Having a college _degree_ (not to be confused with _education_) puts somewhat of a floor on employee performance: it indicates that the employee was able to sit through a minimal required number of classes.
So if you have sufficiently many applicants and just need to cut down the pool quickly without worrying too much about false negatives, testing for a college degree is a simple way to do it.
And the system is self-perpetuating: the more people are enabled to go to college by student loans, whether they learn anything there or not, the more likely it becomes that someone without a college degree simply couldn't handle dealing with a few years of make-work for some reason. And if they can't handle that, they're not a good fit for most office jobs.
So what has happened is that college is becoming more and more of a proxy for the exact skills a high school diploma was supposed to indicate (ability to sit there and follow directions), while high schools become harder to fail out of and post-secondary degrees start to be the real marker of "has an education"... This is really not good for anyone other than people who make money running colleges.
Your IBM example is different from the trend of requiring a college diploma for secretarial work in that it actually involves positions where the sorts of skills college might teach (critical reasoning, understanding of societal context, etc) are useful.
For example, America produces wheat at a much lower cost than China. But China has explicit policies in place to raise the domestic price of wheat to the point where consumers don't buy more of it than they can produce domestically. This sort of sucks for Chinese consumers, obviously, but who cares if people are getting enough food....
More to the point of your question, what America has that China needs right this second is consumers. A trade war does mean higher prices on manufactures goods for US consumers; it also means higher unemployment in China. And higher unemployment and the ensuing political instability is something the Chinese government desperately doesn't want. This is why they've been doing everything they can to export their unemployment to the rest of the world (which is what their currency operations and general economic policy is all about). This does involve repressing domestic demand, unfortunately, which leaves them even more dependent on exports and foreign demand to keep people employed...
So far the US has been allowing higher unemployment here in exchange for cheaper products. This works so long as the unemployment rate doesn't get too high so people vote more for cheap stuff than for more jobs. Whether we've gotten to "too high" yet is unclear.
The point is that if your librarian has that data then your government can subpoena it. And there is a long history of governments using such data in pernicious ways (e.g. "blacklist anyone who has checked out the Communist Manifesto from employment").
For the rest, I suggest googling the phrase "I've got nothing to hide" and doing some reading....
Ah, maybe this is a matter of different mores in different countries? In the US, at least, librarians take not keeping this sort of history around very seriously in general.
As for not giving up any rights.... "I've got nothing to hide" at its finest.
This may come as a surprise to you, but the vast majority of users under the age of 6 and over the age of 45 or so cannot in fact use a mouse at all well, in my experience.
And for the over-45 crowd it's not even an issue of never having learned it as much as an issue of getting older and fine motor control starting to go.
Er, I just realized there's some confusion here (on my end) between Pepper and Parrot. So ignore what I said here while I go and read up on Parrot details.;)
1) Many project can't use GPL'd code. Opera comes to mind, say.
2) Being GPL'd in this case just means you can fork it, not that you can affect its development or prevent changes that break you. Google has complete power to evolve Parrot in any way it wants, and if it happens to break other browsers that happen to be using Parrot or breaks websites they have no choice but to deal. It's not quite as bad as being locked into something you _can't_ fork, of course. But don't pretend like it's "not much of a lock in".
> 1) You loose a lot in translating code. No > optimizations.
This is true of any intermediate representation not designed with a particular language in mind, to some extent. The emscripten compiler does in fact do a fair amount of optimization, and it's operating on LLVM bitcode that's already had some optimization passes applied.
> 2) You can not use libraries native to the language.
In the specific case of emscripten, again, you just compile the libraries,
> 3) You end up having to write/correct the > translators with "if (typeof > myFunction=='undefined') {"
Why correct? You do have to write the translator; you do it once.
> ActiveX is a response to NPAPI with gapping holes > in system security.
Yes, and an attempt at single-vendor lockin of the web. Pepper is missing the security holes, with any luck, but has the lockin thing in spades.
