It is with the advent of the Internet that work which is in the public domain can now be made widely available. Despite the (IMHO) excessively long copyright terms, much classical literature and music is in the public domain, and this is added to every year.
Do you see safeguarding the public domain as part of your role?
I've seen previous Q&A sessions to this where the consequences of false accusations were played down "a tiny proportion"... "mistakes immediately rectified".
But such platitudes miss the point which is this: the point about such accusations being made under penalty of perjury is that the infringement should be investigated *before* the accusation is made. For example, sending a notice about the Source code for Open Office because....SRPMS/OpenOffice.... contains the strings MS and Office doesn't constitute an investigation - indeed, if a human being versed in the art had read the filenames, let alone reviewed the file contents, they would have realised that the notice should not have been sent.
It is not acceptable to dismiss this and other cases as rare and isolated incidents: the need for *all* such accusations to be substantiated is the whole point behind the "under penalty of perjury" thing, and organisations which send such notices out without proper review should be prosecuted for perjury. Yet it would appear that these serious allegations are made with "grep | mail".
Actually true "pirates" kill the crew of ships, and steal the vessel and cargo. This violent crime is still very much part of life for seafarers in some parts of the world today.
What you are referring to is large scale commercial illegal copying.
Apologies for responding to a "more than GNU and Linux troll" but...
The important parts of a GNU/Linux system are almost all either GNU or Linux. The kernel is Linux. The compiler is GNU, the linker is GNU, the C libraries are GNU. Most of the core utilities are GNU. Indeed, the first part of any port of GNU/Linux to a different OS is *not* work on the kernel, but work on GCC.
Very few of the applications run on GNU/Linux systems are GNU, as has correctly been identified, just as many Solaris boxes run Oracle. But the bit which makes a running system is almost entirely GNU and Linux.
Dunstan - using GNU software since Linus Torvalds was still at school.
He filed, as has been stated, very general patents a long time ago. He then used a loophole in patent law which is peculiar to the USA. While a patent in any country usually runs for a number of years from the date of initial filing (typically 20 years), the USPTO allowed an amended filing to get a new filing date, rather than the original one. By repeatedly filing amendments, the patents can last far longer than the intended term.
Lemelson used this technique to run patents from the 1960s forward to the 90s, and at a time of his choosing issued "cease and desist" letters to almost everybody who used barcodes for anything. Given the possibility of triple damages if they continued using barcodes without a licence, and given their reliance on barcodes in their processes, just about every recipient of the letter paid money for a licence.
The technique used was to pervert a provision intended to allow an inventor to refine his ideas and instead exploit it for entirely litigious and not at all technical reasons. To pervert such a concession and succeed in legally extorting over $1billion from the wealth creating sector of the economy is repulsive.
We may often hear the Microsoft tax being referred to in this forum, but when you pay the MS tax you get something in return - an operating system which is not very good but adequate for many purposes. In contrast, many manufacturing companies were ambushed by the Lemelson tax, and ended up having to lay people off in order to carry on their business - and got nothing in return.
At one level Microsoft has so much cash in the bank that it could live off the hump for years and years and years. They identified the problem: lack of a recurring revenue stream, and the need to sell more OS/Office licences to create revenue.
There are two solutions for this problem: 1) Develop a strong services and solutions offering, where business will trust you with their IT and pay lots of money for good service 2) Invent a way to squeeze recurring revenue out of your installed user base without offering anything substantially more
IBM chose (1), Microsoft chose (2).
Consider the phases of IT: firstly there was the traditional IBM phase where by far the largest cost was hardware, even allowing for teams of people writing in-house software. This characterises the period up to, say, 1980, and by 1990 IBM was almost dead on its feet; secondly there was the phase where commercial packaged software was a major part of IT decision making, starting with putting Lotus 1-2-3 in front of decision makers, and continuing through the Windows/Office age. This phase was characterised by the PHB saying "I want 10,000 computers running Wordperfect and Lotus".
