Well that's kinda my point. You wouldn't think that 'computer program' was too harsh to say to the customers, but marketing types seem to think so, hence 'software product'. Or 'solution'. Consider also the process by which a directory became a 'folder' in Win95.
Yes, you can trick your way around private in C++, and that is a security weakness of C++.
Hardly - in a C++ program running on almost any OS, the libraries execute with the same privelege level as the code which calls them. There is nothing you can do by bypassing the 'private' access controls that you couldn't do anyway. Now, if the libraries executed with elevated privilege, for example a device driver library which had direct access to certain ports while the code calling it did not, it would certainly be a security hole to poke directly into the private data of that library. But when you have one process, one address space and one privilege level, no amount of private data access control is going to make things any more secure.
It's understandable for marketeers and Microsoft to say 'software product' as a euphemism for 'computer program', but do hackers have to start doing it as well?
in consideration of using a vendor's software you agree to their licensing agreement.
I think this is begging the question. Who is to say that you don't have the right to use the software anyway? If you had that right to start with, then it couldn't form consideration for a contract.
Only if you assume that by default you have no rights to install and run the software you paid for does it become possible to talk about agreeing to conditions in exchange for getting some rights. But it's that assumption that people are challenging. If you bought some software you already have the right to use it, and no further contract with the vendor is necessary.
Whenever a Java applet wants to create a window it will have a prominent border saying 'Warning: applet window'. This is so that applets can't spoof dialogue boxes belonging to other applications.
Web browsers ought to do the same, although sticking a border round every image might be overkill. Perhaps some hairy heuristic could work out what looks 'rather like' a system window and disguise it appropriately. I'm talking here of images embedded in web pages (which can still look near enough like a dialogue box to fool novice users). Popup windows, of course, are the spawn of Satan and should be blocked by default anyway.
I just don't see how the price can scale linearly with the amount of space. $1000/month standing charge plus ($10/month)/gigabyte would be more reasonable.
Maybe all screenshots should be banned from software reviews. They don't tell you a great deal and seem to be an excuse for the reviewer not to write anything.
I have better things to do than to throttle users.
Strange, my instinct would be to say 'I have better things to do than track down users in person, I'd rather solve the problem automatically by installing software that allocates bandwidth fairly'.
And I'm sure you don't need those RH9 ISOs right this very damn second. They can wait.
I completely agree. But the users are not typing some command that says 'get me the ISOs this very second and screw everyone else on the network'. They are just saying wget redhat-9.0.iso or equivalent. If the network is unable to understand that other users want to download too and insists on using the whole T1 pipe for this one connection, that's hardly the user's fault.
The Windows profile brokenness you talk about isn't to do with clogging the pipe to the outside world, but with breaking things on the local network. I was just referring to the Internet connection. Of course there may be other good reasons to prohibit downloading large files into a user's profile.
We explicitly designed the system for email, web research, and not much else.
Fair enough. But when someone steps outside those intended uses, the system should not fall over or become unusable for everyone else. That's just fragile design. If you explicitly designed the system for certain uses, then part of the job is to make sure those uses are still available even when some host is doing something silly.
I don't mean you should allow the downloading of ISOs at full speed: it's not in the requirements for your system. But if someone does something outside those requirements the system must not break.
What the article talks about is a difference between two kinds of standards. Those that codify existing practice (SMTP, IP, ANSI C, HTTP 0.9, most of the early Internet standards I think) and those which attempt to create a new standard from scratch. He doesn't like the second kind.
I think it's similar to the argument that says you shouldn't set a program's design in stone before it is implemented, because until you have a working implementation you can't know what the best design would be (nor indeed what the requirements will become). And I have a lot of sympathy with that.
But while a few years of anarchy followed by a period of standardization can work well in some cases, you can't seriously suggest that in areas where there are big upfront costs to get into the market it is better to let people waste effort thrashing around with a dozen different formats or protocols until one of them wins 'in the marketplace'. (And we all know that 'the marketplace' is often lousy at picking the best technical solution, worse even than standards committees.) Mobile phones are a great example. You need to have an agreed standard before you start manufacturing, not afterwards.
If new standard creation is politically motivated by companies who have a potential new product to promote, so what? That's surely preferable to having no standard at all, launching several new products with incompatible formats or protocols, and then years later trying to document and standardize whichever random one of them seems to be the winner. Case in point: where is the standards document and process for MS Word file format?
Can't you throttle based on source and destination addresses? So every host within your network gets its fair share of the bandwidth (whether incoming or outgoing). This wouldn't discriminate against P2P protocols or in favour of SMTP, ssh etc - but it would effectively discriminate in favour of hosts which have only light traffic (their packets would always get through while packets from busier hosts might get dropped).
