And can you give an example of anyone ever sueing Microsoft or Sun successfully when something went wrong with their COMMERCIAL software?
Bwa hahahahaha. I can't think of anything more futile.
Read one of the software licenses agreements for the commercial software you think you are so "not STUPID" to use. You will learn that they are not liable for anything.
THEY aren't stupid. They know that if they were liable for every time their software broke and people lost time and money, they would be sued into oblivion. So they specifically make you agree (through EULA's) that they are NOT liable.
The Linux-specific parts of Unreal Tournament will be open sourced. (Credit to Blue's News where I saw this - I thought it would be relevant here.)
From Brandon "GreenMarine" Reinhart's.plan:
I have permission from Tim to open source certain parts of the engine for the purposes of improving Linux support. What I would like to do is make available the code I have written for Glide, Mesa, input, and audio. I would appreciate Linux programmers taking the time to submit improvements to me at brandon@epicgames.com. I will look into putting up a webpage for this and perhaps a discussion forum. Please email me your input or ideas.
I'm leaving to go home for Thanksgiving tomorrow, so I may not be able to start on this until I get back on Friday. Right now I'm working on the Mesa renderer, because I would like to release that soon, if possible.
The code that I open source will not allow you to really dig very deep into the engine, but it will let you do a lot of stuff to the renderer and X windows interface.
Since my time is split between non-Linux related projects, the Linux port could no doubt be improved if it gets some attention from experienced Linux programmers.
This is excellent - it might also help other people porting games to Linux. An obvious question is, what open source license? The released code will have to be linked against the proprietary game engine. So I guess the GPL wouldn't work, right?
...Whose finger do you want on the ALT-CONTROL-DELETE button?
It's Control - Alt - Delete!
And what the heck is that supposed to mean, anyway? What does the president conceptually have the power to "reboot"? It sure isn't the economy, or any industry. Maybe education? Or maybe he's referring to "logging on", NT-style.
I bet some staffer just came up with a phrase that sounded sort of techno-power-cool and seemed appropriate for Microsoft (snicker), and didn't bother to define it, or even think very hard about it. And Gore just spouted it off.
After the 2.4 kernel comes out, and Redhat releases a distribution based on it, Redhat should get Mindcraft to do another set of benchmarks of Windows 2000 (which may be out by then) vs. the latest Redhat distro.
The hardware and benchmark configuration could be specified to something more realistic, but still high end - other posters here have described many ways this could be done. (Gigabet ethernet, dynamic content, stability testing, etc, etc.)
I presume Linux could win that handily. Now, as many people here are pointing out, benchmarks are kind of meaningless relative to day-to-day real life situations... but I'd still like to see the headline:
June 4, 2000 Redhat Press Release: "Mindcraft Benchmarks show Redhat Linux 7 faster, more stable than Windows 2000"
"In Mindcrafts recent testing of the latest releases of Windows 2000 and Redhat Linux, Linux was the fastest for for web serving and file serving... Redhat CEO thanks Linux developers world-wide... We're the best for servers, the desktop is next... "
What Microsoft really has a monopoly in is the implementation of the Win32 API's and the Microsoft Office file formats.
The best outcome of the trial would be to force them to publish complete, accurate specifications for all the Win32 API's and the MS Office file formats.
Then, the WINE project would be able to make much faster progress to making a real Windows competitor. And StarOffice, KOffice, etc. would be able to read and write MS Office documents.
At that point, it would become feasible to switch to Linux/BSD/BeOs and still be interoperable with the rest of the business world. Then, Microsoft would have to compete on the quality, features, and price of their software.
This reminds me of the movie "Strange Days", which involves full sensory recordings that could be played back (experienced) by anyone.
One thing that really stuck with me from that movie was peripheral to the plot.
The main character gave a birthday gift to a disabled, wheel-chair bound, hacker-type friend. It was a recording from the point of view of a man running along a beach with a beautiful girl, with the wind blowing and the waves rolling in, splashing over their feet...
