Q: But can't you just turn it off? A: Sure - unless your system administrator configures your machine in such a way that TCPA is mandatory, you can always turn it off. You can then run your PC with administrator privileges, and use insecure applications.
What that translates into is that you can run Linux and Linux applications (or other non-Microsoft operating systems) on the PC without having to worry about this nonsense. It would keep you from playing copy-controlled proprietary content (because you wouldn't be able to present the right credentials to the remote site to get the data), but that's just fine as far as I'm concerned. I think we couldn't hope for a better booster to Linux market penetration or open content than this.
Do you really think Intel is going to ship processors that require a key to run any piece of software? Intel's competitors would be overjoyed. There are dozens of chip manufacturers just itching to fill the void with RISC chips.k
What this will do is allow the OS to put the CPU into a mode where it only executes signed code. If you don't want it, you don't use it.
Besides, with JITs and interpreters, it doesn't really help security anyway--a lot of Microsoft's security problems are with interpreted code. You know: stupid little stuff, like VBscript and JScript, and if they can't get those interpreters to do the right checking, they won't enforce signatures correctly either.
Palladium is like another one of those gee-whiz ideas that people at Microsoft are prone to fixating on. You know, like putting a database into the kernel or using object marshalling for writing application data to disk. Ideas the rest of the world had and discarded decades ago because they are so stupid.
Sorry that you missed my shorthand remark. My point is that misbehaved Java threads generally don't crash each other. Misbehaved C threads easily do. That's why it is OK to build a multithreaded Java web server evev if you don't need the threading, while it is silly to build a multithreaded C web server unless the threading is really essential to your application.
Well, first note that the article talks about corporate usage, not general usage. Also, it's hard to compare the Google statistics to other statistics around the web. Are those unique clients? By cookies? Do they count hits from Sherlock and similar tools? In terms of number of raw hits, looking around web server statistics around the web, the numbers seem more even.
At the very least, even the Google statistics suggest that the figures for Linux and OS X (which is a fraction of Macintosh usage) are in the same ballpark. Linux isn't the obscure cousin of Macintosh OS X even on Google.
Don't get me wrong: I think Macintosh OS X is great, and the more successful it is, the better (well, up to around 25% market share--beyond that, anything goes bad). But Macintosh users would do well to start respecting Linux as a viable desktop choice as well rather than badmouthing it (as many are wont to do).
15Mbytes? So what? Even if none of that data were shared virtual memory, that's peanuts on modern machines. You can easily run 100 of those (=1.5Gbyte) simultaneously on a single machine--you'll bring the processor and network to its knees before you run out of memory.
Note that JVMs are much bigger, but unlike Apache, a JVM can actually do threading safely in a single address space.
Sorry, but you must have misunderstood. I made no assertion about the quality of Linux threading. Threading is simply the wrong concurrency model for this application: there is no need to run multiple web server requests in the same address space; in fact, doing so is harmful.
Just because you have a hammer (kernel threading) on your particular pet operating system (BSD? Solaris? NT?) doesn't mean that you need to go around hitting everything with it.
people quickly discover that they don't need your module all that much anyhow.
And how is that bad? Commercial software houses have an incentive to confuse users into buying zillions of useless packages, needed or not. But for open source software, both the maintainer's and the user's interests are aligned: if it isn't needed, nobody should waste their time on it.
The forking model used in Apache 1.x works great on UNIX platforms and is, for practical purposes, all that is needed. Apache 2 with its threading support is likely to be less reliable and harder to extend for a performance gain that is meaningless to almost every site in existence.
I wouldn't be surprised if many UNIX users don't ever go for this and Apache 1.x just branches off into a separate project. Apache 2 can turn into some kind of specialized Apache derivative for platforms that just can't handle forking; we shouldn't keep burdening UNIX software with accomodating those other kludgy operating systems.
If I did that, fine positioning would go out the window. You have to face the fact that human hands are only capable of so much range and so much precision--no mouse setting can get around that.
That should have been Linux is up to twice as common as Macintosh (by which I mean, accounts for twice as many hits). And, again, keep in mind that this is all Macintosh platforms, not just OSX.
Check for yourself: you can find statistics from various web sites through Google.
I was looking at a lot of web browser statistics around the web recently. I found pretty consistently that Linux was at least as common as all Macintosh platforms combined, and on many sites quite as common.
I'm not surprised either. KDE and Gnome are easily set up to behave almost indistinguishably from Windows--non-techies often can't tell the difference. And Linux comes with a complete suite of applications--OpenOffice and Mozilla really do satisfy the needs of most users.
The biggest problem with Linux, in my opinion, is the excruciatingly painful way in which drivers and other kernel extensions are installed--often involving recompiling the kernel. Even the most painless driver distributions (e.g., nVidia) require much more computer know-how to install than the average user can muster. In corporate settings, this doesn't matter that much--the IT department probably likes it that people can't just plug things in. But in the small business and home market, it matters big time.
