Unfortunately, I am quite convinced that you're wrong. You're assuming that with DRM, the media is secure, which I just cannot believe. There will always be a way to crack it, and the larger pirates will be able to do so, be it by stealing the encryption key from the source, finding a bug in Windows or by simply cracking the system. As long as there is the capability to play back the media losslessly on someone's system, there will be the capability to capture it.
Thus, the pirates have little to lose with DRM (the only thing it will do is to make it harder to copy something, but not impossible), while the average customer just wanting to make a mix CD for his car or tape a TV show has much more to lose, since they usually will be unable to crack the encryption (and there will always be companies that place to restrictive terms on their media). In fact, the larger pirate networks may even support DRM, since it may be able to make the media companies feel falsely secure. Needless to say, once a single pirate has cracked the protection on piece of media, that piece of media is free for all, forever.
The conclusion is the DRM has no positive consequences for anyone.
I hadn't heard of that before, and even now it's a bit hard for me to believe that. If hotplugging PS/2 equipment could truly be damaging to the system, I can't imagine that the designers wouldn't have gone through more effort to secure the connectors far more. It isn't exactly just a few times that PS/2 equipment has come loose and disconnected "on its own". When that happens, it is the natural reaction of anyone without intimate knowledge of the system to just reconnect it again, and if such a simple act could damage the system, it would be easy for the victims to sue the computer manufacturer.
USB can be plugged in and unplugged while the machine is running. PS/2 cannot.
This is an unimportant point in the context, but I still feel the need to point out that it's wrong. PS/2 devices can be plugged and unplugged during operation, and they even have hotplug detect. In Linux, it doesn't matter the slightest whether you put the mouse in the keyboard plug or the other way around. It's just Windows that doesn't support it, PS/2 does (maybe Windows does support it these days -- I haven't used it regularly since the days of W98).
That said, I completely and fully agree that USB is still better in every possible regard (except design simplicity, that is).
You know, I don't think RMS really cares about the practical benefits of free software. If you want someone who cares about that, try Eric Raymond instead. RMS believes everyone should use free software for ideological reasons (we all know them -- help your neighbor, etc. -- if not, just read up on gnu.org). Mind you, this is not a bashing of RMS, because to a very large extent I agree with him -- I'm just not being as active about it as he is. I just don't really think he's prepared any procedure to measure the practical benefits that these people will experience...
In the meantime, Microsoft almost seems tentative in their position about standards compliance versus backwards compatibility.
I don't understand why there has to be a conflict at all here, though. IE, like most other browsers, already has a quirks-mode/compliance-mode separation. Why not just go on doing the same old, bad job on the pages rendered on quirks mode, and then render correctly and compliantly on pages that specify proper DOCTYPEs etc.? It seems to me that the old "backwards compatibility" argument is just a bad excuse for Microsoft not to comply to standards.
Call me infidel if you will, but I would argue that the bottom left corner is indeed the One True position for the Control key. It is far easier to reach with the pinky finger than the leftmost middle row is. When using Emacs commands like C-x C-f with the Control key in the bottom left, notice that if you press Control with your pinky, X with your long finger and F with your index finger, the fingers line up nicely in a natural straight line. Then do the same with the Control key in the leftmost middle row, and note how you have to stretch the pinky finger surely, Oh Good Lord, an entire extra inch!
I'm amazed! Never have I seen anyone be so thoroughly wrong about the history of Unix. I know this is Slashdot, so most people here should know these things, but just in case someone gets a wrong impression, I feel I should at least clear up a few things.
He claims that the first version of Unix came with paging. To even credit the PDP-7 with paging capability is rather amazing. Unix used whole-process swapping until only much later in its development. If I'm not entirely mistaken, wasn't paging implemented in BSD3?
He appears to claim that Unix invented time sharing! I don't think I have to elaborate on that, really...
He also claims that "[Unix] would let two people use a computer at the same time." Not only is it false (it supported as many as there were terminals wired in), I also find it a bit funny that two-people time sharing would have been considered impressive at the time.
He seems to imply that "Uniplexed Information and Computing System" was an actual, official name of the system. To begin with, "Unics" wasn't really meant to be expanded -- it was just a pun on the "Multics" acronym (that is, a pun on the acronym, not on its expansion).
To mention that Unix was rewritten in C without mentioning that C was invented for that very purpose is of course not "incorrect", but I would argue that it is a rather important omission.