Now we are into the next phase, where both hardware and packaged software are commodities within a solution or service. This is why companies such as EDS, CSC and IBM (and smaller players in this market) matter more than Microsoft. If Ballmer thinks that some new technological gizmo will get people spending again then he's wrong: there may well be a lot of individual buyers for new toys, but neither the business desktop nor my mum need or want a new killer technology. They need, and already have, a working toolset to send email, browse the web (and use web enabled applications) and create documents. Essentially we have now commoditised the information rather than the software (yes, I know this process isn't complete, but it's under way).
Now the good bit: Microsoft has so much cash that it needn't deal with this issue for years yet. IBM got into deep doodoo before reinventing itself. Microsoft is showing the signs that it expects to spend several years yet digging the hole in the same place.
Cheaper than patenting extensions to this, by publishing a set of extensions it makes those extensions unpatentable.
Though you have to publish in a journal which the Patent Office would recognise. Alternatively, you simply write a patent spec for the extensions, pay the initial filing fee, then let the thing lapse - that's enough to establish prior art.
How much does initial patent filing cost with the USPTO? I know that it is a nominal sum at the UK Patent Office.
How are Microsoft's interests best served? Simple: by making sure this suit goes on as long as possible. So this licensing deal is a good cover for them to put money into SCO to delay the point where SCO goes bust and the lawsuit gets rapidly settled by creditors. By toying with SCO in this way, they get to talk about the "impending lawsuit" for longer.
The public comments about IP protection are minor asides: the real value to them is having thousands of sales blokes able to keep repeating "... and the outstanding lawsuit..." every time they have a customer who might use a Linux solution.
Remember, this comes about a week after it came out that MS have directed their sales for "not to lose to Linux at any cost". They will play this for all it's worth - it's like an astroturf campaign which fell into their lap.
The argument about OpenOffice and AbiWord being able to read MS Office files is specious - just because these formats have been partially reverse engineered doesn't alter their being proprietary unpublished formats. OTOH, OO is a completely published format, in effect a XML schema (which is stored in a zipped archive format). But it isn't a case of OO vs MS Word, this is about the whole subject of how machine readable information should be passed between agencies.
The mandating of proprietary unpublished formats by government or its agencies is inexcusable. Document interchange in writeable formats is really, really bad - they should at least be passed in Postscript or PDF (yes, I know they too can be modified, but it takes a bit of effort, and if it really matters then they can be digitally signed), and tables of data can be moved as an ASCII CSV file (again, signed if needed).
This is the most insightful comment I've seen on the subject of software patents. Thinking back to my introduction to patents, for something to be patentable (in the UK at least) it must be:
1. Novel 2. Inventive 3. Capable of physical embodiment.
And over many decades (centuries even) patent practice has developed and matured. The same case can be made of trademark and copyright law - there is a long trail of established case law. This body of case law will help not only in dealing with disputes but also in guiding the patent offices when awarding patents. And not only case law, but maturity in the process of examining and granting patents.
The advent of software patents (in the US, still don't have them over here) is a step change, and introduces the patent process to an arena where there is no case law, and no established maturity in the process of examining and granting patents.
Now, the US patent office could tackle this in two ways: a) they could set the bar for the granting of software patents very high, and themselves get involved in wrangling with big corporations about patents which they have declined, or b) they could just grant any application which comes in, in which case they will not be involved in any disputes between patent holders and alleged infringers.
Whatever the merits of the two cases, it is now too late: there is a large body of software patents which, instead of being use to protect an inventor from having his ideas copied, is used by large corporations to selectively bully other corporations (large and small) in a game of bluffing poker played with legal fees.
The only silver lining is that all patents expire, and being able to cite an expired patent which covers what you're doing is a cast iron defence (assuming you waited until it expired before distributing your version).
And the dark cloud on the horizon? The possibility of patent terms being extended, in the same way as copyright terms, by similarly Mickey Mouse organisations.
No, I find the Slashdot discussions useful. The mainstream media can't give the scientific insight which so many of us crave, and assuming you browse/. at a suitable threshold there are a number of intelligent contributions which stimulate the braincells.