Civilized monarchs, back when they still had such people, would consider it a sign of confidence, style, and good behavior to hire and pay people who disagreed and argued with them. They also found such people useful to have around because they'd get them thinking in new ways. Some of those people were fools (their job title, not their intelligence), others philosophers. HP needs some good fools and some good philosophers if they want to make it.
Well, whether it's faster or not is an experimental question. For me, I'm pretty certain it's not faster because I can't move across the whole screen with a single sweep of the mouse. Keep in mind, too, that if the mouse is to the far right of a menu entry, the target represented by that menu is very small (as viewed from the mouse position).
Also, one has to be careful interpreting such data. Let's say, for the sake of argument, that hitting the Mac-style menu is still faster even on a large screen, at least for people with their mice adjusted in a certain way. You still have extra eye movements and head movements and a shift of attention when you try to find the menu and when you try to get back to your application. The overall action may still be slower, or it may be more disruptive to one's work, even if just the act of activating the menu is faster.
Companies like IBM have a large contingent of people that loathe Microsoft and aren't afraid to speak out about it, yet IBM seems to be able to deal with both Microsoft and its customers just fine, and IBM is able to deliver a wide variety of systems. In fact, from a customer's point of view this is good: if a company has Windows, UNIX, mainframe, and open source people inside it, it is much more likely that advice and recommendations from the company will be based on technical considerations, rather than whatever is available. With a company that only bets on Microsoft, you already know the answer to any of your questions is going to be "Microsoft", whether that makes sense or not.
If HP is so threatened by a single person like Perens, they must really be in deep trouble. Apparently, The New HP is trying hard to become The New Unisys. Too bad--DEC and HP used to be nice companies. Compaq just keeps eating up one company after another, digesting them, well, and you know what comes out the other end.
No: it is easy to write UNreliable software if your methodology is not up to scratch. Given proper techniques and rigour, writing reliable software in C++ is *easy*.
Given that you are so brilliant, go work for Microsoft or Netscape or any of a number of big companies, who evidently lack the "proper technique and rigour" to produce reliable software. If it's so easy, you should be able to fix their numerous bugs and problems in no time.
I've been programming for many years and can now happily churn out C/C++ applications (many of which are multi-threaded) which (post-testing) never suffer unrecoverable errors.
And if you stay in the business for many more years, you'll realize the hubris and folly of that statement.
The myth that C/C++ is difficult
I made no statement to the effect of whether C/C++ is hard or not. I actually think C is a pretty easy language to learn. But it can be hard to write reliable software in otherwise easy languages, and it can be hard to verify compilers for easy languages.
In fact, C++ is one of the most difficult languages to verify a compiler for, and I have personally tracked down a number of bugs in both commercial and open source compilers. The fact that C++ compilers are hard to verify alone would be a reason not to use the language on critical tasks, even all other issues aside.
cause the developers don't have to work as hard to produce games (windows ce ports)
Actually, it's NT ports. In any case, developers had to work extra hard to get the stuff running under Windows in the first place.
Sony could realize a similar advantage by going more strongly with open standards for the PS2: OpenGL, POSIX, etc. That way, developers could keep roughly the same code for PS2, Windows, Macintosh, and Linux, even if they need some platform tweaks on each.
Note that they are most likely using GNU software. Here is a list of the software development environments for these chips, and Here is the European Space Agency's web page for the tools and emulator.
It's easy to store and transport energy cleanly on a large scale: using hydrogen. In fact, in addition to locally generated wind energy, solar energy generated in the Sahara and other deserts and shipped around the world as hydrogen could also contribute to a clean, renewable energy infrastructure.
(Besides, your premise is wrong: wind is not limited to spring and fall in many places.)
No, it's much better not to have to worry about upgrading individual software packages manually at all. But Macintosh is almost as primitive as Windows when it comes to software installation: on Windows, you fiddle around with installer dialog boxes, on Macintosh, you muck around in the file system. Come on, move into the 21st century.
A little knowledge is a dangerous thing. All things being equal, putting targets at the edge of the screen is good. But all things are not equal. When I work in a window on the top, right hand side of my 23" screen, aiming the mouse for a menu at the top left hand side of the screen gives me a tiny target--this is not faster than putting the menu at the top of the window. Putting menus at the top of the screen was a good choice for the original Mac. It's a stupid choice for the huge, high resolution screens we have today.
In fact, if speed mattered that much, the best thing to use would menus, preferably pie menus that pop up under the mouse. This is the way that GUIs were going before Apple came along. But Apple, in their infinite wisdom, popularized menu bars.
