Erlang is an Ericsson product, and they've released or assisted on a number of other products. Sony - well, they're famous for closed products, rootkits and the walkman. I'm not overly convinced I trust those guys to honour prior license agreements (there are plenty of products that were GPL that have been made proprietary with the open source variant deleted from the catalogue completely). I am very concerned.
Not only that, the US has been demanding the payback of loans made during the wars (loans with interest, I might add) that would never have been required if the US hadn't initially sided with Germany and confiscated Allied property in the US. Hell, had the confiscations not happened, US involvement might never have been needed - Europe might have been able to afford an armed forces capable of resisting the Germans.
Exactly. Reference texts exist for a reason. So long as you know the basics, the reference texts are all you ever need to be able to do the specifics.
And thanks for your post. Sure I have the skills, but I'm a freak in that respect. (I learned something of everything, which is why I had 10 BGP4+ IPv6 tunnels running in 1996 and was hosting 1/3rd of the IPv6 transatlantic backbone for a while.) But there's no need for me to be a freak in that respect. Anyone with a BSD or Linux box can use Quagga or Zebra to learn the core elements of IOS. Anyone with a BSD or Linux box, Webmin and a text editor (emacs works fine) can learn about how servers are configured and what configuration option in what configuration file does what. Or you can read the docs, but experimenting is usually a better way to learn.
You are correct, which is why there are plugins for Tcl, etc. Selenium is technically a client-side scripting language. Flash is blob, so not a technically a script, but it is still run client-side. Dunno if the attempt to run Python client-side ever got finished. There used to be a Mozilla plugin that supported the X11 protocol. The SCRIPT tag has a TYPE attribute which gives the media type. There is nothing in the spec that says this must be Javascript, it can be anything at all - though it helps if the browser supports it. The spec, then defines any language that can be described with the media type as being a valid language.
If people want features X, Y or Z not in the official specification for language L, use a different media type. It's not like the code will run anybloodyplace else anyway.
Show me 3D that requires that the page define whether it is to be rendered in software or hardware and I'll show you a specification that should be burned at the stake along with idiots like yourself.
I don't give a frag whether a given piece of 3D is rendered using SVG, VRML, GDML, OpenGL, DX11, PHIGS, Renderman, Maya, Rhino, Blender, a GPU, one of those insanely high-end nVidia modules that uses more power than every other computer in the house combined, that Chinese supercomputer built out of GPUs, or a cheese sandwich, so long as it is rendered correctly. The software should detect what options exist, use the best one (according to the built-in algorithm) and allow the user to switch to another.
There's bugger all HTML5 has to do with that process and only a moronic imbecile could think otherwise.
Rule 101, taught to EVERY BLOODY CS STUDENT ALIVE, is NEVER EVER EVER test your own code. You WILL miss things. ALWAYS have the code tested by someone else or - in the case of Extreme Programming methods where you write the test harnesses in advance - by something else. But NEVER test it yourself.
I read the judgement when it came out. I have worked with companies carrying out the court instructions. What's your excuse?
Yes, you can replace the entire kernel. Not just with another Linux kernel - anything that supports the Linux ABI will work, so you could replace the Linux kernel with Lynx if you wanted. FreeBSD should also work. There are probably others.
It goes beyond Unix. Intel defined the Intel Binary Compatibility Standard to facilitate ANY OS whatsoever running ANY software from ANY OTHER OS, provided both were written to the spec. Thus, there's nothing to prevent you from running a Solaris application dynamically linked to a Unixware library all under the Linux kernel. Yes, even Intel believed that vendor interoperability was important.
Internal to the kernel, it makes bugger all difference whether you're running the graphics through Framebuffer, KGI, a proprietary driver that can bind to X or a graphics-to-ASCII-art converter (yes, they exist). Everything still works exactly the same, except that Doom looks a bit odd. It still works, though.
Actually, yes. A lot of people "bother with that". In the scientific computing world, it means you don't have to care *whose* BLAS or LAPACK library you use. This is great. You can design using standalone libraries designed specifically to assist debugging and run against parallel libraries optimized for sheer speed - even when they're written by completely different groups.
In the GUI world, do you really care if you're using Motif or Lestif? Or whether that's really SGI's OpenGL or actually the Mesa 3D library? Can you name a single X11 program that breaks when using a custom implementation of X11 rather than the reference version? After all, it links to all kinds of libraries!
