So some artists distribute their work legally on P2P networks. But what is the point? Wouldn't it be much simpler and more convenient just to publish a URL and serve files with http?
People sometimes suggest that mirroring files saves bandwidth, but that can be done with http as well, and in general P2P services are quite wasteful of bandwidth, not choosing the most direct route for sending files but some meandering path between lots of peers. That's because they aren't optimized for efficient network usage, but for avoiding detection of who is sharing which files.
If one day, everyone decided that they didn't want to download any more pirated MP3s, would we still need P2P networks?
The most difficult part of writing a module for CPAN is getting past the bureaucracy: there are at least two different documents giving different instructions on how to submit modules, and the modules@perl.org mailing list is notorious for never replying to any messages sent to it, especially not those asking for help or 'what went wrong?'. It's not like registering a project on Sourceforge or Freshmeat; you have to send mail to register a namespace, then do an ftp upload and then send mail to have your upload 'noticed'. Sounds reasonable enough, except that it is easy to go wrong or fail to jump through a particular hoop, and if that happens you get no feedback at all. I'm sure the CPAN maintainers are busy and hard-working, but the current situation shows there are cases where an automated system (which at least gives error messages when stuff goes wrong) is better than manual administration.
Myself, I have managed to put modules on CPAN, after a few months of wondering why they didn't appear in the main listing. (It turned out I had forgotten to send an updated module list entry; my fault, but still it would have been friendlier for somebody to answer my question about what I did wrong rather than just ignore it.) But uploading newer versions of the same modules has sometimes been troublesome, with the update appearing in my home directory but not the main listing: again, any request for help or guidance on what part of the process I was doing wrong was studiously ignored.
Despite the troublesome upload procedure, CPAN is Perl's biggest asset and other language communities would do well to copy it (as some are). But please don't copy the management system; have some more structured way to submit code where it's clearer what to do at each stage, and there can be warning messages if you do something wrong, rather than just quietly failing to work.
I think one of them routes and the other one routs. But when you stick -er on the end the written difference disappears.
There's a similar question with -ing, which has led some British publications to distinguish between 'routeing' and 'routing', but most people just write routing for both.
The trouble is that independent creation is no defence to patent infringement, so if you did write your own algorithm you could still be sued if it had elements in common with the patent filed. I don't know about the MPEG patent in particular, but in many cases software patents are so broad you cannot write any algorithm that doesn't infringe. They become a government-granted monopoly on the whole application area.
This is one of the reasons why copyright is a better fit for software than patents: if you create something independently, without looking at another's code, you cannot infringe the copyright.
Blake's 7 == Robin Hood in space, except that Robin dies / disappears fairly early on. And it's more kind of 'dark' and not morally clear-cut. Doesn't take itself too seriously though (not like say B5).
(Hmm this reminds me that Danny John-Jules was both Barrington in Maid Marian and the Cat in Red Dwarf, but I don't think that is relevant.)
Yeah of course it costs money and almost certainly more than the profit margin on the drive. The question is what proportion of drives will 'fail' and be returned as warranty replacements. If only 5% or so (which seems very high to me), then the policy is workable.
Of course if every customer decided to have a big warranty-fest and send back lots of perfectly working disks (getting other working second-hand disks in exchange) then it could be a bit of a drain. But I don't believe that is likely.
Anyway I was proposing this warranty policy as a lower-cost alternative to bothering to discriminate between good and bad returns - because the labour costs are lower. Surely you agree that some monkey getting a drive in the post and sending back a replacement (with optionally some techs in the background to run a few automated tests on the returned unit) is cheaper than a more rigid returns policy? The problem would come only when this policy caused a much higher number of warranty returns (but even then it might be cheaper to send out a replacement drive rather than have support people on the phone).
They can't do whatever they want, because they would then be in breach of copyright. It is copyright law, and not any kind of 'agreement' or contract, that gives the GPL its force.
As long as UnitedLinux is based in a country that applies copyright to software, they cannot distribute any code without the permission of its copyright holder (either the author or the FSF, in the case of most Linux stuff). So the copyright holder could sue them if they distributed code without permission - in other words, outside what is specifically permitted by the GPL licence they received.
Er, both Mandrake and SuSE use RPM for package management. I think you are talking about whatever layer they slap on top for resolving dependencies and the like. Mandrake has urpmi (which is free software); SuSE has some new thing for their upcoming distribution (which will probably be non-free, like YaST).
BTW - if you value package management so highly, why dismiss Conectiva? They're the folk behind apt-rpm.
(I realize the parent post may have been a troll but if so it was a fairly cogent troll and worth replying to.)
