> I believe that anyone capable of organizing the hijacking of several commerical airplanes and using them as weapons against both civillian and military targets in an unprovoked sneak attack is not someone that I can make amends with...
Presumably you would also have trouble living in a country with a proven track record of murdering the leaders of other countries to further its own interests, installing and supporting corrupt murderous regimes in numerous countries, killing hundreds of thousands of innocent civilians (Hiroshima/Nagasaki), allowing the slaughter and oppression of the palestinians to continue with no oppositio, and which supported the radical Islamists in Afghanistan in order to give the Russians a hard time. Yes the good ol' US of A.
Desperate people - people with nothing to lose - will do desperate, horrible things. The only answer to to give them some hope, something that for example the Palestinians do not have.
As you will recall, they have been thrown out of the land they lived in for thousands of years, reprived of the vote, their houses confiscated or destroyed.
America does what is in the interests of the wealthy people in America (eg DMCA). What is right and moral comes a very poor second. They preach free trade but have the biggest farm subsidies and import restrictions in the world. America did not honor copyright until it was in its interests. However they are forcing poor contries to honor intellectual property laws, even when this means that millions will die because they cannot afford patented medicines.
I'm sorry but, no, the whole world does not necessarily see the US as the good guys.
"Even though there are some fairly valid reasons as to why, its still fairly egotistical of him - did he ask for a consensus of all the developers releasing "GNU Software"? Does his own technical work make up a large slice of the GNU works used by linux? [No, Emacs does not count as a large slice, despite its footprint.;)]"
Richard Stallman did not 'just' write emacs. Among his other contributions are gcc, gdb and texinfo. There are many more. He also wrote a ton of documentation as well as the GPL and LGPL.
Without Richard Stallman 'Linux' would not exist.
I have met him and I found he expresses his views firmly but he is not a 'raving loony' at all. Yes he is at times unreasonable but I think only to the extent that nothing great was ever accomplished by reasonable men, and a lot is at stake.
Definitely not the person to take to see the PHB though.
Even more relaxed lobsters and nicer food if you float the lobsters in wine until they become unconscious. We did this once and the results were excellent
The basic problem with AI machines is lack of CPU power. To explain....
The human brain has 100,000,000,000 neurons. Each neuron has an average of 1,000 connections to other neurons. It is probably most accurate to think of the connections as being the active components. The cycle time is 30ms.
So you can do about 3e15 computations per second. The typical new desktop these days is about 3e9 instructions per second. So the gap is a factor of about 1,000,000. This gap should close in about 20-30 years, assuming Moore's law continues.
I know this is rough but it does give the flavour of the problem.
Current computers have the processing power of an insect brain, and they are mostly about that smart.
I think computer scientists have done pretty well given the lack of CPU power that is available. A lot of things that computers have trouble with such as vision procession are handled in the brain by brute force - many neurons in parallel doing lots of simple repetitive things.
Assuming Moore's law continues, we are going to see a dramatic closing of the gap between silicon and carbon based intelligence in our life times.
AI is the field to be in over the next few years. Having the processing power is not enough. There will be many theoretical questions to answer before we can build truly smart machines.
1. There is nothing that ensure that even an analog recording will be possible into the future, as a way of delivering fair use. For example watermarking technology could be used to mark content and the studios could strong arm the manufacturers to stop making such cameras. Such cameras are, after all, a circumvention device - a mechanism exists to prevent copying and the cameras bypass this (by ignoring the watermark).
You might say this is hypothetical, but no more hypothetical than any piracy that DeCSS is supposed to allow.
So the 'take a video of the screen' argument does not answer the question.
2. Regional encoding. I buy a DVD while working in the Europe and take it home. I cannot play it. The studios argue that you can buy a second DVD player that works in Europe. First, this is an argument only a wealthy person would make. Second, due to differences in voltages and TV formats (eg NTSC vs PAL) this is actually quite difficult, and invariably results in a far worse picture.
3. I do music analysis, and a lot of researchers do this. This requires access to high quality digitized music. Music is headed for DVDs - don't think you are going to be able to buy CDs for long! Without a way to bypass the encryption, I cannot do this, and I will have to try and tap the analog signal at some point. This means worse quality signal and more dificulties analyzing the music. Again, the technology is moving to a system where all components will incorporate the protections mechanisms and it will only be possible to record using a microphone in front of the speakers. Future use of watermarking technology and recorders that recognize the watermarking may even make it impossible to take a recording. See (1) above.
C, the language DeCSS is written in, is a derivative of Algol. Algol (algorithmically oriented language) was originally developed to express algorithms in precise form. It was only later that peoiple thought - hey, you could write a compiler for this and use it for programming.
If coding is not expressive, why include comments to clarify meaning? Why have meaningful variable names? Why not just publish object code?
> No, it doesn't leave permanent damage. There's a way to remove scars: it's called maturity.
This reminds me of those inane arguments you used to hear about corporal punishment - "well it never did me any harm". Or the arguments that smoking must be harmless because Uncle Fred smoked and lived to 80.
Yes a lot of people are bullied. Yes most people don't go beserk. But not everyone can cope with dialy insults, threats and humiliation. And not all bullying is alike - it goes way beyond 'a bit of teasing'. And while most people may be able to get on with your lives some are less resiliant.
Adults do not have to put up with this kind of behaviour and nor should children.
Why aren't they suing the school for creating an environment where people felt so hopeless, so worthless, that they had nothing to lose by going on a murderous rampage.
> 1.Click on paperclip to bring up a dialog balloon
> 2.Click on "options" button
> 3.Uncheck "Use office assistant" box.
Yes, but how long does it take to work this out. This information is quite hard to find, unless you ask the paperclip.
