdates all of the above, since WebM wasn't around at the time.
The reason VP8 is so similar to H.264 is that this makes it easier to avoid patent issues. The parents that read on the H.264 standard are listed as part of the standardization process. So you start with H.264 as a base and then modify it slightly in the areas the parents cover to make sure you avoid those exact patents.
ARM and x86 in fact have different alignment requirements; things that merely cause a performance degradation due to misalignment on x86 will cause a program termination on ARM.
ARM endianness is.... variable. For most ARM chips you can in fact select at boot-time whiche endianness it should use. Some of the ARM-based handsets out there are little-endian some are big-endian; some (well, a rare few) actually expose that processor-level option and let you choose at boot.
Primitive sizes for ARM and 32-bit x86 more or less match, yes.
In fairness to Fisher, in his context (crop yields) 95% significance happened to be the point at which it became worth it to do the alternative thing being tested, from a probabilistic point of view (in that at that point the expected benefit minus the expected cost became positive).
So it's not like he made up the 95% number from nothing. But then people who don't understand statistics came along and cargo-culted it into various contexts where it may or may not make sense.
Perhaps, but what constitutes "too many"? The government is perhaps providing an amount of services that the state's residents desire and which is perfectly affordable by said residents at their income levels, except for the fact that some of their money is, dare I say, being stolen by others.
It's very easy to oversimplify the issues here; much harder to do a actual thorough analysis.
Fwiw, I do think that California has "too much" government, for my tastes. But I don't live in California either. A number of the people I know who live there feel that the amount of government they have is just fine.
No, I picked California for a good reason. It could balance its budget right now with no additional burden on its citizens if the transfer payments were stopped. I would be _incredibly_ surprised if the bailout, if any, were more than it's sent in taxes, since the latter measures in the trillions over just the last 2 decades.
I really don't follow what you don't understand about the simple equation "budget deficit == transfer payments". I realize it doesn't fit your pet theories, and I agree that California is regulation-happy to a fault, but that doesn't change the fact that if it were not for others effectively leaching off them the residents of California could easily afford those regulations.
> The Bill of Rights is one such place where the courts have pretty much ruled that if it > states the Feds can't do something, the State also cannot.
Right, but my original statement was about laws in general, not just laws that might be infringing on rights guaranteed in the Bill of Rights.
As for state A vs state B.... the bankruptcy thing is not that simple. A simple example is California, which for all its faults net pays money into the federal government (in that the amount of federal money going to California is lower than the aggregate federal taxes paid by Californians). Other states are net federal money recipients. Such transfer payments are a necessary cost of having a currency union, unless you're willing to tolerate the sorts of issues the euro is having right now. But at the same time, the net transfer payments out of California just happen to be about the size of California's budget deficit right now. So one could just as easily argue that the only reason California is "bankrupt" is that some other states are mooching off of its citizens.
In practice, the state A and state B thing is so distorted by federal action that it's hard to make any sort of useful claims about state policies based on observation of population flows.
For the rest, I agree that's what the Founders wanted; that idea died in the 1860s.
There are cases in which the U.S. Constitution limits the powers of the federal government but does not limit the power of the states in a corresponding way. So in fact, states have more latitude in their legislation than the federal government does.
Subject to the Florida constitution and in some cases the US constitution, yes.
> Why can they not just pass a law to allow random breath testing?
One could plausibly argue that such a law violates the 5th amendment of the US constitution, which is generally considered binding on state legislation.
I think we both agree that the DHS are way out of line. That doesn't mean this isn't out of line too.
Sure thing, which is why the Constitution matters here, since this sort of thing would wind up in the Supreme Court.
> A lot of the US bonds are actually inflation-linked
Sure, a lot of them are TIPS. But a lot aren't. And you don't have to get rid of all the debt to make the interest payments (which are the only real problem with carrying debt, for a government) bearable.
Note also that TIPS are linked to the _official_ rate of inflation, which happens to be determined by the US government via a formula they control. Again, no action required by congress; the whole thing is under the purview of the executive branch last I checked (the CPI, for example, is determined by the BLS, which is part of the Department of Labor, using a formula they make up, complete with hedonic adjustments, etc).