Well part of the difference is that the GPL isn't an EULA. You don't have to 'agree' to it and you are not bound by it. However, under copyright law you need the permission of the copyright holder to distribute a work, and that permission is available only under the terms of the GPL (in most cases). An EULA purports to bind you just because you try to run the software, even though running a program is (in most countries) not restricted by copyright.
Basic seems just as much a 'functional language' as Logo... it didn't have the same nice list handling, but made up for it with more flexible data structures in general. Well I guess it depends on which Basic variant you use. One with just GOTO and GOSUB would be pretty nasty, but as long as it has named procedures and functions it's vastly preferable to Logo.
I think that in general, not many were 'introduced to the functional paradigm' by Logo... the Logo programs I've seen in introductions are wholly imperative (REPEAT 50 FD 20, and the occasional imperative TO X procedure declaration). It would have to be a pretty advanced course to get into the functional stuff. Your particular class sounds like that, but it's likely most weren't.
I looked at the original GoF book and gave up on it as almost unreadable. It seemed to have about half a page of mumbling followed by two pages of dry C++ code, repeat for each pattern. I just couldn't be bothered, there are so many more worthwhile things to read for the same effort. Like real code, for a start. (At least I think it was the original GoF book, it might have been some lame cash-in but I don't think so.)
Maybe it's just a prejudice I have against long code samples in books.
I don't think you understand what I meant. If it turns out that people are sending back lots of perfectly-functioning disks, then you can 'refurbish' them (in other words do nothing) and send them out as warranty replacements to the next person. Costs no more than testing the drive.
Maybe Logo was once a decent language - a Lisp variant - but the versions most people use must have been stripped down from that.
I remember Logo as being less useful than Basic - what really sticks in my mind is the brain-dead syntax for string literals, where the first space character ends the string. Hello? Strings with spaces in them?
There might be ways around that; there might be decent control structures (not just REPEAT 50), ways of creating data structures like arrays, and some means of calling external procedures or operating system routines. But they didn't get much emphasis in Logo programming textbooks, perhaps because they'd differ from one pl atform to another.
Ah, looking at the old Logo book on my bookshelf I see that the language had list handling similar to Lisp, its parent. But much clunkier and less elegant. Logo may be a functional language but I wouldn't call it a decent one.
I envisaged a system where the customer could send in the old disk and get a replacement - which might itself be a 'refurbished' warranty replacement from someone else, if it's in GWO. If the customer wants to do that every month, fine by me:-P.
I take your point about the phone support costing money, even if the returns policy is generous. But I still think that tripling the price of the drive to allow for two extra years of warranty is a bit much. Maybe it was just an exaggeration. (Suppose a warranty return costs 5x the price of the drive: then if there's a 5% failure rate - very high - that amounts to a 25% overhead on each unit sold. Not a tripling.)
Yes, it costs money to have the techs look at why the part went bad, but I'd consider that money well spent. It's not a downside of having longer warranties that you get the opportunity of examining the failures - it's a benefit. Although if you are worried about the cost you could just throw away all faulty, returned drives.
Yep - it would force governments to either allow unrestricted access, or block the whole site. But the fact that China backed down from blocking Google as a whole (it was banned, and then un-banned, apparently) suggests that they recognize its usefulness. There's at least some chance that China would choose to allow access rather than blocking the web's best search engine. And if all major websites followed suit, they'd have no choice but to allow access.
This was the original promise of the Internet for freedom of speech worldwide - 'you can't restrict it, it's all or nothing'. Sadly it did not turn out like that. But some steps could still be taken in that direction.
If drive costs are low, you don't really need to explain about master and slave, etc. All you need to do is say 'send it in, and we'll send a replacement'. That replacement can be either a new drive or a 'refurbished' one that was previously sent in for a warranty claim (ie, there was nothing wrong with it). I don't think the administrative costs are that high.
Hardware may be getting cheaper but time isn't. Restoring from backups is a real PITA and so is having to order and install a replacement drive. Having hardware that lasts a long time is just as important as it always was.
Why would a three-year warranty double the cost of the drive? At worst all you have to do is send a replacement drive to everyone who claims - plus a small amount in costs for handling and postage. Suppose ten per cent of the drives fail within three years, compared to five per cent within one year. Then extending the warranty from 1 to 3 years would add about 5% to the price.
It could only double the price if a huge proportion of drives failed in three years!
This is why a long warranty is a good thing. The manufacturer is saying, I'm confident that not many drives will fail during the warranty period, and I'm prepared to put my money where my mouth is. It's like the manufacturer placing a bet that few drives will fail. If the manufacturer is confident enough to do that, then it's likely to be a well-made product.
Slashdot Mirror
So some artists distribute their work legally on P2P networks. But what is the point? Wouldn't it be much simpler and more convenient just to publish a URL and serve files with http?