Actually the paperclip is very useful, because you can ask it questions in plain English and mostly get useful answers.
The problem was it popped up too often
And this is the one that really annoyed me: when you clocked on it to go away, it would wait a few seconds, then wink, then eventually disappear. If it was a person I would have punched its lights out.
"- credit creation - the unbundling of the currency from any tangible asset such as the gold standard".
I agree with the rest but this doesn't make sense.
The first problem with credit creation is that here is a constant risk that governments will print money and use inflation to pay off their debts or to finance unsustainable expenditures.
We saw this in the 1970s and in the 1920s. The cost of the uncertainty this creates is enormous - see how even the slightest sign of inflation 'spooks' financial markets.
The second problem is incompetent management of the money supply. This was the actual cause of the great depression. They went off the gold standard and congratulated themselves how clever they were creating the unsustainable 1920s boom by printing money. Eventually the party ended and the stock market crashed. During the next year or so, the money supply fell by 60%. If you do that, you will cause a depression. This is all documented at length in Milton Friedman's book on the monetary history of the US.
Governments have mostly learned how to manage money supply so this problem rarely occurs these days.
People often assume that the 1930s depresssion was a product of free markets, when in reality it was due to incompetent management of regulated money supplies by newbies who didn't know what they were doing.
While the gold standard was in place there were booms and busts, but the busts were usually short lived and the rate of economic growth was far higher than it is now. When you graph the stock market before and after going off the gold standard, there is a knee in the curve - growth is much slower afterwards.
One comment on the book - it makes the very good point that secure title to property is key to economic growth. Having visited a number of third world countries I can confirm this is a huge issue. But it is just a part of the wider problem of the rule of law. When every government decision is dictated by 'connections' or bribes, and when it is impossible to compete with those who have the right connections due to government harrassment, you end up with inefficient entrenched monopolies.
Ghandi saw this problem; in fact his first big protest in India was against a government-mandated monopoly in the production of salt.
The USA is not immune from these problems. Look at the actions of the recording industry to extend their control over consumers and to extent the reach and duration of copyright. The fact that you need tens of millions of dollars to get into congress delivers a system that gives those who are already wealthy and organised what they want.
The USA was very slow to crack down on tobacco companies for this reason. Other countries took effective action a long time ago.
I often wonder whether those in authority really want to stop the illegal drug trade. The economist had a graph recently that showed that heroin prices have been in almost constant decline for 15 years. "The way on drugs is going well and we will succeed". Right, like you won the war in Vietnam.
The article states they they have been able to break the crypto. I personally assume that since they removed the export barriers on 128 bit crypto they can break it.
I suspect that all this carry on is to make people think they are safe to use PGP, when in reality they probably are not. So people will use PGP and get their messages decoded.
The bottom line is if those people want to find out what you are saying they will. Even if they have to park a van out the front of your house and watch your screen and read your mouse clicks.
One can also assume, based on known past practices, that the US uses their crypto capabilities to obgain commercial advantage for US companies by feeding them information obtained from cracking secret commmunications.
Not everyone speeds. Inexperienced drivers and those who do not think, speed.
Most young people vastly overestimate their driving ability. After a few close calls they often modify their behaviour. The road tool kills huge numbers of people every year. Speed is a major factor, because it reduces the time you have to react and it increases the damage done when you hit. Don't forget the energy is proportional to the square of the speed.
A lot of the posters seem to assume that the system needs to be perfect for it to be useful. In fact the system could just turn itself off when it doesn't know about a road.
A possible alternative to the 17 CDs and the GPS would be a simple system of bar codes on the roads with a bar code reader under every car. I would welcome such a system, because it is often hard to keep track of the changing limits as you drive along, and this system would pick up the changes automatically. You just set the cruise control to speed limit - 2.
One point I think a lot of people are missing is the phenomenon of 'worse is better'.
When Unix replaced some mainframes, the mainframers pointed out that Unix lacked some features of mainframes. The same when NT replaced some commercial Unix systems. The same when Linux started replacing some NT systems.
It didn't stop the replacements happening. Mainly because the replacement was cheaper.
The bottom line is that if the replacement is cheaper and good enough, then it will win.
Mainframe revenue is falling, Unix vendors are vanishing rapidly. Microsoft's share price has halved. All because of competition from systems that are not as good in some respects.
Linux does not have to be better than NT in every respect to win. It just needs to be good enough.
Over time the gaps will be filled, by free software or commercial products. System monitoring tools, job scheduling etc.
> Yeah, right. Ok, lets address stuff in order:...
> Sure the Pentium 4 doesn't perform great on code not optimized for it. But neither did the 486, the Pentium, or the Pentium Pro. And which would you prefer to have right now, a 250MHz 386, or a 1GHz Pentium III?
Well actually the P4 is the first X86 from Intel that is slower on the targetted workload. The P-Pro was slower on 16 bit code but it was faster on 32 bit code.
Running Quake faster is not good enough - not everyone runs Quake. All the stuff I do will be slower (compilations etc).
It's all very good to crtiicise compiler writers for not optimizing for the latest CPU. The problem here is that in the past, you have not had to severely penalize past CPUs to write code that runs better on new CPUs. But that is just what Intel has done this time. The other problem is that it is not simple to optimize for the P4. The optimization manual is 300 pages and the rules are very complex. It is not reasonable to expect everyone to rewrite the code generation every time. The GCC code generation is hundreds of thousands of lines of code. Anyone who thinks it is easy to implement Intel's suggestions should submit a patch to the gcc patches mailing list.
This was indeed a major point of the article, that AMD and Transmeta do not require code rewrites to run well. But Intel does for the P4. I am working on a compiler and my attitude at the moment is that the P4 is not worth the effort at this stage.