He writes that the first C version of Unix was "Unix System III", while in fact it was, of course, Third Edition (V3). System III was a much later release (~1980?) by the USG.
There are probably many more errors, but I stopped reading when I noticed that my eyes were bleeding.
As far as what they actually are doing, well, it's theft, plain and simple.
I don't know what definition of theft you are using, but the last time I looked, theft was the act of depriving someone of their property. When using an unlicensed copy of software, you are not depriving anyone of their rightful property. You could argue that you deprive them of potential property, but that's not the same thing as depriving them of actual property (if it were, no IP laws would be necessary to begin with).
As for what they actually are doing -- well, I'm not sure. I would guess that the crime is that they said to agree to the EULA without actually following its provisions, which would make it, what, breach of contract?
The issue is simple and obvious from the moral standpoint: Company A offers a product it cost $5 to make for $10. You use said product without paying for it. That's boldly obviously theft
I definitely disagree with that. There's neither any criminal offense nor, IMNSHO, any moral sin associated with using things that one does not own. Noone would require you to sign a form of temporary transfer of property or something like that to enter a computer store and play around with the products on display. Taking them out of the store is, of course, a completely different thing (that would be depriving the store of their rightfully owned property). Visiting the store's legally owned premises without their consent (which they are, of course, free to withdraw in case you do something with which they disagree) is also a different thing.
Well, the submitter's point was the fact that while the IBM PC has definitely been infinitely influential, it was very innovative, and thus didn't fit on the top 25 innovative products list. However, that seems to be a misunderstanding as a whole, since the TFA doesn't mention anything about being "most innovative", but actually is only a list of "top products".
Extensions can be happily installed inside a user's profile directory. It doesn't require write permissions to the Firefox application's directory to install an extension.
The basic problem is, of course, that to even get that far, the user has to have executed unknown/untrusted code, and at that point, it is too late on any system. Since that code can do anything else that the user can otherwise do (by running other code), there's nothing preventing untrusted code from installing itself in ~/.profile, HKLM\...\Run, etc. Take your pick.
There have, of course, been discussions, on most operating systems, about solving this "problem" by simply not allowing a user to run code that hasn't been signed by a sysadmin or similar, but that, too, has obvious problems.
I agree with the OP in sticking to C, Python and Perl, and it is for exactly some of the reasons that you listed. Java is, indeed, adequate for just about every programming task. For almost any programming task, however, there will be a language which is a perfect fit, which makes Java, in its adequateness, inferior for that particular task. I am lazy (as programmers should be), so I really dislike the wordieness of which you speak. I do work as a Java developer, and I have felt very hindered by the language in almost anything I do. Quite often I encounter subproblems that take 10+ lines in Java, which could be solved in a single, small and elegant s-expr in Scheme, or a single line of Perl. It is also quite cumbersome to do serious data structure mangling in Java, considering that you either have to define an entire class just to do simple things, or do so much type casting with the java.util structures that half of the code will just be type casts (and it isn't made any better by the facts that type names in Java tend to be very long and that the java.util collections can't store primitive data types).
I do find Java a good fit for a certain class of programming tasks -- namely those which require the software to run on a diversity of platforms. In that regard, I don't think anything can beat Java. If only Windows shipped with a good JRE...
However, I do disagree with some of your points. For example, no matter how good Swing is for programming, it will always suck since it renders widgets itself. Even though it has pluggable look-and-feel, it isn't possible to always make it look like a native application (try to write a look-and-feel for "the currently selected GTK2 theme", for example...). Also, even though Java's library support may be good, it quite much sucks that it is impossible to do OS-specific calls. I realize that that is because of the WORA principle, but it quite much sucks that it is impossible to, for instance, write Unix programs that use setuid in Java. I cannot imagine that it would hurt to add OS-specific calls for programs that just don't care about being OS-independent (especially not if they were added in seperate packages, like os.posix or os.win32, so that they won't be confused with OS-independent classes), right? By the way -- yes, I know about JNI and, no, I don't think it counts.
Basically, I think that the JVM and the class library are pretty nice (not perfect by a long shot, however), but the Java language itself just sucks. I have often considered making an alternative language that compiles to Java class files.
Actually, they are complying exactly to that, and that's why they're getting fined by the EU. Microsoft claims to "...make available, on commercially reasonable terms, all of the communications protocols..." (emph. mine), and the EU saw through that to mean the free software was excluded. Thus, they told Microsoft it wasn't good enough.