People sometimes suggest that mirroring files saves bandwidth, but that can be done with http as well, and in general P2P services are quite wasteful of bandwidth, not choosing the most direct route for sending files but some meandering path between lots of peers. That's because they aren't optimized for efficient network usage, but for avoiding detection of who is sharing which files.
If one day, everyone decided that they didn't want to download any more pirated MP3s, would we still need P2P networks?
Sure, this will make system administrators obsolete, just as we don't need programmers any more now that we have compilers and RAD tools.
The most difficult part of writing a module for CPAN is getting past the bureaucracy: there are at least two different documents giving different instructions on how to submit modules, and the modules@perl.org mailing list is notorious for never replying to any messages sent to it, especially not those asking for help or 'what went wrong?'. It's not like registering a project on Sourceforge or Freshmeat; you have to send mail to register a namespace, then do an ftp upload and then send mail to have your upload 'noticed'. Sounds reasonable enough, except that it is easy to go wrong or fail to jump through a particular hoop, and if that happens you get no feedback at all. I'm sure the CPAN maintainers are busy and hard-working, but the current situation shows there are cases where an automated system (which at least gives error messages when stuff goes wrong) is better than manual administration.
Myself, I have managed to put modules on CPAN, after a few months of wondering why they didn't appear in the main listing. (It turned out I had forgotten to send an updated module list entry; my fault, but still it would have been friendlier for somebody to answer my question about what I did wrong rather than just ignore it.) But uploading newer versions of the same modules has sometimes been troublesome, with the update appearing in my home directory but not the main listing: again, any request for help or guidance on what part of the process I was doing wrong was studiously ignored.
Despite the troublesome upload procedure, CPAN is Perl's biggest asset and other language communities would do well to copy it (as some are). But please don't copy the management system; have some more structured way to submit code where it's clearer what to do at each stage, and there can be warning messages if you do something wrong, rather than just quietly failing to work.
But does it run Linux?
I suppose they were launching one thread for each concurrent user?
I think one of them routes and the other one routs. But when you stick -er on the end the written difference disappears.
There's a similar question with -ing, which has led some British publications to distinguish between 'routeing' and 'routing', but most people just write routing for both.
The trouble is that independent creation is no defence to patent infringement, so if you did write your own algorithm you could still be sued if it had elements in common with the patent filed. I don't know about the MPEG patent in particular, but in many cases software patents are so broad you cannot write any algorithm that doesn't infringe. They become a government-granted monopoly on the whole application area.
This is one of the reasons why copyright is a better fit for software than patents: if you create something independently, without looking at another's code, you cannot infringe the copyright.
Blake's 7 == Robin Hood in space, except that Robin dies / disappears fairly early on. And it's more kind of 'dark' and not morally clear-cut. Doesn't take itself too seriously though (not like say B5).
(Hmm this reminds me that Danny John-Jules was both Barrington in Maid Marian and the Cat in Red Dwarf, but I don't think that is relevant.)
Yeah I guess you're right - if such a policy really were cheaper then companies would already be doing it.
Yeah of course it costs money and almost certainly more than the profit margin on the drive. The question is what proportion of drives will 'fail' and be returned as warranty replacements. If only 5% or so (which seems very high to me), then the policy is workable.
Of course if every customer decided to have a big warranty-fest and send back lots of perfectly working disks (getting other working second-hand disks in exchange) then it could be a bit of a drain. But I don't believe that is likely.
Anyway I was proposing this warranty policy as a lower-cost alternative to bothering to discriminate between good and bad returns - because the labour costs are lower. Surely you agree that some monkey getting a drive in the post and sending back a replacement (with optionally some techs in the background to run a few automated tests on the returned unit) is cheaper than a more rigid returns policy? The problem would come only when this policy caused a much higher number of warranty returns (but even then it might be cheaper to send out a replacement drive rather than have support people on the phone).
Which is which?
They can't do whatever they want, because they would then be in breach of copyright. It is copyright law, and not any kind of 'agreement' or contract, that gives the GPL its force.
As long as UnitedLinux is based in a country that applies copyright to software, they cannot distribute any code without the permission of its copyright holder (either the author or the FSF, in the case of most Linux stuff). So the copyright holder could sue them if they distributed code without permission - in other words, outside what is specifically permitted by the GPL licence they received.
Er, both Mandrake and SuSE use RPM for package management. I think you are talking about whatever layer they slap on top for resolving dependencies and the like. Mandrake has urpmi (which is free software); SuSE has some new thing for their upcoming distribution (which will probably be non-free, like YaST).
BTW - if you value package management so highly, why dismiss Conectiva? They're the folk behind apt-rpm.
(I realize the parent post may have been a troll but if so it was a fairly cogent troll and worth replying to.)
