It's an issue until market penetration occurs. I doubt the BBC have the resources to do the customer support needed for such less well-propagated formats, even if they're superior. And I think for right now, they're unfortunately committed to the idea of DRM in order to keep their contracts with actors and producers, etc.
It's a shame, really. A media company the size of the BBC could be wonderful at promoting a newer and less DRM-crippled standard, if they had the chutzpah and short-term funding to do it. As I understand it, though, they've just been very disappointed by the adjustments to the television tax (whoops, I mean licensing fee!).
Frankly, expect some smart alecks to be un-DRM'ing the BBC's content and publishing it via www.thepiratebay.org within minutes of each day's or week's programs being published. I'm afraid they've sold their souls to DRM for contractual reasons, and don't realize how quickly it will be penetrated.
No, there is a tax called a "license fee". It's hardly free.
Circumvention also hardly calls for intelligence. The workaround is called www.thepiratebay.org. It only takes one person with a slight bit of intelligence to circumvent the DRM, and then it's downloadable by anyone.
It could be both DRM, and the documented difficulty of supporting multiple media formats. They easily triple your storage and support costs to support Realtime and Quicktime as well, and nothing else has the market penetration. And unless you've actually tried to *install* some of the amazing "Linux server" pieces of festering gob-shite (to borrow a British friend's phrase), I suggest you not underestimate the pain and cost of doing so.
If you don't believe I'm serious about the difficulty, go try to gracefully install the latest "Helix Server" from Real. It's pretty awful to install, and worse to maintain. The only way to do so is to find a pirate copy that has a competent installer wrapped around it, much as the only graceful way to insall NVidia drivers under RedHat operating systems is to get them from the Livna repository in a non-license-compatible format.
Not really. There has to be a quick *patent application*. Violations of the patent date from the submission of the patent. If the fix was applied before the patent was applied, that would be prior art. The patent system recognizes that patents can be violated while the patent is pending, and that the vioilation can be addressed after the patent is finally granted. (I say this as someone who's helped establish prior art and helped establish that similar technologies are not described by the same patent, not as a lawyer.)
There sometimes *is* going back. But it's very painful, very expensive in money and manpower and political capital, and usually occurs only when the original winner proves themselves to be grossly, grossly incompetent.
Ending that kind of relationship is like ending a marriage: people keep saying "what about the customers?" instead of "what about the children?" And getting the agreement cancelled entirely is like getting a marriage annulled, you have to prove pretty serious fraud in the first place.
Sir, being trolled does not mean your comments are unreasonable.
It means that someone has successfully gotten you to waste time or become upset by pretending to be serious about something outrageous. The more they can make you froth and become upset, especially if this can then be used to discredit your other statements to outside witnesses of the discussion, the more effective the troll has been.
Or am I being successfully trolled? It can become difficult to keep track in such a wide-open forum.
Agreed. I've seen critical patches held up for up to a year because the "QA" process was not allowed to test and approve individual components for release, but had to hold up or roll back the entire software release if any individual component failed. There are uses for monolithic development models, but for large projects it's quite crippling. Developers get frustrated and lose the ability to share their work with others. Managers who don't always have the technical expertise needed wind up being gatekeepers on both development and communications.
I'll be delighted to test out git and see if it works well, simply based on Linus's very strong recommendation. The branching model is particularly attractive for live redundant repositories.
You, sir, have apparently been successfully trolled.
But what makes CVS and now SVN usable are decent interfaces: viewvc for repository browsing and tarball downloading, TortoiseCVS and TortoiseSVN for use by people who are not command line gurus. I'll be delighted to see git have similarly good tools, especially for Linux or UNIX work.
In fact, if someone knows a good GUI for Subversion on Linux or UNIX, and knows of good ones for git or mercurial, I'd be delighted to point my acquaintances who deal with projects using them to those interfaces.
You're fairly lucky. These days, the "disassembly" goes on at resource allocation and funding time. Your equipment money goes to buy a professor a laptop, or to send some favored colleague to a conference to do a presentation and keep them on the tenure track, and you never even get to hear about it (Unless you're like me and had to fix their email, meaning they sent you the bounce messages or gave you deliberate access to their mail spool to fix it.)
I hope Lederman did help the student out later, I really do.
True. But it's a lesson to remember: stealing research and materials from graduate students is far, far too common a practice. And it teaches the grad students to themselves steal research and materials from the next generation. I've had too many friends whose research was stolen by advisors or whose experiments were ruined by another professor in the same group "appropriating" their equipment, reserved laboratory time on expensive systems, or even their funding.
Fascinating science with poor equipment is great to do: but stealing equipment should not be encouraged.