I name all my machines after locations in Middle-Earth. Lorien, Rivendell, Mordor, Moria, Gondor, Arnor, Fangorn, Shire, Hobbiton, Bywater, Bree, Weathertop, Celebdil, Harad, Mirkwood, etc.
These are nice for a couple of reasons. Tolkien fans will never run out of names, (Check the index!) and they are mostly easy to pronounce. There's hundreds of Tolkien person-names, too. You could go with the names of the Valar: Manwe, Aule, Orome, Elbereth, Tulkas... or the names of elven-lords: Finwe, Elwe, Fingolfin, Finrod, Feanor, Celebrimbor..., or the names of the dwarfs in the Hobbit: Thorin, Fili, Kili, Bombir, Oin, Gloin, Balin, Dwalin...
You can probably tell I'm a Tolkien fan. Sadly, at work the NT servers are... Beavis and Butthead. Sigh.
Torrey (Azog) Hoffman.
(Azog was the orc who killed Thorin's grandfather - one of the very few named orcs in Tolkien's stories. I chose the name originally for playing Quake on line... I should have changed it when I signed up on Slashdot. Oh well.)
One-time pads are provably unbreakable, (and it's easy to prove).
Since the key is _random_ (and it has to be really random), you can get _anything_ out by changing the decryption key... So there's no way for an attacker to be sure they have guessed the right key when they get a message out that looks sensible.
IIRC, there are also ciphers for which breaking by means better than brute force would mean P != NP, but I don't remember the details on how they work.
The usual reason offered for the secrecy is that there are all sorts of tricks and techniques in the software that improve performance. This generally applies to OpenGL and Direct3D drivers. The best-performing ones are much more than just a thin layer over the hardware. They have a lot of tricky 3D code in them, and may have innovative techniques for using MMX, 3DNow, and the new PIII instruction sets.
These intial stages of the 3D graphics pipeline may not be closely tied to the hardware, and could be "stolen" by other companies for their drivers. The 3D accelerator market is so competitive that even small performance differences can make a difference in sales.
All the more reason to applaud and encourage the companies that open their specifications - especially if they open up all the code in their best Windows OpenGL drivers to be ported to Linux!
This was a fascinating article. I'm just disgusted by the whole situation, though. It's so obvious that patent law, at least for software, completely works against the original intent of the law.
But, it's good for big, rich, companies. So it's unlikely to be changed. Arrrgh.
Even worse is that it does apply to open-source software. I was previously unclear on that, but if I understand the article correctly, even if the programmer is anonymous or hasn't made any money directly off the software, people who use patent-infringing open-source software can be sued.
Yuck.
Torrey Hoffman (Azog)
Re:Q3Demo - Ibaibaibaibaiba...
on
No Next Q3Test
·
· Score: 1
Unplayable on a 300? What do you mean "unplayable???". I was playing it on a 64MB P166 (non-MMX) with a Voodoo 1. It was slow, sure, but I did a lot of playing against my roomate with it. He's not as good, so he got my 450 Mhz-256MB-Nvidia TNT machine. But I still beat him. Heh.
Since several people have posted on "optimality", without knowing exactly how Huffman coding works and how it is optimal, I thought I'd post an explanation. This is based on the chapter in "Introduction to Algorithms" by Corman, Leiserson, and Rivest. (An excellent undergrad textbook.)
First, definitions:
A character code is a code in which each symbol of input is translated to one symbol of output. (Compression is achieved in character codes by using variable length codes. Short code symbols are used for frequently occuring inputs.)
A prefix code is a code in which no symbol is a prefix of another symbol. For example, if you have "10" as a symbol, then you can't have "101" as a symbol in a prefix code. Prefix codes are nice because they greatly simplify encoding and decoding.
It is possible to prove that the optimal data compression achievable by a character code can always be achieved with a prefix code. Furthermore, it is possible to represent codes as a binary tree, in which a symbol "1" represents the left branch, and "0" the right branch. So "101" would be left, right, left.
It is easy to prove that an optimal code for any particular file is represented by a full binary tree, i.e. one in which each node has two leaves. (This is the first exercise of the chapter).