Yesterday (or the day before perhaps) someone posted a link to Feynman's Appendix on the Challenger enquiry about risks being de-emphasised if they had previously not resulted in catastrophe - and there *may* be elements of this flawed analysis involved in the Columbia breakup.
I welcome the opportunity for mainstream news stories to receive the/. treatment, and in this case at this stage starting a new discussion every couple of days provides a refresh to the intelligent discussion.
It is precisely this behaviour from Real which makes me annoyed that the BBC only put content out in Real formats - both audio and video. At the time they made the choice Real was the only mature format available, and they now have such a large installed base of (totally non techie) viewers and listeners with their computers set up for Real that changing would be very disruptive.
I would like to see them make other formats available, but the cost of implemeting streams in multiple formats is probably prohibitive. But it is galling that the Public Service broadcaster, who I and almost everyody in this country is compelled to finance, is channeling millions of customers to an organisation with objectional business practices.
The nub to the key/algorithm business is not necessarily the speed with which it can be changed but the issue of control. As a cryptography user I have no control over the secrecy or otherwise of the algorithm, but when I generate a key pair I am in control of the secrecy of the key.
It would be the fifth - Sunview, NeWS, Openwin, CDE, Gnome.
People tend to forget Sunview because it wasn't X based. Hell, it was kernel based, but it ran reasonably quickly on a 68020 with 4MB of memory across 10Mb ethernet. Sun took their GUI out of the kernel and into user space a few years before Microsoft took their GUI the other way. Go figure.
The arguments about NeWS have been well rehearsed... brilliant innovative technology but Sun kept it proprietary while X was BSD licensed.
Then there was the Openlook vs Motif holy war, during which Scott McNeally was quoted saying Sun would adopt Motif "over my dead body".
As for Gnome, Sun have been putting development effort into Gnome for a couple of years now, working on some of the boring bits. They wouldn't have done this if they didn't intend to use it.
Back in the days before there was a Linux kernel, a lot of people, myself included, were running SunOS4 systems with most of the BSD utilities replaced with GNU versions. Nobody bothers to do this with Solaris nowadays - if you want a GNU based system then you run GNU/Linux.
The big systems in the server room will continue to run Solaris on Sparc - the scalability and RAS features wipe the floor with the alternatives. So you can now expect software developers to need to support two APIs, Solaris and GNU/Linux.
What Sun are doing is to provide choices in how these APIs are made available to users and developers - Solaris on sparc for big iron right down to $2k edge servers, GNU/Linux on Intel for cheap edge servers (with support), Solaris on x86 with a GNU/Linux API utility for mixed mode users.
And providing an easily installable set of GNU replacement utilities takes us back to where we came in. All the end user has to do is choose their option.
The usage of language by people *is* a major issue.
RMS is absolutely correct to object, as I do, to the terms which large companies use to describe the act of illegal copying: Piracy, a violent crime which still goes on in some parts of the world, often resulting in the death of its victims; Theft, a crime where the victim has a piece of property removed and is no longer able to use it.
If I copy a piece of software, or a music CD, it doesn't result in anyone's death, nor does the person I copy it from no longer have the music. What does happen is two things:
a) If the copy I make results in my not buying a royalty bearing copy (from a shop) then the producer of that software/music has lost revenue. If I make a copy rather than just not using the software or listenin to the music then there hasn't been a loss of revenue
b) I have broken the law
Hence the term "Unauthorised copying" or "Illegal copying" would appear to be the most correct term to describe this activity.
I also take issue with those who describe the GPL as "viral". In the case of a computer virus, the significant feature is that it runs on your computer without your censent. On the other hand, if you incorporate some GPL licensed code into your software product you can only do this by taking action to do so. There is no difference, for example, to incorporating a piece of code from another party where the terms of use require you to pay them a royalty. The GPL simply has a different set of conditions attaching should you choose to incorporate it into your software - and in both cases if you don't like the attached conditions then you should simply not incorporate that software.
