Some tasks simply need low-level features. If you want to use a high-level language most of the time, you have got several choices: You can write the low-level parts in a different, low-level language, you can try to outgess the compiler vendor and write source code which compiles to the required machine code using a specific compiler version (in many cases, this is the C approach), or you can use a high-level language which supports low-level programming. The third choice does not have to be the worst, especially if the low-level language features are clearly separated from the high-level ones (C fails miserably in this area).
Remember that many mission-critical computer systems are implemented in Ada, which offers a wide range of very low-level features (interrupt handling, representation clauses, unchecked conversion of objects, and so on). Unlike C, there is also explicit support for machine addresses and address arithmetic.
However, you should keep in mind that only certain types of security problems can be avoided automatically by choosing an appropriate programming language. Buffer overflow and format string bugs are found in almost all C software, but they are not the only cause of problems.
Aegis does not only deal with source code management, but can be used to enforce a development process which includes steps like testing and peer review.
You can implement the same thing using arch or CVS, of course, but Aegis offers much more structure in this regard.
And the argument that you would expect less security holes in GNU/Linux systems because they are not as widely deployed as Microsoft-Windows-based systems is not convincing at all. Security vulnerabilities are there even if hardly anybody uses the piece of software inquestion.
AFAIK, the GNU Ada compiler is not too bad on eliminating unnecessary bounds checking (array bounds checking usually only adds a penalty of a few percent), so something can be done even in the GCC context.
BTW, I would expect that gcj-compiled Java applications start much more quickly than entire VMs. The comparison doesn't seem to take this into account.
On German Usenet (the de,* hierarchy), this is already common practice. In particular, these pseudo-attachments are used to fool OE users to believe that articles carry some kind of mail worm, without really using attachments (so that the posters keep to the letter of netiquette).
However, it doesn't seem to help much, quite a few people are still using Outlook Express. Other newsreaders such as Gnus display some of these pseudo-attachments as real ones, too. (And I don't think this is a bug, it's just built-in uudecode support.)
And Outlook Express has much more critical bugs, for example in quoted-printable handling together with quoting.
Newton's Law of Gravity always holds? Even at the scale of the solar system, there are violations of the Newton "Law" of Gravity.
In addition, most if not all so-called laws in modern physics are statistical. Claiming that they are not universally valid because of their statistical nature is highly misleading. They wouldn't be universally valid in the strict sense even if they were not statistical. The Laws of Physics have never been more than just approximations to reality, and not Truth itself (20th century physics has shown that for earlier centuries, and we shouldn't make the mistake that this phenomenon won't repeat).
Unfortunately, the article misses the main point: Germany has been trying to build a PKI for governmental use since 1997 or so (when legislation was passed to make documents carrying some types of digital signatures equivalent to paper documents).
However, the 1997 law features very high requirements for CAs and the actual implementations of digital signing. Partly because of the high security standards (which look good on paper, but fails in practice--a certified solution was successfully attacked by compromising the hosting general purpose computer), and partly because of incompatibilities, acceptance of this type of signatures was extremely low.
The new digital signature law introduces a new kind of digital signature with lower security standards, and which does not necessarily require additional hardware. Although this is less secure (key theft might be possible), this approach seems to be practical.
At the same time, the compatibility problems are addressed in the Sphinx framework, where KMail and GnuPG are enhanced so that they can exchange messages with other Sphinx-compatible clients.
If I'm not mistaken, the German federal government announced recently that it would promote the use of the low security digital signature in non-critical areas of the federal government. I think this is a good idea; even a digital signature based entirely on software (and not on some smart card which fully implements an assymmetric crypto algorithm) provides more authentication than a simple phone call, and certainly much more non-repudiation (even more than an oral consultation). And this time, the rollout might actually succeed, if the clients get ready soon.
Red Hat bought Cygnus, which was quite profitable in selling GCC support for embedded systems. Nowadays, there are several competitors in this sector, but if Red Hat ceases to exist (or AOL decides that it is no longer interested in embedded systems and the like), this could have a strong impact on the business of lots of companies.
Fortunately, since GCC is free software released under the GPL, these companies do not face artificial barriers when looking for a new source for support, but doing so would involve major changes in any case.
And this is just one example. Over the years, Red Hat piled up a substantial bit of knowledge and work force in the GNU/Linux sector, and AOL's decisions could have a huge impact on it.
If you want to use "PostScript", you need at least a trademark license.
Of course, Adobe PostScript is not just a language specification, it's also an interpreter, and for the latter, you certainly have to pay license fees (and for the standard fonts, too).