Now, Huffman coding is a greedy algorithm that constructs an optimal prefix code. The algorithm builds the tree T corresonding to the optimal code in a bottom-up manner.
Pseudocode: Assume C is a set of n characters, and each c in C is an object with a frequency f(c). A priority queue Q, keyed on f, is used to identify the two least-frequent objects to merge together. The result of the merger of two objects is a new object whose frequency is the sum of the frequencies of the two objects that were merged:
HUFFMAN(C) n = size of C Q = C for i = 1 to (n - 1) {
z = new node()
x = z.left = Q.ExtractMin() y = z.right = Q.ExtractMin() f(z) = f(x) + f(y) Q.Insert(z)
} return Q.ExtractMin()
And that's it!
To prove this is correct, prove that the problem of determining an optimal prefix code: (a) exhibits the greedy-choice property, and (b) exhibits the optimal substructure property
From these two lemmas, the theorem "Huffman coding produces es an optimal prefix code" is trivial to prove. The corollary is that Huffman encoding produces an optimal character code.
Torrey Hoffman (Azog)
Re:Proof that it's optimal? I don't believe that.
on
David Huffman is Dead
·
· Score: 1
What do you mean you don't believe the proof that it's optimal? You think there's an error in the proof? If you insist, I can dig out the proof and post it here, perhaps you can show us the error?
Remember, a proof in computer science is like a proof in math.
If you are just complaining that Huffman coding is not always the best compression method, well, nobody ever claimed it was.
What it is, is a greedy algorithm which constructs a provably optimal "prefix code". Furthermore, it is possible to show that the optimal compression achievable by a character code can always be achieved with a prefix code. A character code is one in which each byte (or other symbol) of the input is translated to a single symbol of output. The compression achieved in Huffman coding is optimal for character codes.
LZW, arithmetic coding, and other compression algorithms are not character codes.
Well, he actually does say why he thinks it's flawed. He called it censorship.
Now that's just dumb. The original Jane's article wasn't news, it was analysis. And it wasn't very good analysis. So Jane's chucked it and is writing a better analysis, with input from Slashdot posters.
This is not censorship. Cringley is completely wrong.
The interesting question: Is Cringely just ignorant, or was he intentionally misleading? If so, why?
One problem with your argument is that you have the hidden assumption that the money saved by not saving the life of the disabled baby would actually be used to prevent people from starving.
But of course it would not be. In our society, preventing people from starving is a much lower priority than about ten thousand stupid government programs. Tragic, but true.
I believe that when human life is involved you absolutely cannot just look at the bottom line. To do so is to start down a terrible path that ends with anyone who is "unproductive" being classified as worthless, and therefore being eliminated.
Can you not imagine the hell of a society where you live in fear every day of not being productive or "good" enough to justify your own life?
That's not how gravity and orbits work. I don't remember the details of Kepler's laws of orbital mechanics, but if you made Uranus lighter without changing it's speed it would fly away from the sun.
I think the most effective use of computer technology by terrorists would be to research vulnerable infrastructure links or nodes, which could then be attacked with conventional methods. This could cause dramatically more disruption than random attacks on infrastructure.
For example: terrorists who wanted to seriously mess up a city could determine what power lines supply the city, and attempt to bring down all of them at once. It is relatively easy to destroy a high-voltage transmission tower. Just destroying a few power lines at random might not have any effect on the population, as the system has redundancy.
The resulting chaos could be effectively used for cover for further operations. A relatively small group, say 50 people with inexpensive equipment and explosives could probably keep a large city completely disrupted for weeks by repeated, coordinated attacks on infrastructure.
Recent events demonstrate that a single backhoe can do a lot of damage to the internet, despite redundant links. What could terrorists do with a few dozen backhoes across the country, carefully orchestrated to take out as many redundant links as possible? If they had some detailed knowlege of how the internet was routed and where the fiber lines are, they could probably destroy coast-to-coast connectivity for at least a day or two. This is the kind of knowlege that "hackers" could probably aquire.
