As you are probably aware, cut-and-paste in Unix/Linux desktop environments have traditionally relied on X to do the dirty work.
The problem with this approach (which Microsoft, Apple, etc. recognized back in the '80s but we still haven't caught up with) is the notion that text is not the only thing that needs to be cut, copied and pasted among and across applications.
Thus, there is inherently a different semantic involved when transferring objects other than text, because X doesn't know how to handle those. X knows nothing of moving a picture between a charting application and a word processor, for instance. Nor should it.
In order to make up for that deficiency, we're faced with a conundrum: do we take the functionality away from X and hand it to the higher-level desktop environment, or do we go with a hybrid approach, letting X still handle the "plain-text" cut-and-paste functionality and letting the desktop environment handle the rest?
I argue that the hybrid approach is terrible for users, as it adds yet another conditional rule users have to follow when trying to complete a task. This, I believe, was what the author was trying to communicate.
Just because your code is stored in CVS does not make it automatically fall under the GPL. That's like saying that because your code was compiled under gcc that it falls under the GPL.
Only if you include CVS code in your project does the GPL apply to it.
RMS forces people to accept the GPL when they run GNU software -- what's the difference?
Tie IP protection to the tax code
on
Fair IP Laws?
·
· Score: 4, Insightful
Intellectual property is a corporate (or individual) asset used to obtain revenue, just like computers, desks, chairs, etc. As such, it is treated as a capital expense under Internal Revenue Code. The full cost of producing the IP must be amortized according to the Code, and amortizations are a time-limited period chosen when the property is first claimed for deduction.
So, why not just tie the protection of said IP to its amortizations? Once the IP has fully capitalized, it should no longer qualify for protection and then fall into the public domain.
This would make IP holders think *very carefully* about how long they want to take to amortize -- it would make them choose a balance between tax savings and IP protection.
In order to DoS Google it doesn't really matter whether you bang on the front door or the back door.
In fact, an attack through the front door will be more likely to succeed because you're hitting the rendering engine, which takes a lot more CPU time (believe it or not) than the search engine.
OTOH the back door is lightweight and is as such advantageous for not only third parties but also Google itself to employ.
Besides, if you're being abused, if you don't want to use technological avenues to keep miscreants away, you can always use legal ones.
Saying LCDs are bad because you can see JPEG compression artifacts is like saying microscopes are bad because you can see germs in them.
JPEG is lossy compression and always has had artifacting. Because you've never seen it before says more about the blurry monitors you're used to than anything else.
Yeah, except Google isn't using Big Box "production" quality servers. They're using the cheapest parts possible -- their software was purposefully designed to be fault-tolerant so that they could do this.
What you name your server doesn't really matter, as long as you have some simply way to map its properties to its name. Having a database around in which the primary key is the hostname and the other columns represent things like location, service, etc. will help you manage these things much more easily.
Management wants to know more than you think they probably do. Back when I was managing servers on several continents, as they were in pairs, I gave them paired names (e.g. ren and stimpy, bonnie and clyde, etc.). It was cute, but as management was getting reports from these servers they were having difficulty keeping track of what they represented. I would often get asked "are bonnie and clyde in London or DC? I can't remember. Will you please make them easier for us to analyze?"
That said, take a good look at your organization and determine who will be analyzing the reports coming out of your servers and what management thinks is the most important thing to know about a server given only its hostname. Is it the location of the server in the racks? Is it the datacenter in which it's housed? Is it the customer to whom the server is leased?
Once you figure out that important bit of data, the rest is just a matter of taste. Here are some examples:
Geographical servers: use the 3-letter airport code, e.g. sfo01, sfo02, etc. If that's further divided by functions, use sfowXX (web), sfomXX (mail), etc.
Rack placement: use the rack number, side (if you're placing servers back to back), and slot location from the bottom up. Example: r17s6. If you're using geographical data centers, use the airport code, e.g. SFOr17s6.
Customer: give each customer a unique four-letter ID that's easy to remember. Granted, you may run into conflicts but some creative naming may help in the matter.