So, yes, every company -- barring Microsoft -- already gives the option to substitute ALL shared libraries. Microsoft is about the sole exception and it is a stupid one.
Does anyone care? Well, define "care". They care that their programs "just work" and that they don't need a billion essentially identical libraries to get them to do so. They care that they can tune and tweak. They care that updating external components or replacing them with something functionally the same will not break anything.
They don't care which specific library is installed, unless there's one optimized the way they want, precisely because things "just work". There's about a dozen standard C libraries - not because anyone seriously thinks people want to get the complete set, but because that lets users tailor their system to their needs, rather than tailoring their needs to what some vendor has decreed.
THAT is why they would bother.
Who cares if the rendering engine is shared? Well, if it's the rendering engine that is shared, take it out of IE and make it an independent shared component. Then people can uninstall IE if they want. Tying the rendering engine into IE and thus preventing people from uninstalling IE is not a sound software design, it is merely an abuse of a monopoly in an effort to gain another monopoly. Which is a criminal act.
People WOULD care if they could replace the rendering engine. There are other HTML5 rendering engines out there and being able to replace one with another would allow me to use whatever look-and-feel I liked without having to replace the GUI entirely. I should not have to replace Explorer, but I point out that you CAN. That people HAVE rewritten Afterstep as an Explorer replacement. That project was damn popular. Why? Because people actually DO like having a say over the L&F.
JS isn't a W3C standard. It can be obtained as a standalone engine. Provided Google's Chrome can support the loading of that engine, it makes not one whit of difference whether Google adds other engines to Chrome or remove their own JS engine. The only requirement for JS compatibility is that SOME JS engine be loadable at SOME point. It doesn't have to be built-in.
In fact, it's probably better if it isn't. Lightweight tools are generally superior tools. Having JS as a plugin would ensure that you could use any JS out there -- if you wanted to -- but that if you didn't want to run JS at all, you had zero overheads. THAT is Superior Design. It is ALSO Classic Unix Design Philosophy.
Actually, the basis of the Windows 95/98 lawsuit and the later IE bundling under XP lawsuit was that libraries ARE crafted to include totally irrelevant code. Indeed, it was Microsoft's position in the lawsuit that Felton's hack could not possibly work BECAUSE they had included such code. (Felton's hack worked because it left the extraneous code intact and in place.)
Nonetheless, even Microsoft disagrees with you. Under oath.
First off, not an excuse. Selenium means that testing one browser or a hundred different brands takes the same time and the same level of complexity.
Second off, no competent vendor has extensions to HTML, CSS or JS. Competent vendors do EVERYTHING in the standards, which are quite powerful enough. It is a mark of incompetency that Microsoft not only does NOT implement the standards, they fill the gaps with proprietary crap.
Third, developers should never test their own code. That is a sign of an untrained and moronic developer.
Fourth, if you are required by a court to SPECIFY all of these APIs and ABIs, then you are violating the law to not specify them. That is absolutely final.
Fifth, if you are required by a court to ALLOW a drop-in replacement for any given set of APIs and ABIs, then you are violating the law to not permit such a replacement. That too is final.
Included is not the same as required. Even if you installed Ubuntu (which has Firefox), you aren't required to install it. It's optional. Thus it isn't tied in. Further, even if you install it, you can later uninstall it when you discover Chrome does most of the stuff Firefox does better.
Firefox isn't in the OS in Linux (or any other OS). Firefox isn't an OS program. It is a user application. There is a HUGE difference. It is hard to describe all the ways it is different without causing the Slashdot machines to run out of space for database files, there are that many.
Removing Firefox from an installation won't cause X11 or the Linux kernel to destabilize. Removing IE will cause Windows 8's kernel to break at the lowest level.
No, that's not because Windows 8 does low-level graphics stuff. Install KGI (the Kernel Graphics Interface) or use Framebuffers extensively. Run X through them. Run KDE through them. Run a browser - say, Opera, Chrome, or even Firefox, through KDE. Now trace through the code and show me where KGI or FB code directly invokes a Firefox library. I dare you to try. Go on! Should be easy.
Yes, users do NOT appreciate any web browser in the OS. They want browsers in Application Space, where a bug won't cause the machine to crash and where switching to something else is easy rather than a 8-month hack with VTune, process-grabbing debuggers and a decompiler.