It would not be expensive, or difficult, or dangerous for terrorists to do these things. It is difficult to guard against. There are thousands of miles of fiber optic cable, clearly marked "Call before you dig!"
Commenting on the lack of female CSC profs, and the age at which the push away from CSC happens:
I'm a male CompSci graduate from the University of Victoria (in BC, Canada). Two years ago when I graduated, the Comp Sci faculty there had at least four women professors, out of a faculty of about 20. One was my thesis advisor. All of those women did quite a bit of teaching of first and second year classes, and were good at it. And, there were several excellent researchers among them.
(In fact, my entire thesis review committee was women - and that was just coincidental - they had the expertise in the subject area.)
Anyway, despite the excellent example of these profs, over the years from 1994 to 1997, the percentage of female computer science students at the university did not markedly increase. In fact, IIRC, it may have decreased.
Now, I think that it's great to encourage young women to get into comp sci and engineering, but usually this is done too late. The time to reach them is when they are much younger, in grades 1 to 5. That's when I decided I liked science and computers.
I believe there's a lot of cultural and peer-pressure "push" in those years for girls to do "girly" things. That's what needs to be counteracted. It will be difficult though, as it's up to elementary school teachers and parents. I suspect that that's where a lot of the problem lies. Some teachers are excellent, but others just maintain the status quo, and indoctrinate more than they educate.
Well, I'm running W2K Beta 3 at home on my Celeron 450 with 256 MB, and it's actually pretty nice. The install went very well, everything worked fine right away. The only problems I have are no drivers for my sound card, and no Open GL on my TNT card. So no Quake III under W2K for me.
But I did get Half-Life running with Direct 3D, (without sound) so it certainly is possible to play some games under Win2K.
I do quite a lot of work in there, mostly programming in Dev Studio 6, and it seems fast enough, although Dev Studio crashes more often than it did under NT 4.
However, I probably will go back to NT 4, just for OpenGL and sound, and a little more stability.
Torrey
Re:Integrated and Random M2 moderation
on
Moderation Ideas
·
· Score: 1
Your point about "manipulating history" is good and, I think, very important.
Leaving aside the discussion of M2, perhaps there should be a limit on moderating historical records, that is, no moderation at all for comments more than (say) 4 days old, and no M2 on moderations more than (say) 7 days old?
This would ensure that history doesn't get rewritten on the sly.
There definitely have been real problems with the GPS rollover, and not just with cheap car units.
My brother works for a local seaplane company, and about 9 of their GPS units quit working completely. These were all from one manufacturer - other units were fine. It's particularly annoying for them, because (being diligent and careful) they had specifically asked the manufacturer if the units were ok, and had been reassured that they were fine.
It took several weeks to get them all fixed. (An expensive software upgrade was required, and it took days to get through the busy phone lines to the company.) In the meantime, the planes were flying without working GPS, which is apparently ok since the weather was good. If there had been a need for navigation by instruments only, that might have been a worse problem.
Adding to the hassle, the software upgrade caused the units to forget all the named coordinate points that had been programmed into them. This will be hours of work to restore.
They won't be buying from that instrument manufacturer any more.
Don't underestimate the problems that these date issues can cause. Dates and times are the hardest fundamental data type to really handle correctly in software development, and are made worse by unpredictable users dealing with badly designed software.
I have an ABit BX6 version 2 running a Celeron 400 at 450 Mhz, which is a pretty conservative overclock.
It is very stable - at least, I have had no problems running Quake 2 and Unreal time demos for over 8 hours at a time. Those are fairly stressful system tests.
And, the extra 50 Mhz (plus bus speed at 75 instead of 66) gives a small but noticable speed improvement.
Slashdot Mirror
I just can't resist...
And can you give an example of anyone ever sueing Microsoft or Sun successfully when something went wrong with their COMMERCIAL software?
Bwa hahahahaha. I can't think of anything more futile.
Read one of the software licenses agreements for the commercial software you think you are so "not STUPID" to use. You will learn that they are not liable for anything.