Looking at Intel's design decisions on the P4 really makes me scratch my head. The business with the shift instructions seems particularly incredible. I have been reading and these optimization guides since the 286 and I am at a loss.
There seem to be three tenable theories.
The megahertz theory. The trouble with this is that to even match the Athlon you need such high clock speeds it is not going to solve the problem for quite a while.
Theory two is that Intel has put multithreading onto the CPU at the cost of a lot of silicon, but hasn't been able to get it going yet. This would explain why they have been so stingy with silicon where it could have been useful. The multithreading could double system throughput by running two threads at the same time on the same chip. That would justify the design decisions made, but this theory is a long shot.
Theory three is that Intel has been taken over by marketing pukes and is doomed. Engineering realities come a distant tenth to PR, Hype etc. Check out http://www.faceintel.com/ for some information about what Intel is like from the inside. This seems the most plausable theory at this stage. Intel has had a long series of debacles so it is not surprising to see another one.
One reasonably sucessful Author told me she makes more from library royalties than from book sales. This is in Australia where libraries do pay royalties.
Here you also have to pay royalties if you play a radio in your store or workshop.
If someone working for me advanced this argument ie "I want a way to avoid taking responsibility", I would fire them.
In fifteen years of working in enterprises, I found very few companies actually provide any worthwhile support. IBM is one, but it is one of few. You can get Linux support from IBM I believe.
Otherwise it is just the usual "retry, reboot, reinstall, upgrade" routine. This is not just from tinpot outfits either. Everyone knows what Microsoft support is like, but they are not unusual in this regard.
The typical support organisation consists of a call centre staffed by people following scripts, plus a large number of marketing types. Actual people who can help solve your problem are thin on the ground in many cases.
I am not exaggerating. MSFT spent more money marketing Windows 95 than developing it. Progress Corp spends 250% as much on marketing as on product development.
Often when you find a problem, the vendor will put out a doc change, saying that you shouldn't do whatever triggered the bug, or in other ways say it is expected behaviour. Then they redefine your bug as an enhancement request. Or they do what MSFT did with multilevel numbering in MS Word 97. Remove the feature and pretend it never existed (I am not talking about multilevel heading numbers).
Another thing they do is ask 'Is patch X on'. If not, even though it does not look like your problem, the ball is back in your court. Another good ploy is to ask for impossible amounts of diagnostic information. "Delete records from the database until the problem goes away and then send us a dump of that record", for example.
For hearing this nonsense, you are paying through the nose for a 'support' contract.
The other thing they do is have very tight upgrade requirements. Older versions are not supported and you are *really* on your own, unless you keep to a breakneck and very expensive upgrade schedule. Up the upgrade creek without the source code, in fact. Upgrading is not a simple matter. You may have a large number of products with complex version interdependencies and OS and hardware dependencies. Sometimes you can't get there from here.
In a live situation it is almost always up to you to find a workaround or bypass. Stop using certain features, reboot every x hours, simplify the system, filter out errors that cause the software to choke, hone your recovery procedures. Swap out hardware until you find the bad component.
I have been in the situation more than once where closed source vendors tell you "We can't [or won't] fix it and we won't let you fix it" (by giving you the source code). What do you do then? In my experience someone within the organisation carries the blame. VPs are very uninterested in hearing that xxx Corporation (NASDAQ XXX) is to blame. They want *you* to fix it *now*.
More than one closed source vendor has gone bankrupt. Then you find they forgot to send the source to the escrow agency. Or they just lose the source code without going bankrupt. Then you get to reverse engineer the file formats, which is tedious after the first few times.
If they don't go bankrupt they get taken over. Then they jack up the maintenance costs by 375%, or the previously strategic product becomes unsupported, you are frog marched to the Gigantic Software Corporation's often inferior product suite, and all your investment in training and product is wasted. You can't read your old files any more.
Some things to look for in a support organisation:
1. A searchable bug and knowledge base.
2. Access to techos, as opposed to Marketing people. Job titles can help but you need to speak to them.
3. Knowledgeable, experienced technical support people
4. Technical support people who are not snowed under. Find out the numbers.
5. A number you can call at any time and they will answer and be able to get someone to help.
6. Clear definitions of problem categories and levels and specific service levels for a) restoration of service and b) provision of a fix for the problem for the various categories.
User groups are very important. Often you will be told "Noone else has this problem so it must be due to something you are doing wrong". Often this is a lie. The vendor must not be able to censor your mailing lists.
1. The 64k segment limit on the 80286. A disaster that held PC back for years.
2. The 432 CPU. Again, a failure partly because it had a 64k segment limit.
3. The overly complex protection model on the 80286 and successors. Made writing OSs and compilers more complicated, apart from wasting silicon.
4. The horrible 'not quite a stack' architecture of the x87 numeric processor. Slows things down by creating a bottleneck at top fo stack and makes writing numeric code or compilers very difficult.
5. Lack of full hardware support for virtualizing protected mode operating systems. Requires cooperation from the target OS to virtualize it. Whereas eg in system 390 the target has a hard time even knowing it is virtualized.
I run a site for a rare eye condition called central serous retinopathy. The URL is http://www.geocities.com/timjosling/csr.html. The site exists because official sources of information are limited and are in some cases marred by conflicts of interest. The conflicts of interest arise because clinics are pushing treatments that have been shown to be ineffective but which are good money earners. There is some speculative information on the site which I label as such.
My site has been instrumental in at least two cases in preventing people being treated for CSR with the very thing - cortisol - which causes it. Because it is a rare condition, incompetent treatment is not uncommon.