Re:Changes over time?
on
MacGyver Physics
·
· Score: 4, Insightful
You don't have to repeat it as many times as possible! That's just wasting time and money. Doing experiments with variations, to confirm what the limits of the theory are and testing related hypotheses is much more effective.
The real reason undergraduates get those classic experiments is to teach them how to do experiments, the limits of their instruments, how to record all relevant data, the difference between accuracy and precision, etc. The big experiment being done is actually on the students themselves, to see if they've learned to do reliable experiments. You absolutely do not want to do sensitive experiments with students whose reliability and even whose honesty have not yet been tested in lab work with known expected results.
Re:Changes over time?
on
MacGyver Physics
·
· Score: 2, Interesting
It does make me wonder what else they cannibalized from other people's work. Perhaps a review of these gentlemen's papers for plagiarism is in order? Or perhaps the grad student should keep an eye on their fiscal behavior and rat them out to the IRS?
Just take away all the dialogue from Lucas. From Jar-Jar Binks needing to light up some big ganja to complete his ethnic caricature, to the "Fu Manchu Menace" of the Trade Federation accents, to taking the amazingly hot and intelligent and skillful character of Padma and turning her into 15 year-old's dream of how princesses would worship their manly manhood while losing every characteristic that made them hot, Lucas's dialogue in the prequels was an absolute failure.
Replacing all the dialogue with old "Bonanza" episodes would be a noticeable improvement.
Yes, I find it very hepful. There are only a few things that really care about atime (such as expiring and deleting files based on last access). For web servers, it's a fairly useless tool.
You've apparently never run into software where version x.y.z of a library from one OS has significant differences from the same library of a different OS. I have: Oracle, in particular, was sensitive to this sort of subtle library difference between Solaris and Linux releases. And libraries and compilers have often had subtle differences between different releases of the same OS.
It's nowhere as bad as Microsoft DLL hell, where the libraries wind up scattered throughout the operating system and overwrite each other without notification or control or the ability to reverse things. But it's been particularly bad with Oracle and its sensitivities, and isn't much better with Java based tools. The particular Java variant, and installing multiple versions gracefully on one system, has been quite painful.
I'm sorry, but the number of times I've seen the "Write Once, Run Anywhere" claim made for new "paradigms" is fairly scary. There are limitiations to all of these approaches. In this case, running software in complete OS emulation mode denies access to hardware features that have not been successfully ported to the virtual environment, enforces limits of the particular underlying VM hosting operating system in fascinating ways, and absolutely punishes the performance of any disk-accessing operations.
There are uses for virtualized environments, but they're hardly a new approach to code portability.
Which Microsoft vulnerability are you referring to as being over 10 years old? CERT and similar vulnerability report sites are not useful this way, because they don't publish the existence of the problem without explicit permission from the software manufacturer. So I've seen vulnerability reports held for over a year by CERT, until Microsoft got around to fixing it. So the apparent "window of vulnerability" was only a few weeks from the finally permitted CERT publication, and the patch being part of the standard Microsoft updates. But the actual vulnerability I saw lasted more than a year.
Just don't wait too long to do your patches or upgrades. I run into far, far too many old systems that are RedHat 7.2 still running critical applications that are no longer supported, or Win98 boxes with data that the owners have never properly backed up or ported ot a contemporary OS. (I shut down the last Win98 box in a large environment last year by bribing the owner of it with free long-distance calls to their kids with Skype on the new system.)
Fortunately, installing critical patches has gotten far easier than it used to be. Patching libc on UNIX systems, or patching sendmail, used to be a nightmare of reconfiguration and reboots from bootstrap tapes. The pain and risk of doing core component updates were part of why the Morris Worm, in 1988, took down so much of the UNIX infrastructure of the time.
These days, they update gracefully in place without even rebooting. The *only* thing that absolutely needs rebooting to install in any OS is the kernel: for other components it's sometimes safer to reboot to make sure everything gets in play, but usually unnecessary.
You just got your Microsoft Certified Software Engineer coupon, didn't you?
I'm sorry, but you need to really do some looking and poking to see just why Windows OS's are traditionally vulnerable. From the oddness needed to allow graphics manipulation for new hardware features for high end games, to the incredibly badly done security models of Internet Explorer, to the unmanageable software installations and cooperation of setting up root kits for DRM purposes, to the foolishness of auto-opening attachments in email and auto-executing CD's without providing user control, etc., etc., etc., Windows has done a major disservice to its customers.
No, I've actually written funding proposals for corporate technological shifts. Every single point I've raised has been one championed by people actually doing the work.