THEY aren't stupid. They know that if they were liable for every time their software broke and people lost time and money, they would be sued into oblivion. So they specifically make you agree (through EULA's) that they are NOT liable.
So who is stupid here?
Why do I bother. It's only an anonymous coward.
Torrey Hoffman (Azog)
The Linux-specific parts of Unreal Tournament will be open sourced. (Credit to Blue's News where I saw this - I thought it would be relevant here.)
.plan:
From Brandon "GreenMarine" Reinhart's
I have permission from Tim to open source certain parts of the engine for the purposes of improving Linux support. What I would like to do is make available the code I have written for Glide, Mesa, input, and audio. I would appreciate Linux programmers taking the time to submit improvements to me at brandon@epicgames.com. I will look into putting up a webpage for this and perhaps a discussion forum. Please email me your input or ideas.
I'm leaving to go home for Thanksgiving tomorrow, so I may not be able to start on this until I get back on Friday. Right now I'm working on the Mesa renderer, because I would like to release that soon, if possible.
The code that I open source will not allow you to really dig very deep into the engine, but it will let you do a lot of stuff to the renderer and X windows interface.
Since my time is split between non-Linux related projects, the Linux port could no doubt be improved if it gets some attention from experienced Linux programmers.
This is excellent - it might also help other people porting games to Linux. An obvious question is, what open source license? The released code will have to be linked against the proprietary game engine. So I guess the GPL wouldn't work, right?
Azog
...Whose finger do you want on the ALT-CONTROL-DELETE button?
It's Control - Alt - Delete!
And what the heck is that supposed to mean, anyway? What does the president conceptually have the power to "reboot"? It sure isn't the economy, or any industry. Maybe education? Or maybe he's referring to "logging on", NT-style.
I bet some staffer just came up with a phrase that sounded sort of techno-power-cool and seemed appropriate for Microsoft (snicker), and didn't bother to define it, or even think very hard about it. And Gore just spouted it off.
Azog the Goblin
After the 2.4 kernel comes out, and Redhat releases a distribution based on it, Redhat should get Mindcraft to do another set of benchmarks of Windows 2000 (which may be out by then) vs. the latest Redhat distro.
The hardware and benchmark configuration could be specified to something more realistic, but still high end - other posters here have described many ways this could be done. (Gigabet ethernet, dynamic content, stability testing, etc, etc.)
I presume Linux could win that handily. Now, as many people here are pointing out, benchmarks are kind of meaningless relative to day-to-day real life situations... but I'd still like to see the headline:
June 4, 2000 Redhat Press Release: "Mindcraft Benchmarks show Redhat Linux 7 faster, more stable than Windows 2000"
"In Mindcrafts recent testing of the latest releases of Windows 2000 and Redhat Linux, Linux was the fastest for for web serving and file serving... Redhat CEO thanks Linux developers world-wide... We're the best for servers, the desktop is next... "
Ahhhh yes. Sell MSFT now, beat the rush.
Torrey Hoffman (Azog)
... factor primes ...
You know, of course, that factoring primes is really easy. Yeah, yeah, everyone knows you meant factor INTO primes - the hard problem.
So I'm being pedantic today.
Torrey Hoffman (Azog).
What Microsoft really has a monopoly in is the implementation of the Win32 API's and the Microsoft Office file formats.
The best outcome of the trial would be to force them to publish complete, accurate specifications for all the Win32 API's and the MS Office file formats.
Then, the WINE project would be able to make much faster progress to making a real Windows competitor. And StarOffice, KOffice, etc. would be able to read and write MS Office documents.
At that point, it would become feasible to switch to Linux/BSD/BeOs and still be interoperable with the rest of the business world. Then, Microsoft would have to compete on the quality, features, and price of their software.
Torrey Hoffman (Azog)
This reminds me of the movie "Strange Days", which involves full sensory recordings that could be played back (experienced) by anyone.
One thing that really stuck with me from that movie was peripheral to the plot.