I am adamantly opposed to this proposal. The WHO should do what everyone else does - put up their web pages, and earn their reputation. Instead what they are proposing is that they become gatekeepers to all health information on the internet. I don't object if they want to create a system whereby you can have a 'WHO approved' logo on your site, but putting one highly bureacratic organisation into such a powerful position is a recipe for disaster.
This is not in line with the internet model which is that people should be able to decide for themselves what they believe. Anyone who believes that officials have their own best interests at heart and that they have no conflicts of interest is naive. This system would also introduce huge delays in getting information online.
The problem is that the apps that are out there do not need the latest and greatest CPU any more.
Once upon a time every time you upgraded you would get heaps more done - I remember when I upgraded to an 11mhz AT - INCREDIBLE!!! You even had to upgrade often just to run something at all. For a long time more CPU was needed to make GUI fast enough - but even unaccelerated X is fast enough on a 333mhz Celeron X.
There are a few exceptions, but even speech recognition does not really benefit enough to pay through the nose for 1ghz.
I can compile GCC in 18 minutes from scratch - that's fast enough.
This hurts MSFT too because a lot of their sales are bundled with new PCs.
At some point there will be a killer app that needs 2.9ghz or whatever. But there is no killer app that needs 1ghz right now. There are lots of things that are not there due to lack of CPU, but they need heaps more than 1ghz - good speech recognition in a noisy office; decoding music.au/wav files into musical scores, a useful version of office assistant and the like - all the AI stuff. It will be a while I think.
Also we are more bandwidth constrained than anything, rather than CPU bound.
> Custom written software like that is one instance where you *can* sue the people who wrote it if it fails, and...
It is very difficult to specify software to the degree of detail where you can actually sue someone. Specs change during development anyway.
Suing someone rarely does any good. The software vendors are experts at this - it is a core competency to negotiate contracts with plenty of holes in them that aren't apparent at the time.
And of course the PHB is breathing down your neck wanting to get started and can't understand why you are being so pedantic about the contract. The sales guy is telling him they are keen to go but 'you man' is holding things up.
The situation is not much better than trying to sue MSFT when powerpoint has a bug.
A good vendor will let you have source code. If not it is a good sign you are about to be done over. At a minimum you should have an escrow agreement so that they place the source at a trusted thrid party so it they go bust or - worse - get taken over by CA - you are not left twisting in the wind.
Finally it is no compensation to sue someone if your plane just crashed. (Airline safety briefing for busy people: If the plane crashes you will die).
All this goes to show that the US congress is a wholly owned subsidiary of big business, just in case anyone hadn't realized. Clinton signed the previous H1B bill just before a fundraising trip to Silicon Valley.
The DMCA Act is another example of the same thing. In fact if you look at the history of copyright and patent law big business has been getting what it wants for 100 years. Since, in fact, the US became an exporter ot intellectual property. Before that the US did not honor intellectual property just as 3rd world countries do not today. Now is is regularly extending the scope of it.
H1B is about driving down the salaries of programmers. As many posters have pointed out H1Bs are in many ways like indentured servants and are in a very weak and vulnerable position.
I know several people who went over (from AU) as H1Bs and for most it was an unhappy experience. Things were not as they were promised and there was nothing they could do about it.
Several people argued that if we do not allow the H1Bs the jobs will move offshore. As someone who works for a multinational corporation I can tell you that running projects across time zones is a nightmare. This is a huge barrier to exporting these jobs. So many companies tried to do this with Y2K work and ended up doing it in house because of the poor results.
A much better way is to have a skills based immigration scheme and allow the required people to get green cards day one. This avoids the 'indentured servant' problem at least and employers will have to pay market rates.
I worked in the US for a while and I was really surprised at the low salaries that people were earning. Especially considering the high cost of living in the US (housing, fresh food, restaurants, high heating bills in the north and cooling bills in the south).
I've seen email introduced to several large organisations, and every time there was a lot of flaming at first. I even used to do it myself.
People don't normally flame each other in person. Sometimes they do on the phone and often on email. Although maybe they want to - if you look at the soap operas they are all people telling each other exactly what they thing of them - exactly what you can't do in real life.
I think the reason for the email flames is that with email, and to some extent with the phone there is a lack of feedback.
You write 'I am disappointed that you did not accept my patch'. No (real time) response.
In normal human interactions if you say something like that, the person looks concerned and responds. With email, any response is much later, if at all. In real life, if the person does not respond you increase the voltage - you might speak a little louder or use body language to get the point across. This is all difficult over email.
So people flame. After they get used to the medium it dies down.
In any deal like this, you need to allow for a lot more management overhead than with an in house project.
Some things that can help
- project will work best if the nature of the work is that requirements are clearly defined up front and it is easy to determine if they have succeeded.
- need to have *frequent* clearly defined deliverables ie weekly. Otherwise you will have no idea how they are going and you will find the project will suddenly slip months in one day.
- require proof of testing. Otherwise they will not test it.
- regular on site visits. You can initiate these at the start of the project, or later on when things go off the rails, as you prefer. There is no substitute for seeing the atmosphere at their place. It's best to have someone there all the time. That way they get used to having you there and you find out what's really going on.
You cannot outsource responsibility. You still have to manage the whole process very tightly. We do design reviews - no coding allowed before the design is approved, and we do code reviews by phone to enforce coding standards and documentation standards.
Slashdot Mirror
> I believe that anyone capable of organizing the hijacking of several commerical airplanes and using them as weapons against both civillian and military targets in an unprovoked sneak attack is not someone that I can make amends with ...