10. Communications protocols. Microsoft will make available, on commercially reasonable terms, all of the communications protocols that it has built into Windows and that are used to facilitate communication with server versions of Windows. To facilitate this, Microsoft will document protocols supported in Windows as part of the product design process. We will also work closely with firms with particular needs to address interoperability scenarios that may require licensing of other protocols.
11. Availability of Microsoft patents. Microsoft will generally license patents on its operating system inventions (other than those that differentiate the appearance of Microsoft's products) on fair and reasonable terms so long as licensees respect Microsoft's intellectual property rights.
(From Microsoft's site)
So in other words, it's a no-go for free software.
I also found the 12th point interesting:
12. Standards. Microsoft is committed to supporting a wide range of industry standards in Windows that developers can use to build interoperable products. Microsoft is committed to contributing to industry standard bodies as well as working to establish standards via ad hoc relationships with others in the industry.
What a commitment! I think I can pretty much count all the open standards that Microsoft can be said to be committed to support on one hand: the IP stack, DNS and HTTP. Even their FTP implementation is half-assed, to say the least (considering how one cannot get out of the initial cwd), and I doubt anyone would argue that Microsoft actually "supports" any of the web standards (that is, if one isn't viewing support of the version of 10 years ago as a "commitment").
The second sentence is interesting in its own right. I, for one, cannot interpret it to mean anything but inventing their own, new standard instead of the ones that already exist and work. "[W]orking to establish standards via ad hoc relationships with others in the industry" doesn't even need a comment...
All the other tenets were pretty well summed up by another poster as "don't poison the customer", "don't shoot the customer", "don't bomb the distributor" and "don't ignore direct orders from a court of law". They are probably reserving the rights to boiling the customer, burying the distributor alive and following order from a court of law at their own pace, however.
In Windoze and OS/2 (am I showing my age here:) it is possible to trap these type of exceptions on a per thread basis. You can then create a "manager" that does effectively what you had in the multi-process scenario. The exception handling code does whatever cleanup it can, and then triggers some action that will cause a thread to be spun up if an existing one bites the dust.
However, if the dying thread has run amok over the heap already, that's not going to help much. Instead, you are likely to get other mysterious failures on other threads 5 hours later with no debugging information. Neither do you get a core file of the dying thread for post mortem debugging.
I'd argue against that. Once you have a parse tree of a C program, you can more or less just output a fixed chunk of assembly for each construct. It isn't entirely true, of course, since the compiler handles things such as register allocation, but when everything comes around, a non-optimizing C compiler isn't much more than a preprocessor for the assembler. It's just a convenient way of writing assembler. Don't get me wrong, though -- I love C just for that reason.
Compare that to actual high-level languages like Haskell or Prolog, where you really need to create an abstract representation of the concepts in the program and examine their interactions with one another in order to even begin to compile it. Or, just look at languages like LISP or Python, which normally aren't compiled since it is much easier to just write a VM than a compiler which would really be able to handle every edge case. These languages aren't just convenient forms of assembler -- they are defined at a "higher level", isolated from the actual instruction set of whatever computer they run on.
I still don't really get it. Probably, that's because I'm not very good with Windows, but how does being able to send arbitrary window messages hurt? From what I remember from Windows programming, window messages are basically just input being sent to a window. How does fooling a window into thinking that the mouse is being moved constitute an attack?
I'd imagine that this could be a problem if a trojan were running in the same session as a window with administrator privileges, but is that all?
I'm not exactly in charge of any large area networks, so I'm probably just ignorant, but why would you want to limit physical Ethernet access to begin with? All your actual services are properly authenticated, aren't they? Is it for DoS prevention or proactive security or something completely else?
I, for one, would definitely trust an SSL channel far more than any phone service. If you type your SSN and/or PIN on the phone, all anyone would have to do is to connect two wires to your local distribution box to listen in. If they have broken Diffie-Hellman kex or RSA to break your SSL connection, though, I would think they'd have larger things to go after than your banking info, on the other hand.
Now when it comes to network printing, Windows domains have got *nix beat.
"Well what you do on *nix is you memorize the printers magic number, no, there is no searching GUI listing of printers setup, then you drop to the command line and pipe the file you want to print to the printer."
That's weird. Last I checked, CUPS has automatic printer sharing (using UDP broadcast). When I set up a LaserJet on my sister's computer, it could instantaneously be used on all other computers on the same link (of course, it's a bit harder when sharing a printer between two links, but all I had to do was to set up cupsd on the router to forward the discovery packets between the links, so it wasn't that bad -- most people probably don't have two subnets anyway). On newer distros, you may have to check a checkbox to enable network discovery, though (for security reasons).