There are already several companies in this space; some have been around for more than two years. Most started out by aggregating dialups, but all recognize that wireless Ethernet is going to reach critical mass at some point in the near future so they've been focusing on it in their business plans:
The proof is in the pudding -- you still have to fsck the filesystem, even with soft updates enabled, if it was left in an inconsistent state before mounting (power off a FreeBSD box without proper shutdown and you'll see).
Not having to fsck is one of the primary advantages of a journalling file system, and soft updates won't give you that, no matter what McKusick says.
Wrong -- prior to the DMCA you only had a right to sue someone for copyright infringement. Thanks to Congress the DMCA has given you extra, heretofore unthinkable rights to control who may analyze, disassemble and modify the bits that represent your work.
Even if the DMCA were repealed (which I hope, though I don't believe, it will be) you wouldn't lose your right to sue for copyright infringement.
Not all of us believe that people should be allowed to benefit from the fruits others' work without compensation (unless, of course, the work is given away). Being anti-DMCA doesn't mean we are anti-copyright. We're merely pointing out the difference between preventing crime (which we don't believe is the state's job in a free society) and prosecuting crime (which is the state's job).
I have no real issues with PPPoE for those who don't mind using it--my Mom uses it because she doesn't need a static IP. The connection time isn't that big a deal, and there's no measurable throughput loss.
My problem is that for security purposes I require a static IP (for example, I only allow FTP connections to my colo'ed server machine from certain IPs), and I purchased service from my value-added ISP under certain terms of service that I pretty much expected would last forever until something better came out.
Now both I and my ISP are getting the rug pulled out under us because of SBC's greediness. None of us would be complaining about the change in TOS if something better was coming along, but instead SBC is fucking us over because they want to push their own content over a line that is supposed to be contain only data I send or request between me and my ISP.
This isn't about address allocation (my ISP has plenty of IP addresses for its customers), this is about corporate greed and taking advantage of a monopoly. I'll say it once or a thousand times; "he who owns the infrastructure will never compete fairly."
Mr. Mundie and the rest of Microsoft have most people chasing a red herring.
The issue is not whether or not Microsoft's software is open source. The issue is not whether or not they like or will ever use the GPL. Being a capitalist entity, neither publishing open source software nor being GPL-friendly should be expected or demanded of them. It's just not logical.
The problem that everyone is concerned about, that prohibits true competition between Microsoft and Open Source developers, is not whether Microsoft's source code is available, but rather whether interoperability can be achieved. What the world needs, and what open source strives to provide, is something that can effectively compete with Microsoft's "standards."
Currently open source software has to compete in a crippled fashion: The developers don't have access to Microsoft's file formats (Word, Excel, etc). They don't have access to the network protocol documentation (e.g. SMB). They don't have access to many hardware forums and manufacturers, because they're not willing to sign NDAs or pay money for the privilege.
What people should be asking for is a level playing field. Making Microsoft open their source code is one way to do it, but it may take away their competitive advantage in other areas (for example, if they build a faster matrix solver) and arguably isn't fair to them. A reasonable alternative, however, would be to ask them to make all of their communication protocols transparent. Once you can interact with other Microsoft users as well as you could if you used Microsoft software yourself, the game is over.
Without such a concession from Microsoft (forced or otherwise), Microsoft's monopoly power will remain unchecked: the Internet will consist of Microsoft clients connecting to Microsoft servers running undocumented application protocols over TCP/UDP. (And you thought forcing HTML email on everyone was bad enough.) If you're smart, you'll either lobby Microsoft (or better, your gov't rep) to make them open their protocols, or you'll buy Microsoft stock.
The total-cost-of-ownership argument on behalf of Linux will disappear (Unix platform vendors like IBM and Sun already offer their Unix OSs at virtually no charge).
What the Gartner Group doesn't understand about TCO (or at least fails to recognize in this report) is that the price of the OS license is only a tiny fraction of TCO. Here's what I think really matters in the calculation of TCO:
Primary documentation. Is the operating system well-documented? Are the manuals (or man pages) accurate, well-written, clear and concise? Linux doesn't score particularly well in the man page department, unfortunately--I think it loses points here. However...
Secondary documentation. How good is the secondary, user-contributed documentation out there relative to other operating systems? Are there well-written, current HOWTOs, user guides, tips and tricks, etc out there? If so, how well are they organized? Linux scores very high here. Other UNIXes are not so lucky. Windows has a lot of users out there--there are a lot of helpful tips on the Net but the documentation is not organized well.