Presumably you would also have trouble living in a country with a proven track record of murdering the leaders of other countries to further its own interests, installing and supporting corrupt murderous regimes in numerous countries, killing hundreds of thousands of innocent civilians (Hiroshima/Nagasaki), allowing the slaughter and oppression of the palestinians to continue with no oppositio, and which supported the radical Islamists in Afghanistan in order to give the Russians a hard time. Yes the good ol' US of A.
Desperate people - people with nothing to lose - will do desperate, horrible things. The only answer to to give them some hope, something that for example the Palestinians do not have.
As you will recall, they have been thrown out of the land they lived in for thousands of years, reprived of the vote, their houses confiscated or destroyed.
America does what is in the interests of the wealthy people in America (eg DMCA). What is right and moral comes a very poor second. They preach free trade but have the biggest farm subsidies and import restrictions in the world. America did not honor copyright until it was in its interests. However they are forcing poor contries to honor intellectual property laws, even when this means that millions will die because they cannot afford patented medicines.
I'm sorry but, no, the whole world does not necessarily see the US as the good guys.
"Even though there are some fairly valid reasons as to why, its still fairly egotistical of him - did he ask for a consensus of all the developers releasing "GNU Software"? Does his own technical work make up a large slice of the GNU works used by linux? [No, Emacs does not count as a large slice, despite its footprint. ;)]"
Richard Stallman did not 'just' write emacs. Among his other contributions are gcc, gdb and texinfo. There are many more. He also wrote a ton of documentation as well as the GPL and LGPL.
Without Richard Stallman 'Linux' would not exist.
I have met him and I found he expresses his views firmly but he is not a 'raving loony' at all. Yes he is at times unreasonable but I think only to the extent that nothing great was ever accomplished by reasonable men, and a lot is at stake.
Definitely not the person to take to see the PHB though.
Even more relaxed lobsters and nicer food if you float the lobsters in wine until they become unconscious. We did this once and the results were excellent
The basic problem with AI machines is lack of CPU power. To explain....
The human brain has 100,000,000,000 neurons. Each neuron has an average of 1,000 connections to other neurons. It is probably most accurate to think of the connections as being the active components. The cycle time is 30ms.
So you can do about 3e15 computations per second. The typical new desktop these days is about 3e9 instructions per second. So the gap is a factor of about 1,000,000. This gap should close in about 20-30 years, assuming Moore's law continues.
I know this is rough but it does give the flavour of the problem.
Current computers have the processing power of an insect brain, and they are mostly about that smart.
I think computer scientists have done pretty well given the lack of CPU power that is available. A lot of things that computers have trouble with such as vision procession are handled in the brain by brute force - many neurons in parallel doing lots of simple repetitive things.
Assuming Moore's law continues, we are going to see a dramatic closing of the gap between silicon and carbon based intelligence in our life times.
AI is the field to be in over the next few years. Having the processing power is not enough. There will be many theoretical questions to answer before we can build truly smart machines.
> Who wants to use sausage software?
Good one - I will be using that!
"All the software is written in ObjectiveC as it came from NeXT... who wrote the ObjectiveC support you'll now find in GCC and GDB".
Yes Jobs gave that to the community after he was threatened with a lawsuit which he knew he would lose. How generous.
1. There is nothing that ensure that even an analog recording will be possible into the future, as a way of delivering fair use. For example watermarking technology could be used to mark content and the studios could strong arm the manufacturers to stop making such cameras. Such cameras are, after all, a circumvention device - a mechanism exists to prevent copying and the cameras bypass this (by ignoring the watermark).
You might say this is hypothetical, but no more hypothetical than any piracy that DeCSS is supposed to allow.
So the 'take a video of the screen' argument does not answer the question.
2. Regional encoding. I buy a DVD while working in the Europe and take it home. I cannot play it. The studios argue that you can buy a second DVD player that works in Europe. First, this is an argument only a wealthy person would make. Second, due to differences in voltages and TV formats (eg NTSC vs PAL) this is actually quite difficult, and invariably results in a far worse picture.
3. I do music analysis, and a lot of researchers do this. This requires access to high quality digitized music. Music is headed for DVDs - don't think you are going to be able to buy CDs for long! Without a way to bypass the encryption, I cannot do this, and I will have to try and tap the analog signal at some point. This means worse quality signal and more dificulties analyzing the music. Again, the technology is moving to a system where all components will incorporate the protections mechanisms and it will only be possible to record using a microphone in front of the speakers. Future use of watermarking technology and recorders that recognize the watermarking may even make it impossible to take a recording. See (1) above.
C, the language DeCSS is written in, is a derivative of Algol. Algol (algorithmically oriented language) was originally developed to express algorithms in precise form. It was only later that peoiple thought - hey, you could write a compiler for this and use it for programming.
If coding is not expressive, why include comments to clarify meaning? Why have meaningful variable names? Why not just publish object code?
> No, it doesn't leave permanent damage. There's a way to remove scars: it's called maturity.
This reminds me of those inane arguments you used to hear about corporal punishment - "well it never did me any harm". Or the arguments that smoking must be harmless because Uncle Fred smoked and lived to 80.
Yes a lot of people are bullied. Yes most people don't go beserk. But not everyone can cope with dialy insults, threats and humiliation. And not all bullying is alike - it goes way beyond 'a bit of teasing'. And while most people may be able to get on with your lives some are less resiliant.
Adults do not have to put up with this kind of behaviour and nor should children.
Why aren't they suing the school for creating an environment where people felt so hopeless, so worthless, that they had nothing to lose by going on a murderous rampage.
> It takes about 30 seconds in Office 2000:
> 1.Click on paperclip to bring up a dialog balloon
> 2.Click on "options" button
> 3.Uncheck "Use office assistant" box.