Well part of the difference is that the GPL isn't an EULA. You don't have to 'agree' to it and you are not bound by it. However, under copyright law you need the permission of the copyright holder to distribute a work, and that permission is available only under the terms of the GPL (in most cases). An EULA purports to bind you just because you try to run the software, even though running a program is (in most countries) not restricted by copyright.
Basic seems just as much a 'functional language' as Logo... it didn't have the same nice list handling, but made up for it with more flexible data structures in general. Well I guess it depends on which Basic variant you use. One with just GOTO and GOSUB would be pretty nasty, but as long as it has named procedures and functions it's vastly preferable to Logo.
I think that in general, not many were 'introduced to the functional paradigm' by Logo... the Logo programs I've seen in introductions are wholly imperative (REPEAT 50 FD 20, and the occasional imperative TO X procedure declaration). It would have to be a pretty advanced course to get into the functional stuff. Your particular class sounds like that, but it's likely most weren't.
I looked at the original GoF book and gave up on it as almost unreadable. It seemed to have about half a page of mumbling followed by two pages of dry C++ code, repeat for each pattern. I just couldn't be bothered, there are so many more worthwhile things to read for the same effort. Like real code, for a start. (At least I think it was the original GoF book, it might have been some lame cash-in but I don't think so.)
Maybe it's just a prejudice I have against long code samples in books.
I don't think you understand what I meant. If it turns out that people are sending back lots of perfectly-functioning disks, then you can 'refurbish' them (in other words do nothing) and send them out as warranty replacements to the next person. Costs no more than testing the drive.
If glass or transparent plastic were a bit cheaper, you could just replace the windows and not need to clean them.
How about a thin plastic film which you spray onto the glass and just peel off when it gets dirty?
Maybe Logo was once a decent language - a Lisp variant - but the versions most people use must have been stripped down from that.
I remember Logo as being less useful than Basic - what really sticks in my mind is the brain-dead syntax for string literals, where the first space character ends the string. Hello? Strings with spaces in them?
There might be ways around that; there might be decent control structures (not just REPEAT 50), ways of creating data structures like arrays, and some means of calling external procedures or operating system routines. But they didn't get much emphasis in Logo programming textbooks, perhaps because they'd differ from one pl atform to another.
Ah, looking at the old Logo book on my bookshelf I see that the language had list handling similar to Lisp, its parent. But much clunkier and less elegant. Logo may be a functional language but I wouldn't call it a decent one.
I envisaged a system where the customer could send in the old disk and get a replacement - which might itself be a 'refurbished' warranty replacement from someone else, if it's in GWO. If the customer wants to do that every month, fine by me :-P.
I take your point about the phone support costing money, even if the returns policy is generous. But I still think that tripling the price of the drive to allow for two extra years of warranty is a bit much. Maybe it was just an exaggeration. (Suppose a warranty return costs 5x the price of the drive: then if there's a 5% failure rate - very high - that amounts to a 25% overhead on each unit sold. Not a tripling.)
Yes, it costs money to have the techs look at why the part went bad, but I'd consider that money well spent. It's not a downside of having longer warranties that you get the opportunity of examining the failures - it's a benefit. Although if you are worried about the cost you could just throw away all faulty, returned drives.
Yep - it would force governments to either allow unrestricted access, or block the whole site. But the fact that China backed down from blocking Google as a whole (it was banned, and then un-banned, apparently) suggests that they recognize its usefulness. There's at least some chance that China would choose to allow access rather than blocking the web's best search engine. And if all major websites followed suit, they'd have no choice but to allow access.
This was the original promise of the Internet for freedom of speech worldwide - 'you can't restrict it, it's all or nothing'. Sadly it did not turn out like that. But some steps could still be taken in that direction.
If drive costs are low, you don't really need to explain about master and slave, etc. All you need to do is say 'send it in, and we'll send a replacement'. That replacement can be either a new drive or a 'refurbished' one that was previously sent in for a warranty claim (ie, there was nothing wrong with it). I don't think the administrative costs are that high.
Hardware may be getting cheaper but time isn't. Restoring from backups is a real PITA and so is having to order and install a replacement drive. Having hardware that lasts a long time is just as important as it always was.
Why would a three-year warranty double the cost of the drive? At worst all you have to do is send a replacement drive to everyone who claims - plus a small amount in costs for handling and postage. Suppose ten per cent of the drives fail within three years, compared to five per cent within one year. Then extending the warranty from 1 to 3 years would add about 5% to the price.
It could only double the price if a huge proportion of drives failed in three years!
This is why a long warranty is a good thing. The manufacturer is saying, I'm confident that not many drives will fail during the warranty period, and I'm prepared to put my money where my mouth is. It's like the manufacturer placing a bet that few drives will fail. If the manufacturer is confident enough to do that, then it's likely to be a well-made product.