Source code availability. It's an old adage that "the source code is the best documentation," and it's hard to argue otherwise. If I really want to know how a certain function call works, or how the kernel is talking to the hardware, I can dig out the glibc or kernel sources and see for myself, and be 100% certain of the accuracy of my conclusions (well, as long as my understanding of the code is correct). Once again, Linux and other free OSes score high here. With commercial UNIXes and Windows, you have to trust the documentation -- which is often not 100% accurate. In addition, if there's just a bug that needs to be fixed, it's often much easier to fix it yourself if you have access to the source code. Waiting for Microsoft or a commercial UNIX vendor to get a patch out to you can be painfully costly if your product or service's success is dependent on your software vendor's turnaround capability.
In summary, what I think really counts the most towards TCO is the relative understanding of the managers who plan system rollouts and the administrators who maintain them once they're out there. I've seen way too many Windows NT admins out there banging their heads on their desks because something doesn't work the way the documentation says it will, and they have no hope of getting it right because Tech Support doesn't know either. This leads to a lot of costly experimentation (labour is much more expensive than hardware) to make things work.
On the other hand, while Linux documentation isn't always perfect, it's plentiful, it's reasonably organized, and if you just can't find what you need from the HOWTOs or the man pages, you can always fall back on the source code.
rm -rf/* is a poor deletion technique anyways. There are several undocumented 'unrm' utilities out there (I've seen one myself) that are basically fancy implementations of 'dd' that can easily recover data from a UFS or ext2 filesystem.
If you really want to wipe your data clean you should use sdelete (Windows) or secure delete (Unix).
I'm not sure I understand why people are calling it a publicity stunt. As far as I can tell, items placed for auction on eBay are legally binding; pud has to sell his site to the highest bidder. What makes anyone think otherwise--that he's not in fact selling it?
Anyways, good for him! At least someone is making money off his Web endeavours.
Neither a freemail account nor a free web hosting account is truly anonymous. They do retain HTTP access logs, and your ISP will be required to turn over their dialup logs pointing the activity to you personally if they are subpoenaed to do so. Remember Yahoo being required to positively identify the poster of the Apple leaks?
If you want true privacy, sign up at Zero Knowledge Systems and/or use a Mixmaster remailer.
Disclaimer: I am not a lawyer. See a competent attorney in your jurisdiction if you need legal advice.
Beware, folks. This opinion is from someone who apparently is employed by a company under the Universal umbrella. Do a 'whois' on naughtydog.com to see for yourself (they made the Crash Bandicoot series of games for Sony PlayStation).
Besides, whether or not MP3.com really broke the law still seems like a gray area to me. It seems odd to me that the judge didn't even bother to examine the scheme closely and ponder the unintended legal consequences that would result from such a hasty decision like that.
I hope MP3.com appeals. I really want to see a better clarification of what the law allows and what it doesn't, because the District Court is a rather low place on the totem pole.
Slashdot Mirror
I've taken the time to test the ActiveState Komodo tools -- GUI response times are exceedingly slow, especially on Linux.
I think XUL/XPCOM have promise outside of Mozilla but they need serious performance enhancements before I'd consider them mature.
As you are probably aware, cut-and-paste in Unix/Linux desktop environments have traditionally relied on X to do the dirty work.
The problem with this approach (which Microsoft, Apple, etc. recognized back in the '80s but we still haven't caught up with) is the notion that text is not the only thing that needs to be cut, copied and pasted among and across applications.
Thus, there is inherently a different semantic involved when transferring objects other than text, because X doesn't know how to handle those.
X knows nothing of moving a picture between a charting application and a word processor, for instance. Nor should it.
In order to make up for that deficiency, we're faced with a conundrum: do we take the functionality away from X and hand it to the higher-level desktop environment, or do we go with a hybrid approach, letting X still handle the "plain-text" cut-and-paste functionality and letting the desktop environment handle the rest?
I argue that the hybrid approach is terrible for users, as it adds yet another conditional rule users have to follow when trying to complete a task. This, I believe, was what the author was trying to communicate.