If all the UAs could agree on a single VM and if that VM were not controlled by any one company, then hitching the web to that VM might be OK. But unless you want to give a single company control over development of the web, this is not that likely to happen in the near future...
If you want to embed multiple language VMs in the browser, you immediately start having issues with cross-VM reference cycles causing leaks. The infrastructure needed for breaking those is.... nontrivial. You also get very complicated interactions performance characteristics as the VMs interact. Note that historically browsers have had issues just solving these problems for the JS VM and C++ DOM, without adding more VMs into the mix.
A better bet may be having a single VM that multiple languages can compile to (a la JVM, which all sorts of things other than Java can target). But at that point, perhaps that VM should simply be the JS VM. For example, you can already compile C code to JS. See http://syntensity.com/static/jsconf_eu_Emscripten_lo.pdf for the recent state of that. Now you do get a slowdown over, say, gcc -O3; the slides list this slowdown as somewhere between 2x and 8x depending on the workload. But JS VMs are pretty continuously improving, and it's not clear how fast running C code on the JVM, say, would be... Would be interesting to have those numbers.
Because a lot of people want to actually _try_ a device before buying it. Little things like how the keyboard feels, how the screen looks, that sort of thing.
I dearly wish I could just go to a physical store and try a bunch of different e-readers instead of spending tens of hours poring over online reviews trying to tease the bits of information I care about out of them.
Chances are, it does. The state of hardware acceleration of graphics on Windows is way better in all browsers right now than on non-Windows platforms...
We obviously do quality control, but not in the full set of circumstances (malware running on the same machine, various extensions, etc) that our users run in, because there are unfortunately just too many possible configurations.
We also track crashes via the crash reporter and fix them. This works great for crashes lots of people hit, but not so well for a crash only a few users hit.
Would you mind going to about:crashes in your Firefox and letting me know what those links are? That would hopefully give me enough information to figure out why you in particular are crashing. If you're willing to do that, I would really appreciate it.
For the rest, there's a bunch of memory improvements in Firefox 8 even compared to Firefox 7, tons of bugfixes, lots of user expeience and web standards improvements, and so forth. I can't help it that the Slashdot summary writers just listed 3 items out of a much longer list in the original blog post, which is itself not exhausrive.
I have no idea which problem you're talking about, so I don't know whether it's fixed.... I personally don't work on the guts of text layout, so you'll have to help me out here by pointing to a bug report.
Is there a bug on file on this? We try really hard to not break websites, obviously; if we broke something without realizing it we would _really_ like to know.
Instead they've been bundling Chrome with all sorts of other software (e.g. installing or updating Flash on windows will helpfully install Chrome for you unless you explicitly opt out).
No one designing a website now seriously designs IE-only. That's years in the past.
However, plenty of people design WebKit-only instead of standards, especially for "mobile-targeted" sites (where non-WebKit UAs face serious compatibility issues to the point that Microsoft and Mozilla have both considered adding support for some -webkit CSS properties on mobile just to get sites to stop breaking).
Note only that, but both Apple and Google have been pushing various non-standard extensions at web developers, and web developers are listening.
So in fact, Chrome or Safari taking market share from other browsers _reduces_ the pressure to design to standards at this point... This wasn't the case 2-3 years ago, but it's the case now.
Slashdot Mirror
NSS is maintained by its module owners, who happen to work at Google at the moment. At least one of them is on the Chrome team.
Mozilla hosts the bug tracker and code repository.
So NSS is maintained by Mozilla about the same way as the Linux kernel is maintained by kernel.org.
Having a college _degree_ (not to be confused with _education_) puts somewhat of a floor on employee performance: it indicates that the employee was able to sit through a minimal required number of classes.
So if you have sufficiently many applicants and just need to cut down the pool quickly without worrying too much about false negatives, testing for a college degree is a simple way to do it.
And the system is self-perpetuating: the more people are enabled to go to college by student loans, whether they learn anything there or not, the more likely it becomes that someone without a college degree simply couldn't handle dealing with a few years of make-work for some reason. And if they can't handle that, they're not a good fit for most office jobs.