Mac OS X apps always create a new document (or viewer, main window, etc., as appropriate to the app) when you activate them,
Maybe that's what they are supposed to do, but they don't. Mozilla is an example of an application that doesn't, and there are others. People click on the dock icon and... nothing seems to happen.
Trying to write such a system in C/C++ strikes me as rather stupid. It is extremely hard to write reliable software in C/C++. That may not matter much for desktop applications, but it matters when billions of dollars are in the balance.
They obviously don't need very high performance, since it runs on 1970s hardware, but they do need high reliability and low development costs.
That means that they should be using a safe, secure high-level language. Something with a virtual machine might be a good idea so that it will be easy to adapt to new hardware platforms: you verify the virtual machine on the new machine and then have reasonable confidence that your code runs.
If they want something in widespread use, a home-built Java byte-code interpreter (not a JIT--they are too buggy) might be a reasonable choice--it's well specified and there are lots of people who know how to program it. They should probably avoid JNI like the plague and instead add new bytecodes for I/O and communications and verify them the same way that they do the virtual machine itself.. VLISP might be another good choice--or at least a source of ideas for how to implement a verified Java interpreter--DARPA already has paid for its development.
And they should hire someone who doesn't recommed COTS with C++, lest we see the next shuttle go up in flames again.
If there was no protection to intellectual property, people would not be encouraged to share knowledge with others. Writers would not write, inventors would not invent, artists would not.
If I look at the crap that people put out today with the protection of IP laws, I think this would actually be a good thing. The profit motive seems to lead to lots of writing, art, and inventions, but little of it any good.
History shows that the artists, writers, and inventors who remain seem to do well enough. You know, like most of the great artists, scientists, and writers who lived before the 20th century and mostly worked for hire, on comission, or supported themselves with other work.
Slashdot Mirror
Q: But can't you just turn it off? A: Sure - unless your system administrator configures your machine in such a way that TCPA is mandatory, you can always turn it off. You can then run your PC with administrator privileges, and use insecure applications.
What that translates into is that you can run Linux and Linux applications (or other non-Microsoft operating systems) on the PC without having to worry about this nonsense. It would keep you from playing copy-controlled proprietary content (because you wouldn't be able to present the right credentials to the remote site to get the data), but that's just fine as far as I'm concerned. I think we couldn't hope for a better booster to Linux market penetration or open content than this.
What this will do is allow the OS to put the CPU into a mode where it only executes signed code. If you don't want it, you don't use it.
Besides, with JITs and interpreters, it doesn't really help security anyway--a lot of Microsoft's security problems are with interpreted code. You know: stupid little stuff, like VBscript and JScript, and if they can't get those interpreters to do the right checking, they won't enforce signatures correctly either.
Palladium is like another one of those gee-whiz ideas that people at Microsoft are prone to fixating on. You know, like putting a database into the kernel or using object marshalling for writing application data to disk. Ideas the rest of the world had and discarded decades ago because they are so stupid.
Sorry that you missed my shorthand remark. My point is that misbehaved Java threads generally don't crash each other. Misbehaved C threads easily do. That's why it is OK to build a multithreaded Java web server evev if you don't need the threading, while it is silly to build a multithreaded C web server unless the threading is really essential to your application.
At the very least, even the Google statistics suggest that the figures for Linux and OS X (which is a fraction of Macintosh usage) are in the same ballpark. Linux isn't the obscure cousin of Macintosh OS X even on Google.
Don't get me wrong: I think Macintosh OS X is great, and the more successful it is, the better (well, up to around 25% market share--beyond that, anything goes bad). But Macintosh users would do well to start respecting Linux as a viable desktop choice as well rather than badmouthing it (as many are wont to do).
Note that JVMs are much bigger, but unlike Apache, a JVM can actually do threading safely in a single address space.
Just because you have a hammer (kernel threading) on your particular pet operating system (BSD? Solaris? NT?) doesn't mean that you need to go around hitting everything with it.
And how is that bad? Commercial software houses have an incentive to confuse users into buying zillions of useless packages, needed or not. But for open source software, both the maintainer's and the user's interests are aligned: if it isn't needed, nobody should waste their time on it.
I wouldn't be surprised if many UNIX users don't ever go for this and Apache 1.x just branches off into a separate project. Apache 2 can turn into some kind of specialized Apache derivative for platforms that just can't handle forking; we shouldn't keep burdening UNIX software with accomodating those other kludgy operating systems.
If I did that, fine positioning would go out the window. You have to face the fact that human hands are only capable of so much range and so much precision--no mouse setting can get around that.
That should have been Linux is up to twice as common as Macintosh (by which I mean, accounts for twice as many hits). And, again, keep in mind that this is all Macintosh platforms, not just OSX.
Check for yourself: you can find statistics from various web sites through Google.