The main character gave a birthday gift to a disabled, wheel-chair bound, hacker-type friend. It was a recording from the point of view of a man running along a beach with a beautiful girl, with the wind blowing and the waves rolling in, splashing over their feet...
Torrey Hoffman (Azog)
I name all my machines after locations in Middle-Earth. Lorien, Rivendell, Mordor, Moria, Gondor, Arnor, Fangorn, Shire, Hobbiton, Bywater, Bree, Weathertop, Celebdil, Harad, Mirkwood, etc.
...
These are nice for a couple of reasons. Tolkien fans will never run out of names, (Check the index!) and they are mostly easy to pronounce. There's hundreds of Tolkien person-names, too. You could go with the names of the Valar: Manwe, Aule, Orome, Elbereth, Tulkas... or the names of elven-lords: Finwe, Elwe, Fingolfin, Finrod, Feanor, Celebrimbor..., or the names of the dwarfs in the Hobbit: Thorin, Fili, Kili, Bombir, Oin, Gloin, Balin, Dwalin
You can probably tell I'm a Tolkien fan. Sadly, at work the NT servers are... Beavis and Butthead. Sigh.
Torrey (Azog) Hoffman.
(Azog was the orc who killed Thorin's grandfather - one of the very few named orcs in Tolkien's stories. I chose the name originally for playing Quake on line... I should have changed it when I signed up on Slashdot. Oh well.)
One-time pads are provably unbreakable, (and it's easy to prove).
Since the key is _random_ (and it has to be really random), you can get _anything_ out by changing the decryption key... So there's no way for an attacker to be sure they have guessed the right key when they get a message out that looks sensible.
IIRC, there are also ciphers for which breaking by means better than brute force would mean P != NP, but I don't remember the details on how they work.
Torrey (Azog)
The usual reason offered for the secrecy is that there are all sorts of tricks and techniques in the software that improve performance. This generally applies to OpenGL and Direct3D drivers. The best-performing ones are much more than just a thin layer over the hardware. They have a lot of tricky 3D code in them, and may have innovative techniques for using MMX, 3DNow, and the new PIII instruction sets.
These intial stages of the 3D graphics pipeline may not be closely tied to the hardware, and could be "stolen" by other companies for their drivers. The 3D accelerator market is so competitive that even small performance differences can make a difference in sales.
All the more reason to applaud and encourage the companies that open their specifications - especially if they open up all the code in their best Windows OpenGL drivers to be ported to Linux!
Torrey Hoffman (Azog)
This was a fascinating article. I'm just disgusted by the whole situation, though. It's so obvious that patent law, at least for software, completely works against the original intent of the law.
But, it's good for big, rich, companies. So it's unlikely to be changed. Arrrgh.
Even worse is that it does apply to open-source software. I was previously unclear on that, but if I understand the article correctly, even if the programmer is anonymous or hasn't made any money directly off the software, people who use patent-infringing open-source software can be sued.
Yuck.
Torrey Hoffman (Azog)
Unplayable on a 300? What do you mean "unplayable???". I was playing it on a 64MB P166 (non-MMX) with a Voodoo 1. It was slow, sure, but I did a lot of playing against my roomate with it. He's not as good, so he got my 450 Mhz-256MB-Nvidia TNT machine. But I still beat him. Heh.
Azog (Torrey Hoffman)
First, definitions:
A character code is a code in which each symbol of input is translated to one symbol of output. (Compression is achieved in character codes by using variable length codes. Short code symbols are used for frequently occuring inputs.)
A prefix code is a code in which no symbol is a prefix of another symbol. For example, if you have "10" as a symbol, then you can't have "101" as a symbol in a prefix code. Prefix codes are nice because they greatly simplify encoding and decoding.
It is possible to prove that the optimal data compression achievable by a character code can always be achieved with a prefix code. Furthermore, it is possible to represent codes as a binary tree, in which a symbol "1" represents the left branch, and "0" the right branch. So "101" would be left, right, left.