So what has happened is that college is becoming more and more of a proxy for the exact skills a high school diploma was supposed to indicate (ability to sit there and follow directions), while high schools become harder to fail out of and post-secondary degrees start to be the real marker of "has an education"... This is really not good for anyone other than people who make money running colleges.
Your IBM example is different from the trend of requiring a college diploma for secretarial work in that it actually involves positions where the sorts of skills college might teach (critical reasoning, understanding of societal context, etc) are useful.
> what does America produce that China needs?
Needs or wants?
For example, America produces wheat at a much lower cost than China. But China has explicit policies in place to raise the domestic price of wheat to the point where consumers don't buy more of it than they can produce domestically. This sort of sucks for Chinese consumers, obviously, but who cares if people are getting enough food....
More to the point of your question, what America has that China needs right this second is consumers. A trade war does mean higher prices on manufactures goods for US consumers; it also means higher unemployment in China. And higher unemployment and the ensuing political instability is something the Chinese government desperately doesn't want. This is why they've been doing everything they can to export their unemployment to the rest of the world (which is what their currency operations and general economic policy is all about). This does involve repressing domestic demand, unfortunately, which leaves them even more dependent on exports and foreign demand to keep people employed...
So far the US has been allowing higher unemployment here in exchange for cheaper products. This works so long as the unemployment rate doesn't get too high so people vote more for cheap stuff than for more jobs. Whether we've gotten to "too high" yet is unclear.
This, like many W3C specs from the early to mid 2000s, has nothing to do with the web and everything to do with enterprise backend stuff.
The point is that if your librarian has that data then your government can subpoena it. And there is a long history of governments using such data in pernicious ways (e.g. "blacklist anyone who has checked out the Communist Manifesto from employment").
For the rest, I suggest googling the phrase "I've got nothing to hide" and doing some reading....
Ah, maybe this is a matter of different mores in different countries? In the US, at least, librarians take not keeping this sort of history around very seriously in general.
As for not giving up any rights.... "I've got nothing to hide" at its finest.
That sort of thing absolutely does not happen in most library systems. I'm sorry to hear it's happening in yours... Where is that, if I might ask?
This may come as a surprise to you, but the vast majority of users under the age of 6 and over the age of 45 or so cannot in fact use a mouse at all well, in my experience.
And for the over-45 crowd it's not even an issue of never having learned it as much as an issue of getting older and fine motor control starting to go.
Er, I just realized there's some confusion here (on my end) between Pepper and Parrot. So ignore what I said here while I go and read up on Parrot details. ;)
1) Many project can't use GPL'd code. Opera comes to mind, say.
2) Being GPL'd in this case just means you can fork it, not that you can affect its development or prevent changes that break you. Google has complete power to evolve Parrot in any way it wants, and if it happens to break other browsers that happen to be using Parrot or breaks websites they have no choice but to deal. It's not quite as bad as being locked into something you _can't_ fork, of course. But don't pretend like it's "not much of a lock in".
> 1) You loose a lot in translating code. No
> optimizations.
This is true of any intermediate representation not designed with a particular language in mind, to some extent. The emscripten compiler does in fact do a fair amount of optimization, and it's operating on LLVM bitcode that's already had some optimization passes applied.
> 2) You can not use libraries native to the language.
In the specific case of emscripten, again, you just compile the libraries,
> 3) You end up having to write/correct the
> translators with "if (typeof
> myFunction=='undefined') {"
Why correct? You do have to write the translator; you do it once.
> ActiveX is a response to NPAPI with gapping holes
> in system security.
Yes, and an attempt at single-vendor lockin of the web. Pepper is missing the security holes, with any luck, but has the lockin thing in spades.
If all the UAs could agree on a single VM and if that VM were not controlled by any one company, then hitching the web to that VM might be OK. But unless you want to give a single company control over development of the web, this is not that likely to happen in the near future...