Yes, but how long does it take to work this out. This information is quite hard to find, unless you ask the paperclip.
Actually the paperclip is very useful, because you can ask it questions in plain English and mostly get useful answers.
The problem was it popped up too often
And this is the one that really annoyed me: when you clocked on it to go away, it would wait a few seconds, then wink, then eventually disappear. If it was a person I would have punched its lights out.
"- credit creation - the unbundling of the currency from any tangible asset such as the gold standard".
I agree with the rest but this doesn't make sense.
The first problem with credit creation is that here is a constant risk that governments will print money and use inflation to pay off their debts or to finance unsustainable expenditures.
We saw this in the 1970s and in the 1920s. The cost of the uncertainty this creates is enormous - see how even the slightest sign of inflation 'spooks' financial markets.
The second problem is incompetent management of the money supply. This was the actual cause of the great depression. They went off the gold standard and congratulated themselves how clever they were creating the unsustainable 1920s boom by printing money. Eventually the party ended and the stock market crashed. During the next year or so, the money supply fell by 60%. If you do that, you will cause a depression. This is all documented at length in Milton Friedman's book on the monetary history of the US.
Governments have mostly learned how to manage money supply so this problem rarely occurs these days.
People often assume that the 1930s depresssion was a product of free markets, when in reality it was due to incompetent management of regulated money supplies by newbies who didn't know what they were doing.
While the gold standard was in place there were booms and busts, but the busts were usually short lived and the rate of economic growth was far higher than it is now. When you graph the stock market before and after going off the gold standard, there is a knee in the curve - growth is much slower afterwards.
One comment on the book - it makes the very good point that secure title to property is key to economic growth. Having visited a number of third world countries I can confirm this is a huge issue. But it is just a part of the wider problem of the rule of law. When every government decision is dictated by 'connections' or bribes, and when it is impossible to compete with those who have the right connections due to government harrassment, you end up with inefficient entrenched monopolies.
Ghandi saw this problem; in fact his first big protest in India was against a government-mandated monopoly in the production of salt.
The USA is not immune from these problems. Look at the actions of the recording industry to extend their control over consumers and to extent the reach and duration of copyright. The fact that you need tens of millions of dollars to get into congress delivers a system that gives those who are already wealthy and organised what they want.
The USA was very slow to crack down on tobacco companies for this reason. Other countries took effective action a long time ago.
I often wonder whether those in authority really want to stop the illegal drug trade. The economist had a graph recently that showed that heroin prices have been in almost constant decline for 15 years. "The way on drugs is going well and we will succeed". Right, like you won the war in Vietnam.
The article states they they have been able to break the crypto. I personally assume that since they removed the export barriers on 128 bit crypto they can break it.
I suspect that all this carry on is to make people think they are safe to use PGP, when in reality they probably are not. So people will use PGP and get their messages decoded.
The bottom line is if those people want to find out what you are saying they will. Even if they have to park a van out the front of your house and watch your screen and read your mouse clicks.
One can also assume, based on known past practices, that the US uses their crypto capabilities to obgain commercial advantage for US companies by feeding them information obtained from cracking secret commmunications.
Not everyone speeds. Inexperienced drivers and those who do not think, speed.
Most young people vastly overestimate their driving ability. After a few close calls they often modify their behaviour. The road tool kills huge numbers of people every year. Speed is a major factor, because it reduces the time you have to react and it increases the damage done when you hit. Don't forget the energy is proportional to the square of the speed.
A lot of the posters seem to assume that the system needs to be perfect for it to be useful. In fact the system could just turn itself off when it doesn't know about a road.
A possible alternative to the 17 CDs and the GPS would be a simple system of bar codes on the roads with a bar code reader under every car. I would welcome such a system, because it is often hard to keep track of the changing limits as you drive along, and this system would pick up the changes automatically. You just set the cruise control to speed limit - 2.
One point I think a lot of people are missing is the phenomenon of 'worse is better'.
e ws .html
When Unix replaced some mainframes, the mainframers pointed out that Unix lacked some features of mainframes. The same when NT replaced some commercial Unix systems. The same when Linux started replacing some NT systems.
It didn't stop the replacements happening. Mainly because the replacement was cheaper.
The bottom line is that if the replacement is cheaper and good enough, then it will win.
Mainframe revenue is falling, Unix vendors are vanishing rapidly. Microsoft's share price has halved. All because of competition from systems that are not as good in some respects.
Linux does not have to be better than NT in every respect to win. It just needs to be good enough.
Over time the gaps will be filled, by free software or commercial products. System monitoring tools, job scheduling etc.
Here is a very interesting article on this topic
http://www.ai.mit.edu/articles/good-news/good-n
Unless you can get a waiver from the developers,
all of them, you have to open source your additions.
This applies even if you try and encapsulate the GPL code in a library.
LGPL would be OK though, as would BSD. That's what's wrong with BSD IMHO. Companies can
privatize your code and give nothing back.
Don't think you will get away with it. Many have
tried and gotten caught. You will be exposed as
a law breaker and thief.
> Yeah, right. Ok, lets address stuff in order: ...
> Sure the Pentium 4 doesn't perform great on code not optimized for it. But neither did the 486, the Pentium, or the Pentium Pro. And which would you prefer to have right now, a 250MHz 386, or a 1GHz Pentium III?
Well actually the P4 is the first X86 from Intel that is slower on the targetted workload. The P-Pro was slower on 16 bit code but it was faster on 32 bit code.
Running Quake faster is not good enough - not everyone runs Quake. All the stuff I do will be slower (compilations etc).