I'm not surprised either. KDE and Gnome are easily set up to behave almost indistinguishably from Windows--non-techies often can't tell the difference. And Linux comes with a complete suite of applications--OpenOffice and Mozilla really do satisfy the needs of most users.
The biggest problem with Linux, in my opinion, is the excruciatingly painful way in which drivers and other kernel extensions are installed--often involving recompiling the kernel. Even the most painless driver distributions (e.g., nVidia) require much more computer know-how to install than the average user can muster. In corporate settings, this doesn't matter that much--the IT department probably likes it that people can't just plug things in. But in the small business and home market, it matters big time.
Civilized monarchs, back when they still had such people, would consider it a sign of confidence, style, and good behavior to hire and pay people who disagreed and argued with them. They also found such people useful to have around because they'd get them thinking in new ways. Some of those people were fools (their job title, not their intelligence), others philosophers. HP needs some good fools and some good philosophers if they want to make it.
Also, one has to be careful interpreting such data. Let's say, for the sake of argument, that hitting the Mac-style menu is still faster even on a large screen, at least for people with their mice adjusted in a certain way. You still have extra eye movements and head movements and a shift of attention when you try to find the menu and when you try to get back to your application. The overall action may still be slower, or it may be more disruptive to one's work, even if just the act of activating the menu is faster.
If HP is so threatened by a single person like Perens, they must really be in deep trouble. Apparently, The New HP is trying hard to become The New Unisys. Too bad--DEC and HP used to be nice companies. Compaq just keeps eating up one company after another, digesting them, well, and you know what comes out the other end.
Given that you are so brilliant, go work for Microsoft or Netscape or any of a number of big companies, who evidently lack the "proper technique and rigour" to produce reliable software. If it's so easy, you should be able to fix their numerous bugs and problems in no time.
And if you stay in the business for many more years, you'll realize the hubris and folly of that statement.
The myth that C/C++ is difficult
I made no statement to the effect of whether C/C++ is hard or not. I actually think C is a pretty easy language to learn. But it can be hard to write reliable software in otherwise easy languages, and it can be hard to verify compilers for easy languages.
In fact, C++ is one of the most difficult languages to verify a compiler for, and I have personally tracked down a number of bugs in both commercial and open source compilers. The fact that C++ compilers are hard to verify alone would be a reason not to use the language on critical tasks, even all other issues aside.
Actually, it's NT ports. In any case, developers had to work extra hard to get the stuff running under Windows in the first place.
Sony could realize a similar advantage by going more strongly with open standards for the PS2: OpenGL, POSIX, etc. That way, developers could keep roughly the same code for PS2, Windows, Macintosh, and Linux, even if they need some platform tweaks on each.
Note that they are most likely using GNU software. Here is a list of the software development environments for these chips, and Here is the European Space Agency's web page for the tools and emulator.
(Besides, your premise is wrong: wind is not limited to spring and fall in many places.)
No, it's much better not to have to worry about upgrading individual software packages manually at all. But Macintosh is almost as primitive as Windows when it comes to software installation: on Windows, you fiddle around with installer dialog boxes, on Macintosh, you muck around in the file system. Come on, move into the 21st century.
In fact, if speed mattered that much, the best thing to use would menus, preferably pie menus that pop up under the mouse. This is the way that GUIs were going before Apple came along. But Apple, in their infinite wisdom, popularized menu bars.
Maybe that's what they are supposed to do, but they don't. Mozilla is an example of an application that doesn't, and there are others. People click on the dock icon and ... nothing seems to happen.
They obviously don't need very high performance, since it runs on 1970s hardware, but they do need high reliability and low development costs.
That means that they should be using a safe, secure high-level language. Something with a virtual machine might be a good idea so that it will be easy to adapt to new hardware platforms: you verify the virtual machine on the new machine and then have reasonable confidence that your code runs.
If they want something in widespread use, a home-built Java byte-code interpreter (not a JIT--they are too buggy) might be a reasonable choice--it's well specified and there are lots of people who know how to program it. They should probably avoid JNI like the plague and instead add new bytecodes for I/O and communications and verify them the same way that they do the virtual machine itself.. VLISP might be another good choice--or at least a source of ideas for how to implement a verified Java interpreter--DARPA already has paid for its development.
And they should hire someone who doesn't recommed COTS with C++, lest we see the next shuttle go up in flames again.
If I look at the crap that people put out today with the protection of IP laws, I think this would actually be a good thing. The profit motive seems to lead to lots of writing, art, and inventions, but little of it any good.
History shows that the artists, writers, and inventors who remain seem to do well enough. You know, like most of the great artists, scientists, and writers who lived before the 20th century and mostly worked for hire, on comission, or supported themselves with other work.
Indeed. As a Linux developer, I always feel like I'm back in the stone age when I use VisualStudio.