Even in its heyday, IBM made its profits from selling goods, so it couldn't generate the obscene amounts of cash which MS is sitting on. The scandal is that MS has made monopoly profits from software, where the direct cost of sale is almost zero, and rather than doing anything with the money or paying it to shareholders in dividends (and hence paying taxes like the rest of us) they have amassed a cash mountain. So while IBM has bought PwC Consulting, MSFT has *chosen* not to.
So on Microsoft's part, this displays either a different plan or (more likely) a lack of vision. Having made easy money from the Windows/Office monopoly, they don't understand how to run a "normal" business where you have to sell goods or services (where the cost of sale is not insignificant compared to the price tag).
As for IBM, while Global Services is a managed service offering which also performs implementations, PwC Consulting will be mostly implementation based. Their plan will be to use the PwC arm to feed ongoing business into GS (which means that when they report the accounts of the two arms separately IBMGS will show impressive growth).
... who were trialing Ogg earlier this year (and still have ogg feeds up from time to time). They have taken stick in the GB press for adopting Real as their primary format for streaming (because Real player decides to take over granny's computer, etc.).
If all the people who've sacrificed their computers to Real Player could listen to ogg feeds with just an automatic codec update, then it would strengthen the case of those inside the Beeb promoting "free" formats.
The issue with oil is the cost of extraction relative to the market price. If supplies start to dwindle then the price will rise, and supplies which are currently too expensive to be viable will come online.
There's plenty of oil out there, but why mess about in inhospitable places if we can just stick a pipe into the ground in the Middle East.
Slashdot Mirror
It is with the advent of the Internet that work which is in the public domain can now be made widely available. Despite the (IMHO) excessively long copyright terms, much classical literature and music is in the public domain, and this is added to every year.
Do you see safeguarding the public domain as part of your role?
Dunstan
I've seen previous Q&A sessions to this where the consequences of false accusations were played down "a tiny proportion" ... "mistakes immediately rectified".
....SRPMS/OpenOffice.... contains the strings MS and Office doesn't constitute an investigation - indeed, if a human being versed in the art had read the filenames, let alone reviewed the file contents, they would have realised that the notice should not have been sent.
But such platitudes miss the point which is this: the point about such accusations being made under penalty of perjury is that the infringement should be investigated *before* the accusation is made. For example, sending a notice about the Source code for Open Office because
It is not acceptable to dismiss this and other cases as rare and isolated incidents: the need for *all* such accusations to be substantiated is the whole point behind the "under penalty of perjury" thing, and organisations which send such notices out without proper review should be prosecuted for perjury. Yet it would appear that these serious allegations are made with "grep | mail".
Actually true "pirates" kill the crew of ships, and steal the vessel and cargo. This violent crime is still very much part of life for seafarers in some parts of the world today.
What you are referring to is large scale commercial illegal copying.
Dunstan
Apologies for responding to a "more than GNU and Linux troll" but ...
The important parts of a GNU/Linux system are almost all either GNU or Linux. The kernel is Linux. The compiler is GNU, the linker is GNU, the C libraries are GNU. Most of the core utilities are GNU. Indeed, the first part of any port of GNU/Linux to a different OS is *not* work on the kernel, but work on GCC.
Very few of the applications run on GNU/Linux systems are GNU, as has correctly been identified, just as many Solaris boxes run Oracle. But the bit which makes a running system is almost entirely GNU and Linux.
Dunstan - using GNU software since Linus Torvalds was still at school.
OK, I'll villify Lemelson.
He filed, as has been stated, very general patents a long time ago. He then used a loophole in patent law which is peculiar to the USA. While a patent in any country usually runs for a number of years from the date of initial filing (typically 20 years), the USPTO allowed an amended filing to get a new filing date, rather than the original one. By repeatedly filing amendments, the patents can last far longer than the intended term.
Lemelson used this technique to run patents from the 1960s forward to the 90s, and at a time of his choosing issued "cease and desist" letters to almost everybody who used barcodes for anything. Given the possibility of triple damages if they continued using barcodes without a licence, and given their reliance on barcodes in their processes, just about every recipient of the letter paid money for a licence.