What the heck is wrong with us? It should NOT be complicated.
Actually, it isn't that strange, since IPv4 was designed to be used on large, managed networks of minicomputers and mainframes... IPv6 has much better provisions for unmanaged networks.
Computers should not be non-deterministic.
Actually, it has proven that computer systems that involve simultaneous execution of several threads (such as SMP machines or networks of several UP machines) are non-deterministic.
What I don't understand is how they even are able to do this. I mean, if a user doesn't have WGA, it means he hasn't updated his Windows, right? If he hasn't updated his Windows, how is Microsoft supposed to get the software that can turn Windows off onto the user's system? Is it already there by default?
Slashdot Mirror
Thus, the pirates have little to lose with DRM (the only thing it will do is to make it harder to copy something, but not impossible), while the average customer just wanting to make a mix CD for his car or tape a TV show has much more to lose, since they usually will be unable to crack the encryption (and there will always be companies that place to restrictive terms on their media). In fact, the larger pirate networks may even support DRM, since it may be able to make the media companies feel falsely secure. Needless to say, once a single pirate has cracked the protection on piece of media, that piece of media is free for all, forever.
The conclusion is the DRM has no positive consequences for anyone.
I hadn't heard of that before, and even now it's a bit hard for me to believe that. If hotplugging PS/2 equipment could truly be damaging to the system, I can't imagine that the designers wouldn't have gone through more effort to secure the connectors far more. It isn't exactly just a few times that PS/2 equipment has come loose and disconnected "on its own". When that happens, it is the natural reaction of anyone without intimate knowledge of the system to just reconnect it again, and if such a simple act could damage the system, it would be easy for the victims to sue the computer manufacturer.
That said, I completely and fully agree that USB is still better in every possible regard (except design simplicity, that is).
You know, I don't think RMS really cares about the practical benefits of free software. If you want someone who cares about that, try Eric Raymond instead. RMS believes everyone should use free software for ideological reasons (we all know them -- help your neighbor, etc. -- if not, just read up on gnu.org). Mind you, this is not a bashing of RMS, because to a very large extent I agree with him -- I'm just not being as active about it as he is. I just don't really think he's prepared any procedure to measure the practical benefits that these people will experience...
Call me infidel if you will, but I would argue that the bottom left corner is indeed the One True position for the Control key. It is far easier to reach with the pinky finger than the leftmost middle row is. When using Emacs commands like C-x C-f with the Control key in the bottom left, notice that if you press Control with your pinky, X with your long finger and F with your index finger, the fingers line up nicely in a natural straight line. Then do the same with the Control key in the leftmost middle row, and note how you have to stretch the pinky finger surely, Oh Good Lord, an entire extra inch!
- He claims that the first version of Unix came with paging. To even credit the PDP-7 with paging capability is rather amazing. Unix used whole-process swapping until only much later in its development. If I'm not entirely mistaken, wasn't paging implemented in BSD3?
- He appears to claim that Unix invented time sharing! I don't think I have to elaborate on that, really...
- He also claims that "[Unix] would let two people use a computer at the same time." Not only is it false (it supported as many as there were terminals wired in), I also find it a bit funny that two-people time sharing would have been considered impressive at the time.
- He seems to imply that "Uniplexed Information and Computing System" was an actual, official name of the system. To begin with, "Unics" wasn't really meant to be expanded -- it was just a pun on the "Multics" acronym (that is, a pun on the acronym, not on its expansion).
- To mention that Unix was rewritten in C without mentioning that C was invented for that very purpose is of course not "incorrect", but I would argue that it is a rather important omission.
- He writes that the first C version of Unix was "Unix System III", while in fact it was, of course, Third Edition (V3). System III was a much later release (~1980?) by the USG.
There are probably many more errors, but I stopped reading when I noticed that my eyes were bleeding.As for what they actually are doing -- well, I'm not sure. I would guess that the crime is that they said to agree to the EULA without actually following its provisions, which would make it, what, breach of contract?