Adding IPv6 to a network requires, for example, changing the databases and software that manage your resource allocation, hardware ID's, and machine setups. That's neither the kind of cocktail napkin work you describe nor things you can hand to the Microsoft Certified Network Engineers hired as monkey boys. You need people with clues to assess existing infrastructure and migrate it gracefully. Such people are expensive.
For example of a hidden cost of the switchover, look at the RBL systems for blocking spam. IPv6 ruins a lot of its usability. Burdening the spam filters with the switchover is just asking to get a lot of calls and complaints about a system that provides little practical benefit to the end user for the next few budget years. And switchovers of that scale are painful: they interrupt services, and imperial Service Level Agreement contracts. Unless you get a critical mass of people switching over for other reasons, they're just going to stay with IPv4 and better NAT use.
Look, IPv6 has a lot of technical benefits, and in the long term, as household network capable nodes become even more ubiquitous, we will eventually feel the crunch for address space severely. But the pressure just isn't there yet, and just doesn't justify spending the man-hours and budget for it.
Stallman has enough handlers to help keep him out of the realm of shrieking in your face. He's actually matured, and gotten better at convincing strangers that he has some real answers for their concerns.
Richard has been prophetic about this: his concerns at the Novell/Microsoft deal, and about software patents in general, were exposed as completely correct when Microsoft started its recent claims of hundreds of patent violations without naming a single patent.
Samba is an interesting issue. Jeremy Allison resigned from Novell as fast as feasible: as one of the core developers of Samba, I have to assume that Jeremy's work was a big factor in the Microsoft/Novell deal. And with all the little network applicance and storage tools currently on the market, and with almost all of them being Samba based, Microsoft had clear reason to limit the usability of Jeremy's work to direct Novell customers.
Remember also that Novell's Netware tools are still supported: network storage is a big market. And this deal left Microsoft and Novell with a big patent club to be able to use against mutual competitors, without either being restrained by the other company.
Well, given that the British are more video monitored than any country in the world, I think the idea that somehow there is no government snooping there is pretty confused. And have you ever examined how the Value Added Tax records can be used to monitor private business dealings and personal purchases?
Oh, I've gotten a fair chunk of change for selling off a/16. We retained ownership of the domain, but administered for another company that had far more need for it, and stuffed our usage into/28 before I was done. It actually paid for some very useful storage upgrades for us, we got off-site DNS service we didn't have to pay for, and everyone benefitted.
Slashdot Mirror
It's an issue until market penetration occurs. I doubt the BBC have the resources to do the customer support needed for such less well-propagated formats, even if they're superior. And I think for right now, they're unfortunately committed to the idea of DRM in order to keep their contracts with actors and producers, etc.
It's a shame, really. A media company the size of the BBC could be wonderful at promoting a newer and less DRM-crippled standard, if they had the chutzpah and short-term funding to do it. As I understand it, though, they've just been very disappointed by the adjustments to the television tax (whoops, I mean licensing fee!).
Frankly, expect some smart alecks to be un-DRM'ing the BBC's content and publishing it via www.thepiratebay.org within minutes of each day's or week's programs being published. I'm afraid they've sold their souls to DRM for contractual reasons, and don't realize how quickly it will be penetrated.
No, there is a tax called a "license fee". It's hardly free.
Circumvention also hardly calls for intelligence. The workaround is called www.thepiratebay.org. It only takes one person with a slight bit of intelligence to circumvent the DRM, and then it's downloadable by anyone.
It could be both DRM, and the documented difficulty of supporting multiple media formats. They easily triple your storage and support costs to support Realtime and Quicktime as well, and nothing else has the market penetration. And unless you've actually tried to *install* some of the amazing "Linux server" pieces of festering gob-shite (to borrow a British friend's phrase), I suggest you not underestimate the pain and cost of doing so.
If you don't believe I'm serious about the difficulty, go try to gracefully install the latest "Helix Server" from Real. It's pretty awful to install, and worse to maintain. The only way to do so is to find a pirate copy that has a competent installer wrapped around it, much as the only graceful way to insall NVidia drivers under RedHat operating systems is to get them from the Livna repository in a non-license-compatible format.
Not really. There has to be a quick *patent application*. Violations of the patent date from the submission of the patent. If the fix was applied before the patent was applied, that would be prior art. The patent system recognizes that patents can be violated while the patent is pending, and that the vioilation can be addressed after the patent is finally granted. (I say this as someone who's helped establish prior art and helped establish that similar technologies are not described by the same patent, not as a lawyer.)
There sometimes *is* going back. But it's very painful, very expensive in money and manpower and political capital, and usually occurs only when the original winner proves themselves to be grossly, grossly incompetent.