It is easy to prove that an optimal code for any particular file is represented by a full binary tree, i.e. one in which each node has two leaves. (This is the first exercise of the chapter).
Now, Huffman coding is a greedy algorithm that constructs an optimal prefix code. The algorithm builds the tree T corresonding to the optimal code in a bottom-up manner.
Pseudocode:
Assume C is a set of n characters, and each c in C is an object with a frequency f(c). A priority queue Q, keyed on f, is used to identify the two least-frequent objects to merge together. The result of the merger of two objects is a new object whose frequency is the sum of the frequencies of the two objects that were merged:
And that's it!
To prove this is correct, prove that the problem of determining an optimal prefix code: (a) exhibits the greedy-choice property, and (b) exhibits the optimal substructure property
From these two lemmas, the theorem
"Huffman coding produces es an optimal prefix code" is trivial to prove. The corollary is that Huffman encoding produces an optimal character code.
Torrey Hoffman (Azog)
What do you mean you don't believe the proof that it's optimal? You think there's an error in the proof? If you insist, I can dig out the proof and post it here, perhaps you can show us the error?
Remember, a proof in computer science is like a proof in math.
If you are just complaining that Huffman coding is not always the best compression method, well, nobody ever claimed it was.
What it is, is a greedy algorithm which constructs a provably optimal "prefix code". Furthermore, it is possible to show that the optimal compression achievable by a character code can always be achieved with a prefix code.
A character code is one in which each byte (or other symbol) of the input is translated to a single symbol of output. The compression achieved in Huffman coding is optimal for character codes.
LZW, arithmetic coding, and other compression algorithms are not character codes.
Torrey Hoffman (Azog)
Well, he actually does say why he thinks it's flawed. He called it censorship.
Now that's just dumb. The original Jane's article wasn't news, it was analysis. And it wasn't very good analysis. So Jane's chucked it and is writing a better analysis, with input from Slashdot posters.
This is not censorship. Cringley is completely wrong.
The interesting question: Is Cringely just ignorant, or was he intentionally misleading? If so, why?
Torrey Hoffman (Azog)
One problem with your argument is that you have the hidden assumption that the money saved by not saving the life of the disabled baby would actually be used to prevent people from starving.
But of course it would not be. In our society, preventing people from starving is a much lower priority than about ten thousand stupid government programs. Tragic, but true.
I believe that when human life is involved you absolutely cannot just look at the bottom line. To do so is to start down a terrible path that ends with anyone who is "unproductive" being classified as worthless, and therefore being eliminated.
Can you not imagine the hell of a society where you live in fear every day of not being productive or "good" enough to justify your own life?
Torrey Hoffman (Azog)
Hmm. So would it be wrong to kill YOU?
If your only natural right is to die, why not? Your killer would only be depriving YOU of a bonus to which YOU have no natural entitlement anyway.
According to your theory, anyway.
But your theory is wrong. It would be wrong to kill you. It is wrong to kill babies.
Torrey Hoffman (Azog)
That's not how gravity and orbits work. I don't remember the details of Kepler's laws of orbital mechanics, but if you made Uranus lighter without changing it's speed it would fly away from the sun.
I think the most effective use of computer technology by terrorists would be to research vulnerable infrastructure links or nodes, which could then be attacked with conventional methods. This could cause dramatically more disruption than random attacks on infrastructure.
For example: terrorists who wanted to seriously mess up a city could determine what power lines supply the city, and attempt to bring down all of them at once. It is relatively easy to destroy a high-voltage transmission tower. Just destroying a few power lines at random might not have any effect on the population, as the system has redundancy.
The resulting chaos could be effectively used for cover for further operations. A relatively small group, say 50 people with inexpensive equipment and explosives could probably keep a large city completely disrupted for weeks by repeated, coordinated attacks on infrastructure.
Recent events demonstrate that a single backhoe can do a lot of damage to the internet, despite redundant links. What could terrorists do with a few dozen backhoes across the country, carefully orchestrated to take out as many redundant links as possible? If they had some detailed knowlege of how the internet was routed and where the fiber lines are, they could probably destroy coast-to-coast connectivity for at least a day or two. This is the kind of knowlege that "hackers" could probably aquire.