The trouble with charging by the bit, as the article mentioned, is that inter-ISP or international traffic is much more expensive. It costs your ISP almost nothing to download stuff within their internal network, but if your ISP is paying for a link across the Atlantic or Pacific then downloads over that can be very expensive.
The real answer network protocols that have provision for charging, with routers publishing a price as part of their routing information. But this would require big infrastructre changes and so (like the message-charging proposals to deal with spam) isn't likely to be implemented. But it's technically the right answer IMHO, since you are paying the real cost of what you use rather than some averaged or marketing-determined price. (A lot of the time that real cost will be 'next to nothing'.)
WTF? If you think it's unfair that one user grabbing the RH9 ISOs can hog the bandwidth and excessively slow down other users, it's up to you to throttle things so that every user gets a fair turn. 'Hunting down like animals' is not a scalable solution with a large number of users, nor a particularly intelligent one. If the network is busy then the CD-image downloader should get only his fair share of the bandwidth; late at night when nobody else is using the network the images could download at full whack.
If your network can be reduced to a crawl just by someone running wget(1) then it is the network that needs fixing. I don't mean you have to rush out and buy more bandwidth just to satisfy the users, but what bandwidth there is should be shared out robustly so that one user can't break stuff for the others.
So if 'a tax writeoff has to do with cost, not potential revenue', does this mean that the only deduction they can make for a $1000 server licence is the cost of manufacturing the CD and printing the licence document?
I didn't consider cost to reputation (though it's hard to see how it could damage MS's reputation if pirated software doesn't work correctly) but only monetary cost. And I assumed that someone pirating software would not be so dumb as to call for support.
Given the huge pirating of Microsoft software that happens in many countries (eg China), it's clear that the cost to MS must be zero, otherwise they'd be losing huge amounts of money.
Anyway the point I wanted to make was that if Microsoft 'gives away' copies of Windows, it doesn't cost them a penny. All it requires is for a Microsoft representative to write a letter saying 'you are now authorized to install Windows on your PCs', and the cost of a stamp.
Of course it's not selling below cost. The cost is zero. You could download Windows XP and Office XP from some filesharing network, install them on every PC you own, and it wouldn't cost Microsoft a penny.
If 5c is the smallest coin, then why bother to call them 'cents' at all? Why not just divide the currency into twentieths to start with?
OK, for banking and stuff you may want finer granularity than 0.05. But then you wouldn't necessarily want 0.01 either. For those areas you could just have an arbitrary number of decimal places.
I think countries like Italy had the right idea - make the currency unit fairly small and then there is no need to subdivide it. A lot of complication goes out of the window. The trouble is that a divisible currency is some kind of status symbol (since countries with small-valued currencies are often those which have historically suffered from high inflation) and so countries want to adopt one. For example France changed from 'old francs' to 'new francs' just to make the currency look more prestigious; the euro has euros and cents (and a special advertising campaign when it was launched in countries that previously had indivisible currencies).
In England non-competition clauses in contracts are unenforceable unless limited to direct competition with the other party. IIRC the original case was of a baker who sold his business and agreed not to set up in competition with the new owners. It was held that the contract was enforceable to prevent him setting up shop in the same town, but not if he moved to a different town. I imagine a similar rule applies in most common law jurisdictions. IANAL and I couldn't check the details of the case - it seems rather difficult to search for court cases on the web. If anyone is an L of an appropriate country, perhaps he or she could confirm what the case was called and suggest a good search engine to look through summaries of court decisions.
In employment law, rather than competition law in general, I don't think there is a requirement for the employer to pay you during the non-compete period. If I want that, I can choose to insist on it in a contract; or I can choose to agree to some other terms. That's real freedom.
(Of course it doesn't follow that freedom is always in the employee's interest: for example a minimum wage reduces workers' freedom to work for whatever pay they want, yet it can often increase the prosperity of the average worker.)
AFAIK the machines reboot and reinstall after every use, so trojaning them should be fairly difficult. I don't think sshing from a public machine is any worse than buying stuff with a credit card from a public access machine. I'm not that paranoid.
Most people know that Postscript is Turing-complete, but I think the game of Life is too. So for a _real_ challenge you could implement a Postscript interpreter in Life (provided you find some way to encode the input and some way to do 'showpage').
Actually, I don't know of any working Turing machine or lambda-calculus implementations in Life, but I believe it is thought possible.
I wonder whether the RHUG people will be able to build Jazilla using gcj and so create a native binary package. Then we could see whether it is faster or slower than Mozilla.