Ending that kind of relationship is like ending a marriage: people keep saying "what about the customers?" instead of "what about the children?" And getting the agreement cancelled entirely is like getting a marriage annulled, you have to prove pretty serious fraud in the first place.
Sir, being trolled does not mean your comments are unreasonable.
It means that someone has successfully gotten you to waste time or become upset by pretending to be serious about something outrageous. The more they can make you froth and become upset, especially if this can then be used to discredit your other statements to outside witnesses of the discussion, the more effective the troll has been.
Or am I being successfully trolled? It can become difficult to keep track in such a wide-open forum.
Agreed. I've seen critical patches held up for up to a year because the "QA" process was not allowed to test and approve individual components for release, but had to hold up or roll back the entire software release if any individual component failed. There are uses for monolithic development models, but for large projects it's quite crippling. Developers get frustrated and lose the ability to share their work with others. Managers who don't always have the technical expertise needed wind up being gatekeepers on both development and communications.
I'll be delighted to test out git and see if it works well, simply based on Linus's very strong recommendation. The branching model is particularly attractive for live redundant repositories.
You, sir, have apparently been successfully trolled.
But what makes CVS and now SVN usable are decent interfaces: viewvc for repository browsing and tarball downloading, TortoiseCVS and TortoiseSVN for use by people who are not command line gurus. I'll be delighted to see git have similarly good tools, especially for Linux or UNIX work.
In fact, if someone knows a good GUI for Subversion on Linux or UNIX, and knows of good ones for git or mercurial, I'd be delighted to point my acquaintances who deal with projects using them to those interfaces.
Or the "secret sauce" for Linus's work is actually soap and water?
You're fairly lucky. These days, the "disassembly" goes on at resource allocation and funding time. Your equipment money goes to buy a professor a laptop, or to send some favored colleague to a conference to do a presentation and keep them on the tenure track, and you never even get to hear about it (Unless you're like me and had to fix their email, meaning they sent you the bounce messages or gave you deliberate access to their mail spool to fix it.)
I hope Lederman did help the student out later, I really do.
True. But it's a lesson to remember: stealing research and materials from graduate students is far, far too common a practice. And it teaches the grad students to themselves steal research and materials from the next generation. I've had too many friends whose research was stolen by advisors or whose experiments were ruined by another professor in the same group "appropriating" their equipment, reserved laboratory time on expensive systems, or even their funding.
Fascinating science with poor equipment is great to do: but stealing equipment should not be encouraged.
You don't have to repeat it as many times as possible! That's just wasting time and money. Doing experiments with variations, to confirm what the limits of the theory are and testing related hypotheses is much more effective.
The real reason undergraduates get those classic experiments is to teach them how to do experiments, the limits of their instruments, how to record all relevant data, the difference between accuracy and precision, etc. The big experiment being done is actually on the students themselves, to see if they've learned to do reliable experiments. You absolutely do not want to do sensitive experiments with students whose reliability and even whose honesty have not yet been tested in lab work with known expected results.
It does make me wonder what else they cannibalized from other people's work. Perhaps a review of these gentlemen's papers for plagiarism is in order? Or perhaps the grad student should keep an eye on their fiscal behavior and rat them out to the IRS?
Just take away all the dialogue from Lucas. From Jar-Jar Binks needing to light up some big ganja to complete his ethnic caricature, to the "Fu Manchu Menace" of the Trade Federation accents, to taking the amazingly hot and intelligent and skillful character of Padma and turning her into 15 year-old's dream of how princesses would worship their manly manhood while losing every characteristic that made them hot, Lucas's dialogue in the prequels was an absolute failure.
Replacing all the dialogue with old "Bonanza" episodes would be a noticeable improvement.
Yes, I find it very hepful. There are only a few things that really care about atime (such as expiring and deleting files based on last access). For web servers, it's a fairly useless tool.
You've apparently never run into software where version x.y.z of a library from one OS has significant differences from the same library of a different OS. I have: Oracle, in particular, was sensitive to this sort of subtle library difference between Solaris and Linux releases. And libraries and compilers have often had subtle differences between different releases of the same OS.
It's nowhere as bad as Microsoft DLL hell, where the libraries wind up scattered throughout the operating system and overwrite each other without notification or control or the ability to reverse things. But it's been particularly bad with Oracle and its sensitivities, and isn't much better with Java based tools. The particular Java variant, and installing multiple versions gracefully on one system, has been quite painful.
I'm sorry, but the number of times I've seen the "Write Once, Run Anywhere" claim made for new "paradigms" is fairly scary. There are limitiations to all of these approaches. In this case, running software in complete OS emulation mode denies access to hardware features that have not been successfully ported to the virtual environment, enforces limits of the particular underlying VM hosting operating system in fascinating ways, and absolutely punishes the performance of any disk-accessing operations.