There are rumors that moving the IIS document root directory is quite complicated. Do you really have to change registry entries, without support from the GUI?
Yes, this is an interesting side effect. However, USB might result in additional latencies in comparison to PCI cards.
In addition, the specs on the web page do not mention if it is possible to sync to digital signal sources (and do all the processing with this signal rate). If all internal processing is locked to 96 kHz, the quality in the more useful modes is probably less than optimal.
... technology to use your old phone lines as fast link layer for IP, they start to sell you Voice Over IP solutions so that you can use your phone again.
(Yes, I know that they claim you can continue to use the same line for speach communication, and I hope that they don't implement this using Voice Over IP.)
If I recall correctly a lecture I heard back in 1995 or so presented research results according to which cancer cells, unlike other body cells, can exist (or reproduce? I dont recall.) without time limit in lab conditions. There seems to be no built-in time bomb for cancer cells.
So it shouldn't be too surprising if further evidence shows for strong links between the aging process and natural cancer prevention.
I guess they are kidding: 512 MB DDR RAM is nothing, even by today's standards. I guess people will hit the 4 GB limit on traditional x86 desktops even before the end of 2004.
There's a rule that today's hard disk capacities are RAM capacities in five to seven years. By this estimate, we're going to hit 4 GB during 2003, I suppose.
That's exactly the difference between the Open Source and Free Software Movements: Open Source supporters like Open Source because it's better in some way, but Free Software supporters like Free Software because it's free (i.e. libre).
Some people even think that freedom is more important than advancement of technology.
IP Multicasting is already availabe, and multicast-based services have worked reliably despite the load that was placed on general news content the few days following 2001-09-11, which is quite remarkable. (Well, IRC and Usenet kept working, too...)
Unfortunately, Joe Average does not demand multicasting support, so you have to look very closely in order to find an ISP which supports it. AFAIK, here in Germany, you can get multicast support almost everywhere, but of course at rates which are not affordable for personal use.
In theory, multicasting is very interesting for ISPs, too: you receive the traffic once and account it seperately for each customer. Unfortunately, multicasting requires quite an investment to get started, both in man hours and hardware (although most hardware nowadays supports multicasting, but maybe not in an optimal way).
But I think that most of us can agree that charging the schools to drum some secular morality into children's heads is a bad idea -- and that is what the article
proposes.
Maybe my cultural heritage is different than yours, but I don't have a big problem with that.
Schools always transport hidden messages, moral and political. The methods are often very subtle, perhaps not even employed consciously. I don't know what is taught at US schools, but you can only be absolutely neutral if you leave out the controversial topics (such as some parts of history, global politics, or contraceptives), and that's not good either.
On the other hand, I don't think such education programs will work, especially since we are dealing with an extremely abstract concept here.
So you think that moral/legal issues should be determined according to how they fit in with your petty agenda, is that right?
Not quite. Illegal copying is -- illegal. The fact that most illegal copying is not prosecuted doesn't change that. To me, it seems that a certain amount of copying (especially for home use) is tolerated by the software vendors. Otherwise, most people couldn't use the same software on their home PCs as they run at work. (In fact, previous Microsoft EULAs explicitly permitted to install the same copy of an office program (not operating system) both at work and at home.)
I believe that the current situation (copying proprietary software is usually illegal, but nevertheless common practice) misleds people about the importance of copyright law: they believe that it's irrelevant for their daily life. Most still have to make the experience that they want to share something interesting with their friends, but can't, because of copyright restrictions. If copyright was actually tightly enforced, people would start to see that it cuts directly into their daily life and start to oppose copyright laws, especially those which benefit publishers.
So where does the "moral education" of children into corporate-endorsed views fit in your agenda, hmmm?
The fact that copying is limited by law is not only a corporate-endorsed view, it's the truth: the law is quite clear -- and the law has been put into place by your elected representatives, not by the corporations.
Actually, I think those who copy software illegally because they need it for some reason and can afford to buy it should be punished to the extent possible under current law. Why? The availability of copyrighted software at no charge was and is one of the main causes the Free Software Movement does not reach the masses. Why do you need Free Software if you can get almost any software for free from your friends?
If people are forced to use what they can use legally, we would soon see a tremendous increase in manpower available in Free Software projects, and even if it's just users reporting bugs and making suggestions.