Now whether the executive branch _would_ do this is a good question; there would be some serious credibility loss involved. But they've certainly done a good bit of massaging of the CPI in the past, and the credibility loss may perceived to be a lesser evil than an outright default. Hard to tell what goes through politicians' heads sometimes.
> That's not defaulting. That's paying it back with printed bills. And the US won't do that > except as a last resort because it would destroy the US economy, while simply defaulting > wouldn't do so as immediately.
I think both would do it pretty equally, since a debt default would mean that the US suddenly couldn't secure foreign (and likely domestic) financing, and hence couldn't run deficits, at which point things would get pretty bad in the real economy, no matter how the feds try to deal with the issue.
As for paying back the debt... I don't think anyone seriously expects any country to pay back its national debt entirely; no country has ever done so and there's little incentive to do it. The best one can hope for is that interest payments on the debt outpace inflation of the currency and that any particular bond is paid off when it matures. That's good enough for anyone actually investing in bonds (as opposed to entities that hold bonds just for various other purposes, like the Social Security trust fund, the Federal Reserve, the PBoC, etc). And those properties can be satisfied as long as inflation is not too bad and the ratio of debt to gdp doesn't get too large. Of course that requires that there be gdp growth...
Fwiw, at least my high school history book (15 years ago) had a page-long section or so on Debs (not to mention a page-long section on the Espionage Act of 1917).
That was compared to about 2-3 pages total on US involvement in WWI...
> Those could be stripped tomorrow with a simple act of Congress.
The US could default on its whole debt by a simple act of Congress, as far as I can tell (though the 14th amendment is vague enough that I might be telling wrong).
Heck, it could effectively default on it by administrative action: just print money until the debt is inflated away.
I'm not going to comment on the $202 trillion figure, which does indeed involve some shenanigans, but the common figure for the US debt excludes the social security trust fund, which is about as "debt" as you can get (it holds US government bonds, like any other holder of US government debt). That's about $2.5 trillion of debt we don't like to talk about (source: http://www.ssa.gov/oact/STATS/table4a3.html ).
Back on original topic, I agree that the two-party system's implementation is broken. Just the biased rules for getting on ballots in various states seem like they should be unconstitutional (in the sense that if they're not now, we need some amendments).
Of course, being stable makes it more limited. It can't totally rejigger your browser window, say; if it could it would break when the design of that window changed, right?
Now will extension authors write to the stable API or continue to do the Windows equivalent of writing custom drivers as part of their app? Who knows.
> it's really REALLY hard to utilize any of regular expression stream tokenizers
Uh... you can't tokenize HTML (correctly at least) with regular expressions. If you're trying to, you just lose. If you're doing it for security reasons, you _really_ lose.
> so "dozens" turns to "few" under scrutiny..
Uh... "dozens" and "a few dozen" are in fact pretty much the same last I checked. "few" and "a few dozen" are not. Please do read what I wrote instead of just trolling?
> i'm sure there was at least 1 bug attributable to form control spec changes.
Absolutely.;)
> you've obviously never seen any codebase that i maintain
Clearly.
> did your mother name you "BZ"?
Nope, but it provides a convenient handle for Slashdot and irc and whatnot, being short. It's not like I make a big secret over who I am, exactly. Here; I'll even help you out:
Re:Will it support languages other than JavaScript
on
Firefox 4 Beta 8 Up
·
· Score: 1
> IIRC it's also possible to write the bulk of an extension in C++ and then hook it into the > browser chrome with Javascript, or something along those lines.
That's not quite the same thing. In that case the C++ would be running into its own world. What Firefox used to support is just putting:
in your XUL, which would run just like JS runs in chrome, with complete access to the DOM, etc. Again, no one used it.
> But this discussion isn't really about extensions, it's about web-based applications...
Right; there's just not much evidence that if any one browser (but not others) were to support this then anyone would use it. In fact, there's evidence to the contrary: even in the limited context of extensions, where compat with other browsers is not a concern, no one used it.