The technique used was to pervert a provision intended to allow an inventor to refine his ideas and instead exploit it for entirely litigious and not at all technical reasons. To pervert such a concession and succeed in legally extorting over $1billion from the wealth creating sector of the economy is repulsive.
We may often hear the Microsoft tax being referred to in this forum, but when you pay the MS tax you get something in return - an operating system which is not very good but adequate for many purposes. In contrast, many manufacturing companies were ambushed by the Lemelson tax, and ended up having to lay people off in order to carry on their business - and got nothing in return.
Don Corleone would have been proud of Lemelson.
Dunstan
At one level Microsoft has so much cash in the bank that it could live off the hump for years and years and years. They identified the problem: lack of a recurring revenue stream, and the need to sell more OS/Office licences to create revenue.
There are two solutions for this problem:
1) Develop a strong services and solutions offering, where business will trust you with their IT and pay lots of money for good service
2) Invent a way to squeeze recurring revenue out of your installed user base without offering anything substantially more
IBM chose (1), Microsoft chose (2).
Consider the phases of IT: firstly there was the traditional IBM phase where by far the largest cost was hardware, even allowing for teams of people writing in-house software. This characterises the period up to, say, 1980, and by 1990 IBM was almost dead on its feet; secondly there was the phase where commercial packaged software was a major part of IT decision making, starting with putting Lotus 1-2-3 in front of decision makers, and continuing through the Windows/Office age. This phase was characterised by the PHB saying "I want 10,000 computers running Wordperfect and Lotus".
Now we are into the next phase, where both hardware and packaged software are commodities within a solution or service. This is why companies such as EDS, CSC and IBM (and smaller players in this market) matter more than Microsoft. If Ballmer thinks that some new technological gizmo will get people spending again then he's wrong: there may well be a lot of individual buyers for new toys, but neither the business desktop nor my mum need or want a new killer technology. They need, and already have, a working toolset to send email, browse the web (and use web enabled applications) and create documents. Essentially we have now commoditised the information rather than the software (yes, I know this process isn't complete, but it's under way).
Now the good bit: Microsoft has so much cash that it needn't deal with this issue for years yet. IBM got into deep doodoo before reinventing itself. Microsoft is showing the signs that it expects to spend several years yet digging the hole in the same place.
Dunstan
Cheaper than patenting extensions to this, by publishing a set of extensions it makes those extensions unpatentable.
Though you have to publish in a journal which the Patent Office would recognise. Alternatively, you simply write a patent spec for the extensions, pay the initial filing fee, then let the thing lapse - that's enough to establish prior art.
How much does initial patent filing cost with the USPTO? I know that it is a nominal sum at the UK Patent Office.
Dunstan
Piracy is a violent crime which continues in many parts of the world, often resulting in the death of its victims.
What the BSA are talking about here is the illegal copying of software.
Dunstan
Looks as if they missed the statistics lesson where you learn the difference between a correlation and a cause.
Dunstan
How are Microsoft's interests best served? Simple: by making sure this suit goes on as long as possible. So this licensing deal is a good cover for them to put money into SCO to delay the point where SCO goes bust and the lawsuit gets rapidly settled by creditors. By toying with SCO in this way, they get to talk about the "impending lawsuit" for longer.
... and the outstanding lawsuit ..." every time they have a customer who might use a Linux solution.
The public comments about IP protection are minor asides: the real value to them is having thousands of sales blokes able to keep repeating "
Remember, this comes about a week after it came out that MS have directed their sales for "not to lose to Linux at any cost". They will play this for all it's worth - it's like an astroturf campaign which fell into their lap.
Dunstan
The argument about OpenOffice and AbiWord being able to read MS Office files is specious - just because these formats have been partially reverse engineered doesn't alter their being proprietary unpublished formats. OTOH, OO is a completely published format, in effect a XML schema (which is stored in a zipped archive format). But it isn't a case of OO vs MS Word, this is about the whole subject of how machine readable information should be passed between agencies.