Slashdot Mirror
Some tasks simply need low-level features. If you want to use a high-level language most of the time, you have got several choices: You can write the low-level parts in a different, low-level language, you can try to outgess the compiler vendor and write source code which compiles to the required machine code using a specific compiler version (in many cases, this is the C approach), or you can use a high-level language which supports low-level programming. The third choice does not have to be the worst, especially if the low-level language features are clearly separated from the high-level ones (C fails miserably in this area).
Remember that many mission-critical computer systems are implemented in Ada, which offers a wide range of very low-level features (interrupt handling, representation clauses, unchecked conversion of objects, and so on). Unlike C, there is also explicit support for machine addresses and address arithmetic.
However, you should keep in mind that only certain types of security problems can be avoided automatically by choosing an appropriate programming language. Buffer overflow and format string bugs are found in almost all C software, but they are not the only cause of problems.
Aegis does not only deal with source code management, but can be used to enforce a development process which includes steps like testing and peer review.
You can implement the same thing using arch or CVS, of course, but Aegis offers much more structure in this regard.
And the argument that you would expect less security holes in GNU/Linux systems because they are not as widely deployed as Microsoft-Windows-based systems is not convincing at all. Security vulnerabilities are there even if hardly anybody uses the piece of software inquestion.
AFAIK, the GNU Ada compiler is not too bad on eliminating unnecessary bounds checking (array bounds checking usually only adds a penalty of a few percent), so something can be done even in the GCC context.
BTW, I would expect that gcj-compiled Java applications start much more quickly than entire VMs. The comparison doesn't seem to take this into account.
On German Usenet (the de,* hierarchy), this is already common practice. In particular, these pseudo-attachments are used to fool OE users to believe that articles carry some kind of mail worm, without really using attachments (so that the posters keep to the letter of netiquette).
However, it doesn't seem to help much, quite a few people are still using Outlook Express. Other newsreaders such as Gnus display some of these pseudo-attachments as real ones, too. (And I don't think this is a bug, it's just built-in uudecode support.)
And Outlook Express has much more critical bugs, for example in quoted-printable handling together with quoting.
Newton's Law of Gravity always holds? Even at the scale of the solar system, there are violations of the Newton "Law" of Gravity.
In addition, most if not all so-called laws in modern physics are statistical. Claiming that they are not universally valid because of their statistical nature is highly misleading. They wouldn't be universally valid in the strict sense even if they were not statistical. The Laws of Physics have never been more than just approximations to reality, and not Truth itself (20th century physics has shown that for earlier centuries, and we shouldn't make the mistake that this phenomenon won't repeat).
if my business plans didn't work out.
(Read the final paragraphs of the announcement. Why do they stress that they are solvent?)
Unfortunately, the article misses the main point: Germany has been trying to build a PKI for governmental use since 1997 or so (when legislation was passed to make documents carrying some types of digital signatures equivalent to paper documents).
However, the 1997 law features very high requirements for CAs and the actual implementations of digital signing. Partly because of the high security standards (which look good on paper, but fails in practice--a certified solution was successfully attacked by compromising the hosting general purpose computer), and partly because of incompatibilities, acceptance of this type of signatures was extremely low.
The new digital signature law introduces a new kind of digital signature with lower security standards, and which does not necessarily require additional hardware. Although this is less secure (key theft might be possible), this approach seems to be practical.
At the same time, the compatibility problems are addressed in the Sphinx framework, where KMail and GnuPG are enhanced so that they can exchange messages with other Sphinx-compatible clients.
If I'm not mistaken, the German federal government announced recently that it would promote the use of the low security digital signature in non-critical areas of the federal government. I think this is a good idea; even a digital signature based entirely on software (and not on some smart card which fully implements an assymmetric crypto algorithm) provides more authentication than a simple phone call, and certainly much more non-repudiation (even more than an oral consultation). And this time, the rollout might actually succeed, if the clients get ready soon.
Red Hat bought Cygnus, which was quite profitable in selling GCC support for embedded systems. Nowadays, there are several competitors in this sector, but if Red Hat ceases to exist (or AOL decides that it is no longer interested in embedded systems and the like), this could have a strong impact on the business of lots of companies.
Fortunately, since GCC is free software released under the GPL, these companies do not face artificial barriers when looking for a new source for support, but doing so would involve major changes in any case.
And this is just one example. Over the years, Red Hat piled up a substantial bit of knowledge and work force in the GNU/Linux sector, and AOL's decisions could have a huge impact on it.
If you want to use "PostScript", you need at least a trademark license.
Of course, Adobe PostScript is not just a language specification, it's also an interpreter, and for the latter, you certainly have to pay license fees (and for the standard fonts, too).