I don't. GNOME should permit any library that is API and ABI compatible and should not depend on specific implementations of anything. Used to be that GNOME did NOT depend on specific implementations, that you could choose between anything that provided identical functionality. Technically, since the source is out there, that's still the case but it should never have been the case that they restricted themselves to one solution alone.
Nonetheless, GNOME is not an Operating System, the Linux kernel won't break if you don't install GNOME (or indeed X). Whereas, Windows' kernel WILL break if IE is missing. Thus, your comparison is flawed. Probably knowingly.
A brain surgeon (with computing experience) would point out that standalone rendering engines have existed for years - and have existed for longer than any of the current browsers out there. Having the HTML5 rendering engine in a standalone DLL that could be replaced by anyone else's HTML5 rendering engine would NOT be an OS tie-in.
Since HTML5 rendering engines do NOT need a browser (since they can be standalone), a browser is NOT needed for this.
However, if you absolutely insist that a browser provide the library, a published specification (as per the requirements of the anti-trust suit, I might add) of exactly what functions are needed in the library, what name they must have and what ABI they must use, ANY web browser could be used. This is lawful under the requirements. A tie-in is NOT.
This is a flagrant violation of the law, which Microsoft will get away with because nobody dares start controversial lawsuits in an election year. Nonetheless, it IS illegal and it IS unnecessarily illegal. It is done this way for one reason and one alone - to kill competition. That is ALL it is being done for. It isn't for convenience and it isn't for the HTML5.
The OS is a kernel plus core system libraries. It has ZERO relationship to how the output is displayed. Which is why I not only should be able to run KDE and AfterStep under Windows, I can.
What I CANNOT do is run Internet Explorer on Linux. So what if it's compiled for another OS and I don't have the source? I don't have the source for Solaris-x86 Oracle but I CAN run that under Linux (different OS and no source) just fine. Have been able to for years.
Yes, when you open a file panel or a network browser under Windows, you are using IE. The desktop is IE. The control panel is IE. Friggin' everything is IE! Even if you install another browser, you CANNOT tell those components to use it. So, yes, if you use Windows, you MUST use IE. You have no choice. And must you use Windows? Well, yes. Many web applications aren't written to international standards, they're written to Microsoft-proprietary functionality within IE. This WILL worsen, with this news about IE and Windows 8, just as it worsened considerably after Microsoft violated the Windows 95 injunction by releasing the bundled IE as Windows 98.
The competition is hurting something chronic. IE has rising usage figures. Firefox is starting to slide. Opera is sliding badly. Chrome may run foul of the Apple vs Google battle-to-the-death. (And one of them WILL die in it, if they don't back off.) Linux has never been fairly or reasonably offered as a desktop choice by anyone other than the OLPC group - and even they are now getting into bed with Microsoft.
Microsoft is a devout monopolist and it WILL kill anything that threatens that monopoly, no matter how savage or ugly they have to get to do so.
There is nothing in Linux which requires Firefox. Firefox is pre-installed, but only on specific distros. Other distros include other browsers, or no browser at all. (You don't need one - wget is perfectly good.)
This is different than with IE and Windows. If you remove IE, components totally unrelated to web browsing or the Internet WILL fail, because the libraries are crafted to include totally irrelevant code that is critical for other components. Because Microsoft do not publish the specs for these libraries, crafting replacements that ONLY have the bits needed for the rest of the system to function is almost impossible. Not completely impossible, just very very very hard.
There simply isn't any comparison between willful sabotage of the user and a simple pre-install, even if your claim that Firefox was pre-installed with Linux was correct.
Indeed. Didn't the US do this for a longish while, signing up to the International treaties after they'd got the good stuff?
I'm not saying that China is "faultless" or the US is "all bad", but let's face facts - America would not have made the progress it has if it had respected European patent laws and European property rights. If it wants to claim China is in the wrong, then I have nothing against that provided it is NOT for the purpose of maintaining a hegemony obtained solely through the same practices. If China is guilty, then American corporations and the American government owe Europe a percentage of the profits secured through IP theft.
Sure, that might push the US into recession. Isolating China and closing down all counterfeit goods plus genuine goods based on stolen IP would not merely put China into recession, it would bankrupt it. If you're willing to do the latter, you should be man enough to accept the former.
The good news is that 100+ years of compound interest for some of the products and 60+ years of accumulated value in the case of property illegally confiscated from British and other Allied nations during WW2 should cover the combined debts of Italy, Greece, Spain and Ireland, and leave enough left over for the heads of State to put in advance orders for GTA 5.