I definitely disagree with that. There's neither any criminal offense nor, IMNSHO, any moral sin associated with using things that one does not own. Noone would require you to sign a form of temporary transfer of property or something like that to enter a computer store and play around with the products on display. Taking them out of the store is, of course, a completely different thing (that would be depriving the store of their rightfully owned property). Visiting the store's legally owned premises without their consent (which they are, of course, free to withdraw in case you do something with which they disagree) is also a different thing.Well, the submitter's point was the fact that while the IBM PC has definitely been infinitely influential, it was very innovative, and thus didn't fit on the top 25 innovative products list. However, that seems to be a misunderstanding as a whole, since the TFA doesn't mention anything about being "most innovative", but actually is only a list of "top products".
Someone else suggested a panorama (read: vista) of a scrap dump. I still find that a rather good idea.
There have, of course, been discussions, on most operating systems, about solving this "problem" by simply not allowing a user to run code that hasn't been signed by a sysadmin or similar, but that, too, has obvious problems.
I do find Java a good fit for a certain class of programming tasks -- namely those which require the software to run on a diversity of platforms. In that regard, I don't think anything can beat Java. If only Windows shipped with a good JRE...
However, I do disagree with some of your points. For example, no matter how good Swing is for programming, it will always suck since it renders widgets itself. Even though it has pluggable look-and-feel, it isn't possible to always make it look like a native application (try to write a look-and-feel for "the currently selected GTK2 theme", for example...). Also, even though Java's library support may be good, it quite much sucks that it is impossible to do OS-specific calls. I realize that that is because of the WORA principle, but it quite much sucks that it is impossible to, for instance, write Unix programs that use setuid in Java. I cannot imagine that it would hurt to add OS-specific calls for programs that just don't care about being OS-independent (especially not if they were added in seperate packages, like os.posix or os.win32, so that they won't be confused with OS-independent classes), right? By the way -- yes, I know about JNI and, no, I don't think it counts.
Basically, I think that the JVM and the class library are pretty nice (not perfect by a long shot, however), but the Java language itself just sucks. I have often considered making an alternative language that compiles to Java class files.
Actually, they are complying exactly to that, and that's why they're getting fined by the EU. Microsoft claims to "...make available, on commercially reasonable terms, all of the communications protocols..." (emph. mine), and the EU saw through that to mean the free software was excluded. Thus, they told Microsoft it wasn't good enough.
So in other words, it's a no-go for free software.
I also found the 12th point interesting:
What a commitment! I think I can pretty much count all the open standards that Microsoft can be said to be committed to support on one hand: the IP stack, DNS and HTTP. Even their FTP implementation is half-assed, to say the least (considering how one cannot get out of the initial cwd), and I doubt anyone would argue that Microsoft actually "supports" any of the web standards (that is, if one isn't viewing support of the version of 10 years ago as a "commitment").The second sentence is interesting in its own right. I, for one, cannot interpret it to mean anything but inventing their own, new standard instead of the ones that already exist and work. "[W]orking to establish standards via ad hoc relationships with others in the industry" doesn't even need a comment...
All the other tenets were pretty well summed up by another poster as "don't poison the customer", "don't shoot the customer", "don't bomb the distributor" and "don't ignore direct orders from a court of law". They are probably reserving the rights to boiling the customer, burying the distributor alive and following order from a court of law at their own pace, however.
One would think, that if the code is destroyed, it won't be in Linux, and therefore no copyrights infringed, no?
Compare that to actual high-level languages like Haskell or Prolog, where you really need to create an abstract representation of the concepts in the program and examine their interactions with one another in order to even begin to compile it. Or, just look at languages like LISP or Python, which normally aren't compiled since it is much easier to just write a VM than a compiler which would really be able to handle every edge case. These languages aren't just convenient forms of assembler -- they are defined at a "higher level", isolated from the actual instruction set of whatever computer they run on.
I'd imagine that this could be a problem if a trojan were running in the same session as a window with administrator privileges, but is that all?
I'm not exactly in charge of any large area networks, so I'm probably just ignorant, but why would you want to limit physical Ethernet access to begin with? All your actual services are properly authenticated, aren't they? Is it for DoS prevention or proactive security or something completely else?
I, for one, would definitely trust an SSL channel far more than any phone service. If you type your SSN and/or PIN on the phone, all anyone would have to do is to connect two wires to your local distribution box to listen in. If they have broken Diffie-Hellman kex or RSA to break your SSL connection, though, I would think they'd have larger things to go after than your banking info, on the other hand.
What I don't understand is how they even are able to do this. I mean, if a user doesn't have WGA, it means he hasn't updated his Windows, right? If he hasn't updated his Windows, how is Microsoft supposed to get the software that can turn Windows off onto the user's system? Is it already there by default?