The mandating of proprietary unpublished formats by government or its agencies is inexcusable. Document interchange in writeable formats is really, really bad - they should at least be passed in Postscript or PDF (yes, I know they too can be modified, but it takes a bit of effort, and if it really matters then they can be digitally signed), and tables of data can be moved as an ASCII CSV file (again, signed if needed).
Dunstan - if it ain't plain it ain't text
Mod parent as insightful
This is the most insightful comment I've seen on the subject of software patents. Thinking back to my introduction to patents, for something to be patentable (in the UK at least) it must be:
1. Novel
2. Inventive
3. Capable of physical embodiment.
And over many decades (centuries even) patent practice has developed and matured. The same case can be made of trademark and copyright law - there is a long trail of established case law. This body of case law will help not only in dealing with disputes but also in guiding the patent offices when awarding patents. And not only case law, but maturity in the process of examining and granting patents.
The advent of software patents (in the US, still don't have them over here) is a step change, and introduces the patent process to an arena where there is no case law, and no established maturity in the process of examining and granting patents.
Now, the US patent office could tackle this in two ways:
a) they could set the bar for the granting of software patents very high, and themselves get involved in wrangling with big corporations about patents which they have declined, or
b) they could just grant any application which comes in, in which case they will not be involved in any disputes between patent holders and alleged infringers.
Whatever the merits of the two cases, it is now too late: there is a large body of software patents which, instead of being use to protect an inventor from having his ideas copied, is used by large corporations to selectively bully other corporations (large and small) in a game of bluffing poker played with legal fees.
The only silver lining is that all patents expire, and being able to cite an expired patent which covers what you're doing is a cast iron defence (assuming you waited until it expired before distributing your version).
And the dark cloud on the horizon? The possibility of patent terms being extended, in the same way as copyright terms, by similarly Mickey Mouse organisations.
Dunstan
No, I find the Slashdot discussions useful. The mainstream media can't give the scientific insight which so many of us crave, and assuming you browse /. at a suitable threshold there are a number of intelligent contributions which stimulate the braincells.
/. treatment, and in this case at this stage starting a new discussion every couple of days provides a refresh to the intelligent discussion.
Yesterday (or the day before perhaps) someone posted a link to Feynman's Appendix on the Challenger enquiry about risks being de-emphasised if they had previously not resulted in catastrophe - and there *may* be elements of this flawed analysis involved in the Columbia breakup.
I welcome the opportunity for mainstream news stories to receive the
Dunstan
It is precisely this behaviour from Real which makes me annoyed that the BBC only put content out in Real formats - both audio and video. At the time they made the choice Real was the only mature format available, and they now have such a large installed base of (totally non techie) viewers and listeners with their computers set up for Real that changing would be very disruptive.
I would like to see them make other formats available, but the cost of implemeting streams in multiple formats is probably prohibitive. But it is galling that the Public Service broadcaster, who I and almost everyody in this country is compelled to finance, is channeling millions of customers to an organisation with objectional business practices.
Dunstan
The nub to the key/algorithm business is not necessarily the speed with which it can be changed but the issue of control. As a cryptography user I have no control over the secrecy or otherwise of the algorithm, but when I generate a key pair I am in control of the secrecy of the key.
Dunstan
One application for these is that you enable RF on your fridge and hey presto, when you're running out of beer your fridge orders some more.
Dunstan
It would be the fifth - Sunview, NeWS, Openwin, CDE, Gnome.
... brilliant innovative technology but Sun kept it proprietary while X was BSD licensed.
People tend to forget Sunview because it wasn't X based. Hell, it was kernel based, but it ran reasonably quickly on a 68020 with 4MB of memory across 10Mb ethernet. Sun took their GUI out of the kernel and into user space a few years before Microsoft took their GUI the other way. Go figure.
The arguments about NeWS have been well rehearsed
Then there was the Openlook vs Motif holy war, during which Scott McNeally was quoted saying Sun would adopt Motif "over my dead body".