I agree with you. The American model produces people that are too inflexible and too inexperienced. The problem with the Japanese model is that it is too closed-loop. In just the same way that you clean fishtanks by keeping some things fresh and the fish original, the Japanese need to split off those elements that have become stale and/or toxic without replacing those with experience. ie: a partially-open look. That would breed innovation without also breeding the toxin of ignorance.
I've admined 386BSD, FreeBSD, NetBSD, OpenBSD, Linux (0.1 - 3.1), SunOS, Solaris, IRIX, HPUX, OSF/1, VxWorks, Windows 200x Server, Windows XP, Windows 7, Plan 9, Inferno, the PDP-11 and OS/X. Not sure I'd call RTAI or Xenomai distinct OS'. Montavista certainly wasn't. Dealt with all three.
This includes direct kernel work (hacking patches together to form the Functionally Overloaded Linux Kernel was damn hard, thanks to massive conflicts), writing drivers for a number of these, in addition to the usual installation, optimization, configuration, backup/restore, fixing of user issues, installing of software - from source or *bleagh* binary (including binaries for other OS', via the IBCS patch that used to exist for Linux), etc, ad nausium.
Ok, shown you one. That's what you asked for, that's what you got.
(Although I wouldn't expect a typical coder/sys admin to start on that kind of range, I've averaged 1 new OS every 2 years that I've worked with computers, keeping reasonably fresh on those not in immediate use. I would expect anyone who has been to University to be capable of learning at least 1 new OS every 4 years even with maintaining all their programming skills.)
Actually, yes. There were some credit card reader vendors in Europe who did not put coders to work on the problem and those card readers did NOT work after Y2K.
Here are other reports of failures - which were quickly fixed after they happened - but were Y2K-caused:
Slashdot Mirror
Will this do?
Erlang is an Ericsson product, and they've released or assisted on a number of other products. Sony - well, they're famous for closed products, rootkits and the walkman. I'm not overly convinced I trust those guys to honour prior license agreements (there are plenty of products that were GPL that have been made proprietary with the open source variant deleted from the catalogue completely). I am very concerned.
Not only that, the US has been demanding the payback of loans made during the wars (loans with interest, I might add) that would never have been required if the US hadn't initially sided with Germany and confiscated Allied property in the US. Hell, had the confiscations not happened, US involvement might never have been needed - Europe might have been able to afford an armed forces capable of resisting the Germans.
Blah
More blah
Happy now, or do you need me to hold your hand whilst you read them?
IE has not been (strictly) userspace since Windows 98.
No it isn't. The Intel Binary Compatibility Standard unifies all of that.
No, no, no and no. Go back to kindergarten where you belong.
Exactly. Reference texts exist for a reason. So long as you know the basics, the reference texts are all you ever need to be able to do the specifics.
And thanks for your post. Sure I have the skills, but I'm a freak in that respect. (I learned something of everything, which is why I had 10 BGP4+ IPv6 tunnels running in 1996 and was hosting 1/3rd of the IPv6 transatlantic backbone for a while.) But there's no need for me to be a freak in that respect. Anyone with a BSD or Linux box can use Quagga or Zebra to learn the core elements of IOS. Anyone with a BSD or Linux box, Webmin and a text editor (emacs works fine) can learn about how servers are configured and what configuration option in what configuration file does what. Or you can read the docs, but experimenting is usually a better way to learn.
You are correct, which is why there are plugins for Tcl, etc. Selenium is technically a client-side scripting language. Flash is blob, so not a technically a script, but it is still run client-side. Dunno if the attempt to run Python client-side ever got finished. There used to be a Mozilla plugin that supported the X11 protocol. The SCRIPT tag has a TYPE attribute which gives the media type. There is nothing in the spec that says this must be Javascript, it can be anything at all - though it helps if the browser supports it. The spec, then defines any language that can be described with the media type as being a valid language.
If people want features X, Y or Z not in the official specification for language L, use a different media type. It's not like the code will run anybloodyplace else anyway.
Show me 3D that requires that the page define whether it is to be rendered in software or hardware and I'll show you a specification that should be burned at the stake along with idiots like yourself.