Slashdot Mirror
Well that's kinda my point. You wouldn't think that 'computer program' was too harsh to say to the customers, but marketing types seem to think so, hence 'software product'. Or 'solution'. Consider also the process by which a directory became a 'folder' in Win95.
Hardly - in a C++ program running on almost any OS, the libraries execute with the same privelege level as the code which calls them. There is nothing you can do by bypassing the 'private' access controls that you couldn't do anyway. Now, if the libraries executed with elevated privilege, for example a device driver library which had direct access to certain ports while the code calling it did not, it would certainly be a security hole to poke directly into the private data of that library. But when you have one process, one address space and one privilege level, no amount of private data access control is going to make things any more secure.
I think this is begging the question. Who is to say that you don't have the right to use the software anyway? If you had that right to start with, then it couldn't form consideration for a contract.
Only if you assume that by default you have no rights to install and run the software you paid for does it become possible to talk about agreeing to conditions in exchange for getting some rights. But it's that assumption that people are challenging. If you bought some software you already have the right to use it, and no further contract with the vendor is necessary.
Whenever a Java applet wants to create a window it will have a prominent border saying 'Warning: applet window'. This is so that applets can't spoof dialogue boxes belonging to other applications.
Web browsers ought to do the same, although sticking a border round every image might be overkill. Perhaps some hairy heuristic could work out what looks 'rather like' a system window and disguise it appropriately. I'm talking here of images embedded in web pages (which can still look near enough like a dialogue box to fool novice users). Popup windows, of course, are the spawn of Satan and should be blocked by default anyway.
I just don't see how the price can scale linearly with the amount of space. $1000/month standing charge plus ($10/month)/gigabyte would be more reasonable.
The ultrasound mouse-go-away device I've seen emits a very short click every few seconds; the click is audible but not headache-causing.
'Abusing his or her privileges' surely must not mean running wget. Running ping -f or nmap, perhaps. But *downloading a file*?
Maybe all screenshots should be banned from software reviews. They don't tell you a great deal and seem to be an excuse for the reviewer not to write anything.
Strange, my instinct would be to say 'I have better things to do than track down users in person, I'd rather solve the problem automatically by installing software that allocates bandwidth fairly'.
I completely agree. But the users are not typing some command that says 'get me the ISOs this very second and screw everyone else on the network'. They are just saying wget redhat-9.0.iso or equivalent. If the network is unable to understand that other users want to download too and insists on using the whole T1 pipe for this one connection, that's hardly the user's fault.
The Windows profile brokenness you talk about isn't to do with clogging the pipe to the outside world, but with breaking things on the local network. I was just referring to the Internet connection. Of course there may be other good reasons to prohibit downloading large files into a user's profile.
Fair enough. But when someone steps outside those intended uses, the system should not fall over or become unusable for everyone else. That's just fragile design. If you explicitly designed the system for certain uses, then part of the job is to make sure those uses are still available even when some host is doing something silly.
I don't mean you should allow the downloading of ISOs at full speed: it's not in the requirements for your system. But if someone does something outside those requirements the system must not break.
What the article talks about is a difference between two kinds of standards. Those that codify existing practice (SMTP, IP, ANSI C, HTTP 0.9, most of the early Internet standards I think) and those which attempt to create a new standard from scratch. He doesn't like the second kind.
I think it's similar to the argument that says you shouldn't set a program's design in stone before it is implemented, because until you have a working implementation you can't know what the best design would be (nor indeed what the requirements will become). And I have a lot of sympathy with that.
But while a few years of anarchy followed by a period of standardization can work well in some cases, you can't seriously suggest that in areas where there are big upfront costs to get into the market it is better to let people waste effort thrashing around with a dozen different formats or protocols until one of them wins 'in the marketplace'. (And we all know that 'the marketplace' is often lousy at picking the best technical solution, worse even than standards committees.) Mobile phones are a great example. You need to have an agreed standard before you start manufacturing, not afterwards.
If new standard creation is politically motivated by companies who have a potential new product to promote, so what? That's surely preferable to having no standard at all, launching several new products with incompatible formats or protocols, and then years later trying to document and standardize whichever random one of them seems to be the winner. Case in point: where is the standards document and process for MS Word file format?
Standard protocols may suck, but at least they suck in well-known and well-understood ways.
Can't you throttle based on source and destination addresses? So every host within your network gets its fair share of the bandwidth (whether incoming or outgoing). This wouldn't discriminate against P2P protocols or in favour of SMTP, ssh etc - but it would effectively discriminate in favour of hosts which have only light traffic (their packets would always get through while packets from busier hosts might get dropped).