It's all very good to crtiicise compiler writers for not optimizing for the latest CPU. The problem here is that in the past, you have not had to severely penalize past CPUs to write code that runs better on new CPUs. But that is just what Intel has done this time. The other problem is that it is not simple to optimize for the P4. The optimization manual is 300 pages and the rules are very complex. It is not reasonable to expect everyone to rewrite the code generation every time. The GCC code generation is hundreds of thousands of lines of code. Anyone who thinks it is easy to implement Intel's suggestions should submit a patch to the gcc patches mailing list.
This was indeed a major point of the article, that AMD and Transmeta do not require code rewrites to run well. But Intel does for the P4. I am working on a compiler and my attitude at the moment is that the P4 is not worth the effort at this stage.
Looking at Intel's design decisions on the P4 really makes me scratch my head. The business with the shift instructions seems particularly incredible. I have been reading and these optimization guides since the 286 and I am at a loss.
There seem to be three tenable theories.
The megahertz theory. The trouble with this is that to even match the Athlon you need such high clock speeds it is not going to solve the problem for quite a while.
Theory two is that Intel has put multithreading onto the CPU at the cost of a lot of silicon, but hasn't been able to get it going yet. This would explain why they have been so stingy with silicon where it could have been useful. The multithreading could double system throughput by running two threads at the same time on the same chip. That would justify the design decisions made, but this theory is a long shot.
Theory three is that Intel has been taken over by marketing pukes and is doomed. Engineering realities come a distant tenth to PR, Hype etc. Check out http://www.faceintel.com/ for some information about what Intel is like from the inside. This seems the most plausable theory at this stage. Intel has had a long series of debacles so it is not surprising to see another one.
I read the
One reasonably sucessful Author told me she makes more from library royalties than from book sales. This is in Australia where libraries do pay royalties.
Here you also have to pay royalties if you play a radio in your store or workshop.
If someone working for me advanced this argument ie "I want a way to avoid taking responsibility", I would fire them.
In fifteen years of working in enterprises, I found very few companies actually provide any worthwhile support. IBM is one, but it is one of few. You can get Linux support from IBM I believe.
Otherwise it is just the usual "retry, reboot, reinstall, upgrade" routine. This is not just from tinpot outfits either. Everyone knows what Microsoft support is like, but they are not unusual in this regard.
The typical support organisation consists of a call centre staffed by people following scripts, plus a large number of marketing types. Actual people who can help solve your problem are thin on the ground in many cases.
I am not exaggerating. MSFT spent more money marketing Windows 95 than developing it. Progress Corp spends 250% as much on marketing as on product development.
Often when you find a problem, the vendor will put out a doc change, saying that you shouldn't do whatever triggered the bug, or in other ways say it is expected behaviour. Then they redefine your bug as an enhancement request. Or they do what MSFT did with multilevel numbering in MS Word 97. Remove the feature and pretend it never existed (I am not talking about multilevel heading numbers).
Another thing they do is ask 'Is patch X on'. If not, even though it does not look like your problem, the ball is back in your court. Another good ploy is to ask for impossible amounts of diagnostic information. "Delete records from the database until the problem goes away and then send us a dump of that record", for example.
For hearing this nonsense, you are paying through the nose for a 'support' contract.
The other thing they do is have very tight upgrade requirements. Older versions are not supported and you are *really* on your own, unless you keep to a breakneck and very expensive upgrade schedule. Up the upgrade creek without the source code, in fact. Upgrading is not a simple matter. You may have a large number of products with complex version interdependencies and OS and hardware dependencies. Sometimes you can't get there from here.
In a live situation it is almost always up to you to find a workaround or bypass. Stop using certain features, reboot every x hours, simplify the system, filter out errors that cause the software to choke, hone your recovery procedures. Swap out hardware until you find the bad component.
I have been in the situation more than once where closed source vendors tell you "We can't [or won't] fix it and we won't let you fix it" (by giving you the source code). What do you do then? In my experience someone within the organisation carries the blame. VPs are very uninterested in hearing that xxx Corporation (NASDAQ XXX) is to blame. They want *you* to fix it *now*.
More than one closed source vendor has gone bankrupt. Then you find they forgot to send the source to the escrow agency. Or they just lose the source code without going bankrupt. Then you get to reverse engineer the file formats, which is tedious after the first few times.
If they don't go bankrupt they get taken over. Then they jack up the maintenance costs by 375%, or the previously strategic product becomes unsupported, you are frog marched to the Gigantic Software Corporation's often inferior product suite, and all your investment in training and product is wasted. You can't read your old files any more.
Some things to look for in a support organisation:
1. A searchable bug and knowledge base.
2. Access to techos, as opposed to Marketing people. Job titles can help but you need to speak to them.
3. Knowledgeable, experienced technical support people
4. Technical support people who are not snowed under. Find out the numbers.
5. A number you can call at any time and they will answer and be able to get someone to help.
6. Clear definitions of problem categories and levels and specific service levels for a) restoration of service and b) provision of a fix for the problem for the various categories.
User groups are very important. Often you will be told "Noone else has this problem so it must be due to something you are doing wrong". Often this is a lie. The vendor must not be able to censor your mailing lists.
1. The 64k segment limit on the 80286. A disaster that held PC back for years.
2. The 432 CPU. Again, a failure partly because it had a 64k segment limit.
3. The overly complex protection model on the 80286 and successors. Made writing OSs and compilers more complicated, apart from wasting silicon.
4. The horrible 'not quite a stack' architecture of the x87 numeric processor. Slows things down by creating a bottleneck at top fo stack and makes writing numeric code or compilers very difficult.
5. Lack of full hardware support for virtualizing protected mode operating systems. Requires cooperation from the target OS to virtualize it. Whereas eg in system 390 the target has a hard time even knowing it is virtualized.