So why would a browser do this? It's a competitive disadvantage (since it slows down C++ and JS DOM manipulation, increases memory usage, etc) to be the first mover on this, it's a lot of work to do even a half-assed job and a _lot_ of work to do it "right", and there is no payoff for doing it...
> 1415 bugs in a mature release of a spec based media rendering engine?
That "spec based" thing...
One simple issue is that the specs are not stable in many cases (even in cases when they exist; in many cases they don't exist). There were dozens of bugs in one small component (form controls) that had to be fixed because the spec has changed yet again. Similar things for the indexeddb spec, parsing, etc, etc.
I mean.... have you tried implementing these specs? Or even finding a stable version of them?
In the case of Mozilla, all issues in the issue tracker are called "bugs". That includes feature requests, requests to give someone access to the version control system, tickets that track bumping the version number from "4.0b7" to "4.0b8", etc, etc.
So you end up with a pretty inflated bug count if you just count the "bugs".
Re:Will it support languages other than JavaScript
on
Firefox 4 Beta 8 Up
·
· Score: 3, Interesting
> Will it finally support languages other than JavaScript for client side programming?
No.
In fact, we're _removing_ such support. We supported using python for chrome (Firefox browser ui, not google's browser) programming for years, and no one used it. It's just a performance drag on the javascript and C++ side of things, so it's being removed.
The fact is, supporting multiple languages in a single runtime without leaking and without nasty performance hits on both is not really all that feasible. Given that, and the near-zero amount of actual use such functionality would get, based on our experience with chrome, it's not worth building it in....
Slashdot Mirror
> Mozilla were the first ones to take a stand,
1) Not quite. Opera actually took their no-H.264 stand first.
2) Apple took a no-Theora stand at the same time as Mozilla took a no-H.264 stand, iirc. That discussion took course over just a few days over email, so the timing of who was "first" is more related to when people checked their inbox than anything else.
Of course Apple's no-WebM stand postdates all of the above, since WebM wasn't around at the time.
The reason VP8 is so similar to H.264 is that this makes it easier to avoid patent issues. The parents that read on the H.264 standard are listed as part of the standardization process. So you start with H.264 as a base and then modify it slightly in the areas the parents cover to make sure you avoid those exact patents.
ARM and x86 in fact have different alignment requirements; things that merely cause a performance degradation due to misalignment on x86 will cause a program termination on ARM.
ARM endianness is .... variable. For most ARM chips you can in fact select at boot-time whiche endianness it should use. Some of the ARM-based handsets out there are little-endian some are big-endian; some (well, a rare few) actually expose that processor-level option and let you choose at boot.
Primitive sizes for ARM and 32-bit x86 more or less match, yes.
In fairness to Fisher, in his context (crop yields) 95% significance happened to be the point at which it became worth it to do the alternative thing being tested, from a probabilistic point of view (in that at that point the expected benefit minus the expected cost became positive).
So it's not like he made up the 95% number from nothing. But then people who don't understand statistics came along and cargo-culted it into various contexts where it may or may not make sense.
Perhaps, but what constitutes "too many"? The government is perhaps providing an amount of services that the state's residents desire and which is perfectly affordable by said residents at their income levels, except for the fact that some of their money is, dare I say, being stolen by others.
It's very easy to oversimplify the issues here; much harder to do a actual thorough analysis.
Fwiw, I do think that California has "too much" government, for my tastes. But I don't live in California either. A number of the people I know who live there feel that the amount of government they have is just fine.
Exactly.
No, I picked California for a good reason. It could balance its budget right now with no additional burden on its citizens if the transfer payments were stopped. I would be _incredibly_ surprised if the bailout, if any, were more than it's sent in taxes, since the latter measures in the trillions over just the last 2 decades.
I really don't follow what you don't understand about the simple equation "budget deficit == transfer payments". I realize it doesn't fit your pet theories, and I agree that California is regulation-happy to a fault, but that doesn't change the fact that if it were not for others effectively leaching off them the residents of California could easily afford those regulations.
> The Bill of Rights is one such place where the courts have pretty much ruled that if it
> states the Feds can't do something, the State also cannot.
Right, but my original statement was about laws in general, not just laws that might be infringing on rights guaranteed in the Bill of Rights.