It would not be expensive, or difficult, or dangerous for terrorists to do these things. It is difficult to guard against. There are thousands of miles of fiber optic cable, clearly marked "Call before you dig!"
Torrey Hoffman (Azog)
Commenting on the lack of female CSC profs, and the age at which the push away from CSC happens:
I'm a male CompSci graduate from the University of Victoria (in BC, Canada). Two years ago when I graduated, the Comp Sci faculty there had at least four women professors, out of a faculty of about 20. One was my thesis advisor. All of those women did quite a bit of teaching of first and second year classes, and were good at it. And, there were several excellent researchers among them.
(In fact, my entire thesis review committee was women - and that was just coincidental - they had the expertise in the subject area.)
Anyway, despite the excellent example of these profs, over the years from 1994 to 1997, the percentage of female computer science students at the university did not markedly increase. In fact, IIRC, it may have decreased.
Now, I think that it's great to encourage young women to get into comp sci and engineering, but usually this is done too late. The time to reach them is when they are much younger, in grades 1 to 5. That's when I decided I liked science and computers.
I believe there's a lot of cultural and peer-pressure "push" in those years for girls to do "girly" things. That's what needs to be counteracted. It will be difficult though, as it's up to elementary school teachers and parents. I suspect that that's where a lot of the problem lies. Some teachers are excellent, but others just maintain the status quo, and indoctrinate more than they educate.
Torrey Hoffman (Azog)
Well, I'm running W2K Beta 3 at home on my Celeron 450 with 256 MB, and it's actually pretty nice. The install went very well, everything worked fine right away. The only problems I have are no drivers for my sound card, and no Open GL on my TNT card. So no Quake III under W2K for me.
But I did get Half-Life running with Direct 3D, (without sound) so it certainly is possible to play some games under Win2K.
I do quite a lot of work in there, mostly programming in Dev Studio 6, and it seems fast enough, although Dev Studio crashes more often than it did under NT 4.
However, I probably will go back to NT 4, just for OpenGL and sound, and a little more stability.
Torrey
Your point about "manipulating history" is good and, I think, very important.
Leaving aside the discussion of M2, perhaps there should be a limit on moderating historical records, that is, no moderation at all for comments more than (say) 4 days old, and no M2 on moderations more than (say) 7 days old?
This would ensure that history doesn't get rewritten on the sly.
Torrey (Azog) Hoffman
There definitely have been real problems with the GPS rollover, and not just with cheap car units.
My brother works for a local seaplane company, and about 9 of their GPS units quit working completely. These were all from one manufacturer - other units were fine. It's particularly annoying for them, because (being diligent and careful) they had specifically asked the manufacturer if the units were ok, and had been reassured that they were fine.
It took several weeks to get them all fixed. (An expensive software upgrade was required, and it took days to get through the busy phone lines to the company.) In the meantime, the planes were flying without working GPS, which is apparently ok since the weather was good. If there had been a need for navigation by instruments only, that might have been a worse problem.
Adding to the hassle, the software upgrade caused the units to forget all the named coordinate points that had been programmed into them. This will be hours of work to restore.
They won't be buying from that instrument manufacturer any more.
Don't underestimate the problems that these date issues can cause. Dates and times are the hardest fundamental data type to really handle correctly in software development, and are made worse by unpredictable users dealing with badly designed software.
Yeah, there's lots of interesting tidbits.
For example, Apple:
1. Apple Confidential - the real story of Apple Inc.
2. Apple: The inside story of intrigue, egomania, and business blunders.
Azog
I have an ABit BX6 version 2 running a Celeron 400 at 450 Mhz, which is a pretty conservative overclock.
It is very stable - at least, I have had no problems running Quake 2 and Unreal time demos for over 8 hours at a time. Those are fairly stressful system tests.
And, the extra 50 Mhz (plus bus speed at 75 instead of 66) gives a small but noticable speed improvement.
Azog