I don't give a frag whether a given piece of 3D is rendered using SVG, VRML, GDML, OpenGL, DX11, PHIGS, Renderman, Maya, Rhino, Blender, a GPU, one of those insanely high-end nVidia modules that uses more power than every other computer in the house combined, that Chinese supercomputer built out of GPUs, or a cheese sandwich, so long as it is rendered correctly. The software should detect what options exist, use the best one (according to the built-in algorithm) and allow the user to switch to another.
There's bugger all HTML5 has to do with that process and only a moronic imbecile could think otherwise.
Rule 101, taught to EVERY BLOODY CS STUDENT ALIVE, is NEVER EVER EVER test your own code. You WILL miss things. ALWAYS have the code tested by someone else or - in the case of Extreme Programming methods where you write the test harnesses in advance - by something else. But NEVER test it yourself.
I read the judgement when it came out. I have worked with companies carrying out the court instructions. What's your excuse?
Yes, you can replace the entire kernel. Not just with another Linux kernel - anything that supports the Linux ABI will work, so you could replace the Linux kernel with Lynx if you wanted. FreeBSD should also work. There are probably others.
It goes beyond Unix. Intel defined the Intel Binary Compatibility Standard to facilitate ANY OS whatsoever running ANY software from ANY OTHER OS, provided both were written to the spec. Thus, there's nothing to prevent you from running a Solaris application dynamically linked to a Unixware library all under the Linux kernel. Yes, even Intel believed that vendor interoperability was important.
Internal to the kernel, it makes bugger all difference whether you're running the graphics through Framebuffer, KGI, a proprietary driver that can bind to X or a graphics-to-ASCII-art converter (yes, they exist). Everything still works exactly the same, except that Doom looks a bit odd. It still works, though.
Actually, yes. A lot of people "bother with that". In the scientific computing world, it means you don't have to care *whose* BLAS or LAPACK library you use. This is great. You can design using standalone libraries designed specifically to assist debugging and run against parallel libraries optimized for sheer speed - even when they're written by completely different groups.
In the GUI world, do you really care if you're using Motif or Lestif? Or whether that's really SGI's OpenGL or actually the Mesa 3D library? Can you name a single X11 program that breaks when using a custom implementation of X11 rather than the reference version? After all, it links to all kinds of libraries!
So, yes, every company -- barring Microsoft -- already gives the option to substitute ALL shared libraries. Microsoft is about the sole exception and it is a stupid one.
Does anyone care? Well, define "care". They care that their programs "just work" and that they don't need a billion essentially identical libraries to get them to do so. They care that they can tune and tweak. They care that updating external components or replacing them with something functionally the same will not break anything.
They don't care which specific library is installed, unless there's one optimized the way they want, precisely because things "just work". There's about a dozen standard C libraries - not because anyone seriously thinks people want to get the complete set, but because that lets users tailor their system to their needs, rather than tailoring their needs to what some vendor has decreed.
THAT is why they would bother.
Who cares if the rendering engine is shared? Well, if it's the rendering engine that is shared, take it out of IE and make it an independent shared component. Then people can uninstall IE if they want. Tying the rendering engine into IE and thus preventing people from uninstalling IE is not a sound software design, it is merely an abuse of a monopoly in an effort to gain another monopoly. Which is a criminal act.
People WOULD care if they could replace the rendering engine. There are other HTML5 rendering engines out there and being able to replace one with another would allow me to use whatever look-and-feel I liked without having to replace the GUI entirely. I should not have to replace Explorer, but I point out that you CAN. That people HAVE rewritten Afterstep as an Explorer replacement. That project was damn popular. Why? Because people actually DO like having a say over the L&F.
Because you can't gain a monopoly on English by exploiting a monopoly on Windows.
You can and WILL gain a monopoly on browsers by exploiting a monopoly on Windows.
The latter is a criminal offense, for which MS has been convicted (twice), and no amount of asking "why" will change that.
JS isn't a W3C standard. It can be obtained as a standalone engine. Provided Google's Chrome can support the loading of that engine, it makes not one whit of difference whether Google adds other engines to Chrome or remove their own JS engine. The only requirement for JS compatibility is that SOME JS engine be loadable at SOME point. It doesn't have to be built-in.
In fact, it's probably better if it isn't. Lightweight tools are generally superior tools. Having JS as a plugin would ensure that you could use any JS out there -- if you wanted to -- but that if you didn't want to run JS at all, you had zero overheads. THAT is Superior Design. It is ALSO Classic Unix Design Philosophy.