The trouble with charging by the bit, as the article mentioned, is that inter-ISP or international traffic is much more expensive. It costs your ISP almost nothing to download stuff within their internal network, but if your ISP is paying for a link across the Atlantic or Pacific then downloads over that can be very expensive.
The real answer network protocols that have provision for charging, with routers publishing a price as part of their routing information. But this would require big infrastructre changes and so (like the message-charging proposals to deal with spam) isn't likely to be implemented. But it's technically the right answer IMHO, since you are paying the real cost of what you use rather than some averaged or marketing-determined price. (A lot of the time that real cost will be 'next to nothing'.)
WTF? If you think it's unfair that one user grabbing the RH9 ISOs can hog the bandwidth and excessively slow down other users, it's up to you to throttle things so that every user gets a fair turn. 'Hunting down like animals' is not a scalable solution with a large number of users, nor a particularly intelligent one. If the network is busy then the CD-image downloader should get only his fair share of the bandwidth; late at night when nobody else is using the network the images could download at full whack.
If your network can be reduced to a crawl just by someone running wget(1) then it is the network that needs fixing. I don't mean you have to rush out and buy more bandwidth just to satisfy the users, but what bandwidth there is should be shared out robustly so that one user can't break stuff for the others.
So if 'a tax writeoff has to do with cost, not potential revenue', does this mean that the only deduction they can make for a $1000 server licence is the cost of manufacturing the CD and printing the licence document?
I didn't consider cost to reputation (though it's hard to see how it could damage MS's reputation if pirated software doesn't work correctly) but only monetary cost. And I assumed that someone pirating software would not be so dumb as to call for support.
Given the huge pirating of Microsoft software that happens in many countries (eg China), it's clear that the cost to MS must be zero, otherwise they'd be losing huge amounts of money.
Anyway the point I wanted to make was that if Microsoft 'gives away' copies of Windows, it doesn't cost them a penny. All it requires is for a Microsoft representative to write a letter saying 'you are now authorized to install Windows on your PCs', and the cost of a stamp.
Of course it's not selling below cost. The cost is zero. You could download Windows XP and Office XP from some filesharing network, install them on every PC you own, and it wouldn't cost Microsoft a penny.
If 5c is the smallest coin, then why bother to call them 'cents' at all? Why not just divide the currency into twentieths to start with?
OK, for banking and stuff you may want finer granularity than 0.05. But then you wouldn't necessarily want 0.01 either. For those areas you could just have an arbitrary number of decimal places.
I think countries like Italy had the right idea - make the currency unit fairly small and then there is no need to subdivide it. A lot of complication goes out of the window. The trouble is that a divisible currency is some kind of status symbol (since countries with small-valued currencies are often those which have historically suffered from high inflation) and so countries want to adopt one. For example France changed from 'old francs' to 'new francs' just to make the currency look more prestigious; the euro has euros and cents (and a special advertising campaign when it was launched in countries that previously had indivisible currencies).
In England non-competition clauses in contracts are unenforceable unless limited to direct competition with the other party. IIRC the original case was of a baker who sold his business and agreed not to set up in competition with the new owners. It was held that the contract was enforceable to prevent him setting up shop in the same town, but not if he moved to a different town. I imagine a similar rule applies in most common law jurisdictions. IANAL and I couldn't check the details of the case - it seems rather difficult to search for court cases on the web. If anyone is an L of an appropriate country, perhaps he or she could confirm what the case was called and suggest a good search engine to look through summaries of court decisions.
In employment law, rather than competition law in general, I don't think there is a requirement for the employer to pay you during the non-compete period. If I want that, I can choose to insist on it in a contract; or I can choose to agree to some other terms. That's real freedom.
(Of course it doesn't follow that freedom is always in the employee's interest: for example a minimum wage reduces workers' freedom to work for whatever pay they want, yet it can often increase the prosperity of the average worker.)
AFAIK the machines reboot and reinstall after every use, so trojaning them should be fairly difficult. I don't think sshing from a public machine is any worse than buying stuff with a credit card from a public access machine. I'm not that paranoid.
Neat. Although the picture is apparently of one particular TM and not a universal TM (though the site claims this is possible).
Most people know that Postscript is Turing-complete, but I think the game of Life is too. So for a _real_ challenge you could implement a Postscript interpreter in Life (provided you find some way to encode the input and some way to do 'showpage').
Actually, I don't know of any working Turing machine or lambda-calculus implementations in Life, but I believe it is thought possible.
I wonder whether the RHUG people will be able to build Jazilla using gcj and so create a native binary package. Then we could see whether it is faster or slower than Mozilla.