As for Gnome, Sun have been putting development effort into Gnome for a couple of years now, working on some of the boring bits. They wouldn't have done this if they didn't intend to use it.
Dunstan
Back in the days before there was a Linux kernel, a lot of people, myself included, were running SunOS4 systems with most of the BSD utilities replaced with GNU versions. Nobody bothers to do this with Solaris nowadays - if you want a GNU based system then you run GNU/Linux.
The big systems in the server room will continue to run Solaris on Sparc - the scalability and RAS features wipe the floor with the alternatives. So you can now expect software developers to need to support two APIs, Solaris and GNU/Linux.
What Sun are doing is to provide choices in how these APIs are made available to users and developers - Solaris on sparc for big iron right down to $2k edge servers, GNU/Linux on Intel for cheap edge servers (with support), Solaris on x86 with a GNU/Linux API utility for mixed mode users.
And providing an easily installable set of GNU replacement utilities takes us back to where we came in. All the end user has to do is choose their option.
Dunstan
The usage of language by people *is* a major issue.
RMS is absolutely correct to object, as I do, to the terms which large companies use to describe the act of illegal copying: Piracy, a violent crime which still goes on in some parts of the world, often resulting in the death of its victims; Theft, a crime where the victim has a piece of property removed and is no longer able to use it.
If I copy a piece of software, or a music CD, it doesn't result in anyone's death, nor does the person I copy it from no longer have the music. What does happen is two things:
a) If the copy I make results in my not buying a royalty bearing copy (from a shop) then the producer of that software/music has lost revenue. If I make a copy rather than just not using the software or listenin to the music then there hasn't been a loss of revenue
b) I have broken the law
Hence the term "Unauthorised copying" or "Illegal copying" would appear to be the most correct term to describe this activity.
I also take issue with those who describe the GPL as "viral". In the case of a computer virus, the significant feature is that it runs on your computer without your censent. On the other hand, if you incorporate some GPL licensed code into your software product you can only do this by taking action to do so. There is no difference, for example, to incorporating a piece of code from another party where the terms of use require you to pay them a royalty. The GPL simply has a different set of conditions attaching should you choose to incorporate it into your software - and in both cases if you don't like the attached conditions then you should simply not incorporate that software.
Dunstan
The message which should be taken away form this is: The BBC is absolutely dependent on standards, and hence uses Free Software to ensure compliance.
Dunstan
Or you just browse through the proxomitron http://home.arcor.de/six/
Dunstan
Even in its heyday, IBM made its profits from selling goods, so it couldn't generate the obscene amounts of cash which MS is sitting on. The scandal is that MS has made monopoly profits from software, where the direct cost of sale is almost zero, and rather than doing anything with the money or paying it to shareholders in dividends (and hence paying taxes like the rest of us) they have amassed a cash mountain. So while IBM has bought PwC Consulting, MSFT has *chosen* not to.
So on Microsoft's part, this displays either a different plan or (more likely) a lack of vision. Having made easy money from the Windows/Office monopoly, they don't understand how to run a "normal" business where you have to sell goods or services (where the cost of sale is not insignificant compared to the price tag).
As for IBM, while Global Services is a managed service offering which also performs implementations, PwC Consulting will be mostly implementation based. Their plan will be to use the PwC arm to feed ongoing business into GS (which means that when they report the accounts of the two arms separately IBMGS will show impressive growth).
Dunstan
... who were trialing Ogg earlier this year (and still have ogg feeds up from time to time). They have taken stick in the GB press for adopting Real as their primary format for streaming (because Real player decides to take over granny's computer, etc.).
If all the people who've sacrificed their computers to Real Player could listen to ogg feeds with just an automatic codec update, then it would strengthen the case of those inside the Beeb promoting "free" formats.
Dunstan
The issue with oil is the cost of extraction relative to the market price. If supplies start to dwindle then the price will rise, and supplies which are currently too expensive to be viable will come online.
There's plenty of oil out there, but why mess about in inhospitable places if we can just stick a pipe into the ground in the Middle East.
Dunstan