As for state A vs state B.... the bankruptcy thing is not that simple. A simple example is California, which for all its faults net pays money into the federal government (in that the amount of federal money going to California is lower than the aggregate federal taxes paid by Californians). Other states are net federal money recipients. Such transfer payments are a necessary cost of having a currency union, unless you're willing to tolerate the sorts of issues the euro is having right now. But at the same time, the net transfer payments out of California just happen to be about the size of California's budget deficit right now. So one could just as easily argue that the only reason California is "bankrupt" is that some other states are mooching off of its citizens.
In practice, the state A and state B thing is so distorted by federal action that it's hard to make any sort of useful claims about state policies based on observation of population flows.
For the rest, I agree that's what the Founders wanted; that idea died in the 1860s.
> Do you think a breathalyzer counts as "witness against himself"
Yes.
> Actually, in all cases the U.S. Constitution.
There are cases in which the U.S. Constitution limits the powers of the federal government but does not limit the power of the states in a corresponding way. So in fact, states have more latitude in their legislation than the federal government does.
> They get to pass their own laws?
Subject to the Florida constitution and in some cases the US constitution, yes.
> Why can they not just pass a law to allow random breath testing?
One could plausibly argue that such a law violates the 5th amendment of the US constitution, which is generally considered binding on state legislation.
I think we both agree that the DHS are way out of line. That doesn't mean this isn't out of line too.
> although I foresee massive class action suits
Sure thing, which is why the Constitution matters here, since this sort of thing would wind up in the Supreme Court.
> A lot of the US bonds are actually inflation-linked
Sure, a lot of them are TIPS. But a lot aren't. And you don't have to get rid of all the debt to make the interest payments (which are the only real problem with carrying debt, for a government) bearable.
Note also that TIPS are linked to the _official_ rate of inflation, which happens to be determined by the US government via a formula they control. Again, no action required by congress; the whole thing is under the purview of the executive branch last I checked (the CPI, for example, is determined by the BLS, which is part of the Department of Labor, using a formula they make up, complete with hedonic adjustments, etc).
Now whether the executive branch _would_ do this is a good question; there would be some serious credibility loss involved. But they've certainly done a good bit of massaging of the CPI in the past, and the credibility loss may perceived to be a lesser evil than an outright default. Hard to tell what goes through politicians' heads sometimes.
> That's not defaulting. That's paying it back with printed bills. And the US won't do that
> except as a last resort because it would destroy the US economy, while simply defaulting
> wouldn't do so as immediately.
I think both would do it pretty equally, since a debt default would mean that the US suddenly couldn't secure foreign (and likely domestic) financing, and hence couldn't run deficits, at which point things would get pretty bad in the real economy, no matter how the feds try to deal with the issue.
As for paying back the debt... I don't think anyone seriously expects any country to pay back its national debt entirely; no country has ever done so and there's little incentive to do it. The best one can hope for is that interest payments on the debt outpace inflation of the currency and that any particular bond is paid off when it matures. That's good enough for anyone actually investing in bonds (as opposed to entities that hold bonds just for various other purposes, like the Social Security trust fund, the Federal Reserve, the PBoC, etc). And those properties can be satisfied as long as inflation is not too bad and the ratio of debt to gdp doesn't get too large. Of course that requires that there be gdp growth...
Fwiw, at least my high school history book (15 years ago) had a page-long section or so on Debs (not to mention a page-long section on the Espionage Act of 1917).
That was compared to about 2-3 pages total on US involvement in WWI...
> Those could be stripped tomorrow with a simple act of Congress.
The US could default on its whole debt by a simple act of Congress, as far as I can tell (though the 14th amendment is vague enough that I might be telling wrong).
Heck, it could effectively default on it by administrative action: just print money until the debt is inflated away.
I'm not going to comment on the $202 trillion figure, which does indeed involve some shenanigans, but the common figure for the US debt excludes the social security trust fund, which is about as "debt" as you can get (it holds US government bonds, like any other holder of US government debt). That's about $2.5 trillion of debt we don't like to talk about (source: http://www.ssa.gov/oact/STATS/table4a3.html ).