Actually, the basis of the Windows 95/98 lawsuit and the later IE bundling under XP lawsuit was that libraries ARE crafted to include totally irrelevant code. Indeed, it was Microsoft's position in the lawsuit that Felton's hack could not possibly work BECAUSE they had included such code. (Felton's hack worked because it left the extraneous code intact and in place.)
Nonetheless, even Microsoft disagrees with you. Under oath.
First off, not an excuse. Selenium means that testing one browser or a hundred different brands takes the same time and the same level of complexity.
Second off, no competent vendor has extensions to HTML, CSS or JS. Competent vendors do EVERYTHING in the standards, which are quite powerful enough. It is a mark of incompetency that Microsoft not only does NOT implement the standards, they fill the gaps with proprietary crap.
Third, developers should never test their own code. That is a sign of an untrained and moronic developer.
Fourth, if you are required by a court to SPECIFY all of these APIs and ABIs, then you are violating the law to not specify them. That is absolutely final.
Fifth, if you are required by a court to ALLOW a drop-in replacement for any given set of APIs and ABIs, then you are violating the law to not permit such a replacement. That too is final.
Included is not the same as required. Even if you installed Ubuntu (which has Firefox), you aren't required to install it. It's optional. Thus it isn't tied in. Further, even if you install it, you can later uninstall it when you discover Chrome does most of the stuff Firefox does better.
Firefox isn't in the OS in Linux (or any other OS). Firefox isn't an OS program. It is a user application. There is a HUGE difference. It is hard to describe all the ways it is different without causing the Slashdot machines to run out of space for database files, there are that many.
Removing Firefox from an installation won't cause X11 or the Linux kernel to destabilize. Removing IE will cause Windows 8's kernel to break at the lowest level.
No, that's not because Windows 8 does low-level graphics stuff. Install KGI (the Kernel Graphics Interface) or use Framebuffers extensively. Run X through them. Run KDE through them. Run a browser - say, Opera, Chrome, or even Firefox, through KDE. Now trace through the code and show me where KGI or FB code directly invokes a Firefox library. I dare you to try. Go on! Should be easy.
Yes, users do NOT appreciate any web browser in the OS. They want browsers in Application Space, where a bug won't cause the machine to crash and where switching to something else is easy rather than a 8-month hack with VTune, process-grabbing debuggers and a decompiler.
I don't. GNOME should permit any library that is API and ABI compatible and should not depend on specific implementations of anything. Used to be that GNOME did NOT depend on specific implementations, that you could choose between anything that provided identical functionality. Technically, since the source is out there, that's still the case but it should never have been the case that they restricted themselves to one solution alone.
Nonetheless, GNOME is not an Operating System, the Linux kernel won't break if you don't install GNOME (or indeed X). Whereas, Windows' kernel WILL break if IE is missing. Thus, your comparison is flawed. Probably knowingly.
A brain surgeon (with computing experience) would point out that standalone rendering engines have existed for years - and have existed for longer than any of the current browsers out there. Having the HTML5 rendering engine in a standalone DLL that could be replaced by anyone else's HTML5 rendering engine would NOT be an OS tie-in.
Since HTML5 rendering engines do NOT need a browser (since they can be standalone), a browser is NOT needed for this.
However, if you absolutely insist that a browser provide the library, a published specification (as per the requirements of the anti-trust suit, I might add) of exactly what functions are needed in the library, what name they must have and what ABI they must use, ANY web browser could be used. This is lawful under the requirements. A tie-in is NOT.
This is a flagrant violation of the law, which Microsoft will get away with because nobody dares start controversial lawsuits in an election year. Nonetheless, it IS illegal and it IS unnecessarily illegal. It is done this way for one reason and one alone - to kill competition. That is ALL it is being done for. It isn't for convenience and it isn't for the HTML5.
Bullshit to all 3 points.
The OS is a kernel plus core system libraries. It has ZERO relationship to how the output is displayed. Which is why I not only should be able to run KDE and AfterStep under Windows, I can.
What I CANNOT do is run Internet Explorer on Linux. So what if it's compiled for another OS and I don't have the source? I don't have the source for Solaris-x86 Oracle but I CAN run that under Linux (different OS and no source) just fine. Have been able to for years.