Yes, but why is Parrot a better approach than just compiling to JavaScript and using the existing JS VM?
(And note that Parrot is much like ActiveX: an unspecified single-vendor kinda thing.)
If you want to embed multiple language VMs in the browser, you immediately start having issues with cross-VM reference cycles causing leaks. The infrastructure needed for breaking those is .... nontrivial. You also get very complicated interactions performance characteristics as the VMs interact. Note that historically browsers have had issues just solving these problems for the JS VM and C++ DOM, without adding more VMs into the mix.
A better bet may be having a single VM that multiple languages can compile to (a la JVM, which all sorts of things other than Java can target). But at that point, perhaps that VM should simply be the JS VM. For example, you can already compile C code to JS. See http://syntensity.com/static/jsconf_eu_Emscripten_lo.pdf for the recent state of that. Now you do get a slowdown over, say, gcc -O3; the slides list this slowdown as somewhere between 2x and 8x depending on the workload. But JS VMs are pretty continuously improving, and it's not clear how fast running C code on the JVM, say, would be... Would be interesting to have those numbers.
All HTTP socket IO is most certainly on a separate thread in necko.
Just to make sure, you see the problem in safe mode (or with all extensions disabled) as well, right?
I'd love to fix this. Reproducing would be a good start. Please point me to a url that shows the problem?
Because a lot of people want to actually _try_ a device before buying it. Little things like how the keyboard feels, how the screen looks, that sort of thing.
I dearly wish I could just go to a physical store and try a bunch of different e-readers instead of spending tens of hours poring over online reviews trying to tease the bits of information I care about out of them.
Chances are, it does. The state of hardware acceleration of graphics on Windows is way better in all browsers right now than on non-Windows platforms...
We obviously do quality control, but not in the full set of circumstances (malware running on the same machine, various extensions, etc) that our users run in, because there are unfortunately just too many possible configurations.
We also track crashes via the crash reporter and fix them. This works great for crashes lots of people hit, but not so well for a crash only a few users hit.
Would you mind going to about:crashes in your Firefox and letting me know what those links are? That would hopefully give me enough information to figure out why you in particular are crashing. If you're willing to do that, I would really appreciate it.
For the rest, there's a bunch of memory improvements in Firefox 8 even compared to Firefox 7, tons of bugfixes, lots of user expeience and web standards improvements, and so forth. I can't help it that the Slashdot summary writers just listed 3 items out of a much longer list in the original blog post, which is itself not exhausrive.
Just to be clear, complaining is great if it's actionable.
"A site got broken, but I won't tell you which site or what broke" is really hard to take action on...
Not that obvious, apparently! I do watch incoming core bug reports and haven't seen one about this.
Thanks for the mozillazine link; I filed https://bugzilla.mozilla.org/show_bug.cgi?id=690806 on this.
And you're very much welcome!
I have no idea which problem you're talking about, so I don't know whether it's fixed.... I personally don't work on the guts of text layout, so you'll have to help me out here by pointing to a bug report.
Is there a bug on file on this? We try really hard to not break websites, obviously; if we broke something without realizing it we would _really_ like to know.
They know people don't use CDs anymore. ;)
Instead they've been bundling Chrome with all sorts of other software (e.g. installing or updating Flash on windows will helpfully install Chrome for you unless you explicitly opt out).
The problem is the number of people creating content that's WebKit-specific. It's not a problem with WebKit per se, just with market positioning.
No one designing a website now seriously designs IE-only. That's years in the past.
However, plenty of people design WebKit-only instead of standards, especially for "mobile-targeted" sites (where non-WebKit UAs face serious compatibility issues to the point that Microsoft and Mozilla have both considered adding support for some -webkit CSS properties on mobile just to get sites to stop breaking).
Note only that, but both Apple and Google have been pushing various non-standard extensions at web developers, and web developers are listening.
So in fact, Chrome or Safari taking market share from other browsers _reduces_ the pressure to design to standards at this point... This wasn't the case 2-3 years ago, but it's the case now.