My site has been instrumental in at least two cases in preventing people being treated for CSR with the very thing - cortisol - which causes it. Because it is a rare condition, incompetent treatment is not uncommon.
I am adamantly opposed to this proposal. The WHO should do what everyone else does - put up their web pages, and earn their reputation. Instead what they are proposing is that they become gatekeepers to all health information on the internet. I don't object if they want to create a system whereby you can have a 'WHO approved' logo on your site, but putting one highly bureacratic organisation into such a powerful position is a recipe for disaster.
This is not in line with the internet model which is that people should be able to decide for themselves what they believe. Anyone who believes that officials have their own best interests at heart and that they have no conflicts of interest is naive. This system would also introduce huge delays in getting information online.
The problem is that the apps that are out there do not need the latest and greatest CPU any more.
.au/wav files into musical scores, a useful version of office assistant and the like - all the AI stuff. It will be a while I think.
Once upon a time every time you upgraded you would get heaps more done - I remember when I upgraded to an 11mhz AT - INCREDIBLE!!! You even had to upgrade often just to run something at all. For a long time more CPU was needed to make GUI fast enough - but even unaccelerated X is fast enough on a 333mhz Celeron X.
There are a few exceptions, but even speech recognition does not really benefit enough to pay through the nose for 1ghz.
I can compile GCC in 18 minutes from scratch - that's fast enough.
This hurts MSFT too because a lot of their sales are bundled with new PCs.
At some point there will be a killer app that needs 2.9ghz or whatever. But there is no killer app that needs 1ghz right now. There are lots of things that are not there due to lack of CPU, but they need heaps more than 1ghz - good speech recognition in a noisy office; decoding music
Also we are more bandwidth constrained than anything, rather than CPU bound.
> Custom written software like that is one instance where you *can* sue the people who wrote it if it fails, and ...
It is very difficult to specify software to the degree of detail where you can actually sue someone. Specs change during development anyway.
Suing someone rarely does any good. The software vendors are experts at this - it is a core competency to negotiate contracts with plenty of holes in them that aren't apparent at the time.
And of course the PHB is breathing down your neck wanting to get started and can't understand why you are being so pedantic about the contract. The sales guy is telling him they are keen to go but 'you man' is holding things up.
The situation is not much better than trying to sue MSFT when powerpoint has a bug.
A good vendor will let you have source code. If not it is a good sign you are about to be done over. At a minimum you should have an escrow agreement so that they place the source at a trusted thrid party so it they go bust or - worse - get taken over by CA - you are not left twisting in the wind.
Finally it is no compensation to sue someone if your plane just crashed. (Airline safety briefing for busy people: If the plane crashes you will die).
All this goes to show that the US congress is a wholly owned subsidiary of big business, just in case anyone hadn't realized. Clinton signed the previous H1B bill just before a fundraising trip to Silicon Valley.
The DMCA Act is another example of the same thing. In fact if you look at the history of copyright and patent law big business has been getting what it wants for 100 years. Since, in fact, the US became an exporter ot intellectual property. Before that the US did not honor intellectual property just as 3rd world countries do not today. Now is is regularly extending the scope of it.
H1B is about driving down the salaries of programmers. As many posters have pointed out H1Bs are in many ways like indentured servants and are in a very weak and vulnerable position.
I know several people who went over (from AU) as H1Bs and for most it was an unhappy experience. Things were not as they were promised and there was nothing they could do about it.
Several people argued that if we do not allow the H1Bs the jobs will move offshore. As someone who works for a multinational corporation I can tell you that running projects across time zones is a nightmare. This is a huge barrier to exporting these jobs. So many companies tried to do this with Y2K work and ended up doing it in house because of the poor results.
A much better way is to have a skills based immigration scheme and allow the required people to get green cards day one. This avoids the 'indentured servant' problem at least and employers will have to pay market rates.
I worked in the US for a while and I was really surprised at the low salaries that people were earning. Especially considering the high cost of living in the US (housing, fresh food, restaurants, high heating bills in the north and cooling bills in the south).
I've seen email introduced to several large organisations, and every time there was a lot of flaming at first. I even used to do it myself.
People don't normally flame each other in person. Sometimes they do on the phone and often on email. Although maybe they want to - if you look at the soap operas they are all people telling each other exactly what they thing of them - exactly what you can't do in real life.
I think the reason for the email flames is that with email, and to some extent with the phone there is a lack of feedback.
You write 'I am disappointed that you did not accept my patch'. No (real time) response.
In normal human interactions if you say something like that, the person looks concerned and responds. With email, any response is much later, if at all. In real life, if the person does not respond you increase the voltage - you might speak a little louder or use body language to get the point across. This is all difficult over email.
So people flame. After they get used to the medium it dies down.
Of course others flame just to wind up people.
Tim Josling
In any deal like this, you need to allow for a lot more management overhead than with an in house project.
Some things that can help
- project will work best if the nature of the work is that requirements are clearly defined up front and it is easy to determine if they have succeeded.
- need to have *frequent* clearly defined deliverables ie weekly. Otherwise you will have no idea how they are going and you will find the project will suddenly slip months in one day.
- require proof of testing. Otherwise they will not test it.
- regular on site visits. You can initiate these at the start of the project, or later on when things go off the rails, as you prefer. There is no substitute for seeing the atmosphere at their place. It's best to have someone there all the time. That way they get used to having you there and you find out what's really going on.
You cannot outsource responsibility. You still have to manage the whole process very tightly. We do design reviews - no coding allowed before the design is approved, and we do code reviews by phone to enforce coding standards and documentation standards.
Good luck.