Yes, when you open a file panel or a network browser under Windows, you are using IE. The desktop is IE. The control panel is IE. Friggin' everything is IE! Even if you install another browser, you CANNOT tell those components to use it. So, yes, if you use Windows, you MUST use IE. You have no choice. And must you use Windows? Well, yes. Many web applications aren't written to international standards, they're written to Microsoft-proprietary functionality within IE. This WILL worsen, with this news about IE and Windows 8, just as it worsened considerably after Microsoft violated the Windows 95 injunction by releasing the bundled IE as Windows 98.
The competition is hurting something chronic. IE has rising usage figures. Firefox is starting to slide. Opera is sliding badly. Chrome may run foul of the Apple vs Google battle-to-the-death. (And one of them WILL die in it, if they don't back off.) Linux has never been fairly or reasonably offered as a desktop choice by anyone other than the OLPC group - and even they are now getting into bed with Microsoft.
Microsoft is a devout monopolist and it WILL kill anything that threatens that monopoly, no matter how savage or ugly they have to get to do so.
There is nothing in Linux which requires Firefox. Firefox is pre-installed, but only on specific distros. Other distros include other browsers, or no browser at all. (You don't need one - wget is perfectly good.)
This is different than with IE and Windows. If you remove IE, components totally unrelated to web browsing or the Internet WILL fail, because the libraries are crafted to include totally irrelevant code that is critical for other components. Because Microsoft do not publish the specs for these libraries, crafting replacements that ONLY have the bits needed for the rest of the system to function is almost impossible. Not completely impossible, just very very very hard.
There simply isn't any comparison between willful sabotage of the user and a simple pre-install, even if your claim that Firefox was pre-installed with Linux was correct.
Indeed. Didn't the US do this for a longish while, signing up to the International treaties after they'd got the good stuff?
I'm not saying that China is "faultless" or the US is "all bad", but let's face facts - America would not have made the progress it has if it had respected European patent laws and European property rights. If it wants to claim China is in the wrong, then I have nothing against that provided it is NOT for the purpose of maintaining a hegemony obtained solely through the same practices. If China is guilty, then American corporations and the American government owe Europe a percentage of the profits secured through IP theft.
Sure, that might push the US into recession. Isolating China and closing down all counterfeit goods plus genuine goods based on stolen IP would not merely put China into recession, it would bankrupt it. If you're willing to do the latter, you should be man enough to accept the former.
The good news is that 100+ years of compound interest for some of the products and 60+ years of accumulated value in the case of property illegally confiscated from British and other Allied nations during WW2 should cover the combined debts of Italy, Greece, Spain and Ireland, and leave enough left over for the heads of State to put in advance orders for GTA 5.
I agree with you. The American model produces people that are too inflexible and too inexperienced. The problem with the Japanese model is that it is too closed-loop. In just the same way that you clean fishtanks by keeping some things fresh and the fish original, the Japanese need to split off those elements that have become stale and/or toxic without replacing those with experience. ie: a partially-open look. That would breed innovation without also breeding the toxin of ignorance.
I've admined 386BSD, FreeBSD, NetBSD, OpenBSD, Linux (0.1 - 3.1), SunOS, Solaris, IRIX, HPUX, OSF/1, VxWorks, Windows 200x Server, Windows XP, Windows 7, Plan 9, Inferno, the PDP-11 and OS/X. Not sure I'd call RTAI or Xenomai distinct OS'. Montavista certainly wasn't. Dealt with all three.
This includes direct kernel work (hacking patches together to form the Functionally Overloaded Linux Kernel was damn hard, thanks to massive conflicts), writing drivers for a number of these, in addition to the usual installation, optimization, configuration, backup/restore, fixing of user issues, installing of software - from source or *bleagh* binary (including binaries for other OS', via the IBCS patch that used to exist for Linux), etc, ad nausium.
Ok, shown you one. That's what you asked for, that's what you got.
(Although I wouldn't expect a typical coder/sys admin to start on that kind of range, I've averaged 1 new OS every 2 years that I've worked with computers, keeping reasonably fresh on those not in immediate use. I would expect anyone who has been to University to be capable of learning at least 1 new OS every 4 years even with maintaining all their programming skills.)
Actually, yes. There were some credit card reader vendors in Europe who did not put coders to work on the problem and those card readers did NOT work after Y2K.
Here are other reports of failures - which were quickly fixed after they happened - but were Y2K-caused:
So, there you have the proof that if nothing was done, problems existed.