There are rumors that moving the IIS document root directory is quite complicated. Do you really have to change registry entries, without support from the GUI?
Have a look at TrustedPC. The technology is almost there.
And Word versioning is compatible with which configuration management systems?
Yes, this is an interesting side effect. However, USB might result in additional latencies in comparison to PCI cards.
In addition, the specs on the web page do not mention if it is possible to sync to digital signal sources (and do all the processing with this signal rate). If all internal processing is locked to 96 kHz, the quality in the more useful modes is probably less than optimal.
Unlike other, more popular software companies which try to hide that they are spying on their users, ClickTillUWin is open about its business.
If you install such software, it is simply your own fault.
VoIP might be cool technology, but I prefer to have a completely different communication channel over which people can report network problems.
... technology to use your old phone lines as fast link layer for IP, they start to sell you Voice Over IP solutions so that you can use your phone again.
(Yes, I know that they claim you can continue to use the same line for speach communication, and I hope that they don't implement this using Voice Over IP.)
If I recall correctly a lecture I heard back in 1995 or so presented research results according to which cancer cells, unlike other body cells, can exist (or reproduce? I dont recall.) without time limit in lab conditions. There seems to be no built-in time bomb for cancer cells.
So it shouldn't be too surprising if further evidence shows for strong links between the aging process and natural cancer prevention.
I guess they are kidding: 512 MB DDR RAM is nothing, even by today's standards. I guess people will hit the 4 GB limit on traditional x86 desktops even before the end of 2004.
There's a rule that today's hard disk capacities are RAM capacities in five to seven years. By this estimate, we're going to hit 4 GB during 2003, I suppose.
That's exactly the difference between the Open Source and Free Software Movements: Open Source supporters like Open Source because it's better in some way, but Free Software supporters like Free Software because it's free (i.e. libre).
Some people even think that freedom is more important than advancement of technology.
IP Multicasting is already availabe, and multicast-based services have worked reliably despite the load that was placed on general news content the few days following 2001-09-11, which is quite remarkable. (Well, IRC and Usenet kept working, too...)
Unfortunately, Joe Average does not demand multicasting support, so you have to look very closely in order to find an ISP which supports it. AFAIK, here in Germany, you can get multicast support almost everywhere, but of course at rates which are not affordable for personal use.
In theory, multicasting is very interesting for ISPs, too: you receive the traffic once and account it seperately for each customer. Unfortunately, multicasting requires quite an investment to get started, both in man hours and hardware (although most hardware nowadays supports multicasting, but maybe not in an optimal way).
Maybe my cultural heritage is different than yours, but I don't have a big problem with that. Schools always transport hidden messages, moral and political. The methods are often very subtle, perhaps not even employed consciously. I don't know what is taught at US schools, but you can only be absolutely neutral if you leave out the controversial topics (such as some parts of history, global politics, or contraceptives), and that's not good either.
On the other hand, I don't think such education programs will work, especially since we are dealing with an extremely abstract concept here.
Not quite. Illegal copying is -- illegal. The fact that most illegal copying is not prosecuted doesn't change that. To me, it seems that a certain amount of copying (especially for home use) is tolerated by the software vendors. Otherwise, most people couldn't use the same software on their home PCs as they run at work. (In fact, previous Microsoft EULAs explicitly permitted to install the same copy of an office program (not operating system) both at work and at home.)
I believe that the current situation (copying proprietary software is usually illegal, but nevertheless common practice) misleds people about the importance of copyright law: they believe that it's irrelevant for their daily life. Most still have to make the experience that they want to share something interesting with their friends, but can't, because of copyright restrictions. If copyright was actually tightly enforced, people would start to see that it cuts directly into their daily life and start to oppose copyright laws, especially those which benefit publishers.
So where does the "moral education" of children into corporate-endorsed views fit in your agenda, hmmm?
The fact that copying is limited by law is not only a corporate-endorsed view, it's the truth: the law is quite clear -- and the law has been put into place by your elected representatives, not by the corporations.
Of course. The laser printer for the early Apple Macintoshes had more MIPS than the computer itself.
Actually, I think those who copy software illegally because they need it for some reason and can afford to buy it should be punished to the extent possible under current law. Why? The availability of copyrighted software at no charge was and is one of the main causes the Free Software Movement does not reach the masses. Why do you need Free Software if you can get almost any software for free from your friends?
If people are forced to use what they can use legally, we would soon see a tremendous increase in manpower available in Free Software projects, and even if it's just users reporting bugs and making suggestions.