Back on original topic, I agree that the two-party system's implementation is broken. Just the biased rules for getting on ballots in various states seem like they should be unconstitutional (in the sense that if they're not now, we need some amendments).
You may be interested in https://jetpack.mozillalabs.com/ which is precisely such a stable API.
Of course, being stable makes it more limited. It can't totally rejigger your browser window, say; if it could it would break when the design of that window changed, right?
Now will extension authors write to the stable API or continue to do the Windows equivalent of writing custom drivers as part of their app? Who knows.
I did not in fact notice that, no. And I gave up, indeed. ;) Thanks for the good advice!
> it's really REALLY hard to utilize any of regular expression stream tokenizers
Uh... you can't tokenize HTML (correctly at least) with regular expressions. If you're trying to, you just lose. If you're doing it for security reasons, you _really_ lose.
> so "dozens" turns to "few" under scrutiny..
Uh... "dozens" and "a few dozen" are in fact pretty much the same last I checked. "few" and "a few dozen" are not. Please do read what I wrote instead of just trolling?
> i'm sure there was at least 1 bug attributable to form control spec changes.
Absolutely. ;)
> you've obviously never seen any codebase that i maintain
Clearly.
> did your mother name you "BZ"?
Nope, but it provides a convenient handle for Slashdot and irc and whatnot, being short. It's not like I make a big secret over who I am, exactly. Here; I'll even help you out:
http://letmegooglethatforyou.com/?q=bz+mozilla
> i most certainly have implemented them,
You've written an HTML parser? A JavaScript interpreter? A JPEG decoder?
Do you and I just disagree on the meaning of the word "implement" here?
> there aren't 1415 form controls... there aren't even 1415 attributes.
Did you miss the part where I said that form controls are a _small_ module that contributed a few dozen make-work bugs to the total?
> the firefox codebase is quite obviously littered with logical failures.
Yep. It sure is. As is every single large codebase of any sort I've looked at.
You're quite welcome!
Er... I keep forgetting that Slashdot's "plain old text" isn't. What you could put in was:
<script type="application/python" src="my-python.py"></script>
> IIRC it's also possible to write the bulk of an extension in C++ and then hook it into the > browser chrome with Javascript, or something along those lines.
That's not quite the same thing. In that case the C++ would be running into its own world. What Firefox used to support is just putting:
in your XUL, which would run just like JS runs in chrome, with complete access to the DOM, etc. Again, no one used it.
> But this discussion isn't really about extensions, it's about web-based applications...
Right; there's just not much evidence that if any one browser (but not others) were to support this then anyone would use it. In fact, there's evidence to the contrary: even in the limited context of extensions, where compat with other browsers is not a concern, no one used it.
So why would a browser do this? It's a competitive disadvantage (since it slows down C++ and JS DOM manipulation, increases memory usage, etc) to be the first mover on this, it's a lot of work to do even a half-assed job and a _lot_ of work to do it "right", and there is no payoff for doing it...
> 1415 bugs in a mature release of a spec based media rendering engine?
That "spec based" thing...
One simple issue is that the specs are not stable in many cases (even in cases when they exist; in many cases they don't exist). There were dozens of bugs in one small component (form controls) that had to be fixed because the spec has changed yet again. Similar things for the indexeddb spec, parsing, etc, etc.
I mean.... have you tried implementing these specs? Or even finding a stable version of them?
In the case of Mozilla, all issues in the issue tracker are called "bugs". That includes feature requests, requests to give someone access to the version control system, tickets that track bumping the version number from "4.0b7" to "4.0b8", etc, etc.
So you end up with a pretty inflated bug count if you just count the "bugs".
> Will it finally support languages other than JavaScript for client side programming?
No.
In fact, we're _removing_ such support. We supported using python for chrome (Firefox browser ui, not google's browser) programming for years, and no one used it. It's just a performance drag on the javascript and C++ side of things, so it's being removed.
The fact is, supporting multiple languages in a single runtime without leaking and without nasty performance hits on both is not really all that feasible. Given that, and the near-zero amount of actual use such functionality would get, based on our experience with chrome, it's not worth building it in....