There are uses for virtualized environments, but they're hardly a new approach to code portability.
Which Microsoft vulnerability are you referring to as being over 10 years old? CERT and similar vulnerability report sites are not useful this way, because they don't publish the existence of the problem without explicit permission from the software manufacturer. So I've seen vulnerability reports held for over a year by CERT, until Microsoft got around to fixing it. So the apparent "window of vulnerability" was only a few weeks from the finally permitted CERT publication, and the patch being part of the standard Microsoft updates. But the actual vulnerability I saw lasted more than a year.
Just don't wait too long to do your patches or upgrades. I run into far, far too many old systems that are RedHat 7.2 still running critical applications that are no longer supported, or Win98 boxes with data that the owners have never properly backed up or ported ot a contemporary OS. (I shut down the last Win98 box in a large environment last year by bribing the owner of it with free long-distance calls to their kids with Skype on the new system.)
Fortunately, installing critical patches has gotten far easier than it used to be. Patching libc on UNIX systems, or patching sendmail, used to be a nightmare of reconfiguration and reboots from bootstrap tapes. The pain and risk of doing core component updates were part of why the Morris Worm, in 1988, took down so much of the UNIX infrastructure of the time.
These days, they update gracefully in place without even rebooting. The *only* thing that absolutely needs rebooting to install in any OS is the kernel: for other components it's sometimes safer to reboot to make sure everything gets in play, but usually unnecessary.
You just got your Microsoft Certified Software Engineer coupon, didn't you?
I'm sorry, but you need to really do some looking and poking to see just why Windows OS's are traditionally vulnerable. From the oddness needed to allow graphics manipulation for new hardware features for high end games, to the incredibly badly done security models of Internet Explorer, to the unmanageable software installations and cooperation of setting up root kits for DRM purposes, to the foolishness of auto-opening attachments in email and auto-executing CD's without providing user control, etc., etc., etc., Windows has done a major disservice to its customers.
No, I've actually written funding proposals for corporate technological shifts. Every single point I've raised has been one championed by people actually doing the work.
Adding IPv6 to a network requires, for example, changing the databases and software that manage your resource allocation, hardware ID's, and machine setups. That's neither the kind of cocktail napkin work you describe nor things you can hand to the Microsoft Certified Network Engineers hired as monkey boys. You need people with clues to assess existing infrastructure and migrate it gracefully. Such people are expensive.
For example of a hidden cost of the switchover, look at the RBL systems for blocking spam. IPv6 ruins a lot of its usability. Burdening the spam filters with the switchover is just asking to get a lot of calls and complaints about a system that provides little practical benefit to the end user for the next few budget years. And switchovers of that scale are painful: they interrupt services, and imperial Service Level Agreement contracts. Unless you get a critical mass of people switching over for other reasons, they're just going to stay with IPv4 and better NAT use.
Look, IPv6 has a lot of technical benefits, and in the long term, as household network capable nodes become even more ubiquitous, we will eventually feel the crunch for address space severely. But the pressure just isn't there yet, and just doesn't justify spending the man-hours and budget for it.
Stallman has enough handlers to help keep him out of the realm of shrieking in your face. He's actually matured, and gotten better at convincing strangers that he has some real answers for their concerns.
Richard has been prophetic about this: his concerns at the Novell/Microsoft deal, and about software patents in general, were exposed as completely correct when Microsoft started its recent claims of hundreds of patent violations without naming a single patent.
Samba is an interesting issue. Jeremy Allison resigned from Novell as fast as feasible: as one of the core developers of Samba, I have to assume that Jeremy's work was a big factor in the Microsoft/Novell deal. And with all the little network applicance and storage tools currently on the market, and with almost all of them being Samba based, Microsoft had clear reason to limit the usability of Jeremy's work to direct Novell customers.
Remember also that Novell's Netware tools are still supported: network storage is a big market. And this deal left Microsoft and Novell with a big patent club to be able to use against mutual competitors, without either being restrained by the other company.
Well, given that the British are more video monitored than any country in the world, I think the idea that somehow there is no government snooping there is pretty confused. And have you ever examined how the Value Added Tax records can be used to monitor private business dealings and personal purchases?
Oh, I've gotten a fair chunk of change for selling off a /16. We retained ownership of the domain, but administered for another company that had far more need for it, and stuffed our usage into /28 before I was done. It actually paid for some very useful storage upgrades for us, we got off-site DNS service we didn't have to pay for, and everyone benefitted.