Just because your code is stored in CVS does not make it automatically fall under the GPL. That's like saying that because your code was compiled under gcc that it falls under the GPL.
Only if you include CVS code in your project does the GPL apply to it.
RMS forces people to accept the GPL when they run GNU software -- what's the difference?
Intellectual property is a corporate (or individual) asset used to obtain revenue, just like computers, desks, chairs, etc. As such, it is treated as a capital expense under Internal Revenue Code. The full cost of producing the IP must be amortized according to the Code, and amortizations are a time-limited period chosen when the property is first claimed for deduction.
So, why not just tie the protection of said IP to its amortizations? Once the IP has fully capitalized, it should no longer qualify for protection and then fall into the public domain.
This would make IP holders think *very carefully* about how long they want to take to amortize -- it would make them choose a balance between tax savings and IP protection.
In order to DoS Google it doesn't really matter whether you bang on the front door or the back door.
In fact, an attack through the front door will be more likely to succeed because you're hitting the rendering engine, which takes a lot more CPU time (believe it or not) than the search engine.
OTOH the back door is lightweight and is as such advantageous for not only third parties but also Google itself to employ.
Besides, if you're being abused, if you don't want to use technological avenues to keep miscreants away, you can always use legal ones.
Saying LCDs are bad because you can see JPEG compression artifacts is like saying microscopes are bad because you can see germs in them.
JPEG is lossy compression and always has had artifacting. Because you've never seen it before says more about the blurry monitors you're used to than anything else.
Yeah, except Google isn't using Big Box "production" quality servers. They're using the cheapest parts possible -- their software was purposefully designed to be fault-tolerant so that they could do this.
Changes in the comments would have resulted in an SCCS delta. It could have been correct the first time.
This is patently false. The WTC was not designed to withstand a 747 collision; it was considered and deemed to expensive at the time.
Source: Frontline, PBS.
Some thoughts I'd like to share:
What you name your server doesn't really matter, as long as you have some simply way to map its properties to its name. Having a database around in which the primary key is the hostname and the other columns represent things like location, service, etc. will help you manage these things much more easily.
Management wants to know more than you think they probably do. Back when I was managing servers on several continents, as they were in pairs, I gave them paired names (e.g. ren and stimpy, bonnie and clyde, etc.). It was cute, but as management was getting reports from these servers they were having difficulty keeping track of what they represented. I would often get asked "are bonnie and clyde in London or DC? I can't remember. Will you please make them easier for us to analyze?"
That said, take a good look at your organization and determine who will be analyzing the reports coming out of your servers and what management thinks is the most important thing to know about a server given only its hostname. Is it the location of the server in the racks? Is it the datacenter in which it's housed? Is it the customer to whom the server is leased?
Once you figure out that important bit of data, the rest is just a matter of taste. Here are some examples:
Geographical servers: use the 3-letter airport code, e.g. sfo01, sfo02, etc. If that's further divided by functions, use sfowXX (web), sfomXX (mail), etc.
Rack placement: use the rack number, side (if you're placing servers back to back), and slot location from the bottom up. Example: r17s6. If you're using geographical data centers, use the airport code, e.g. SFOr17s6.
Customer: give each customer a unique four-letter ID that's easy to remember. Granted, you may run into conflicts but some creative naming may help in the matter.
The proof is in the pudding -- you still have to fsck the filesystem, even with soft updates enabled, if it was left in an inconsistent state before mounting (power off a FreeBSD box without proper shutdown and you'll see).
Not having to fsck is one of the primary advantages of a journalling file system, and soft updates won't give you that, no matter what McKusick says.
Wrong -- prior to the DMCA you only had a right to sue someone for copyright infringement. Thanks to Congress the DMCA has given you extra, heretofore unthinkable rights to control who may analyze, disassemble and modify the bits that represent your work.
Even if the DMCA were repealed (which I hope, though I don't believe, it will be) you wouldn't lose your right to sue for copyright infringement.
Not all of us believe that people should be allowed to benefit from the fruits others' work without compensation (unless, of course, the work is given away). Being anti-DMCA doesn't mean we are anti-copyright. We're merely pointing out the difference between preventing crime (which we don't believe is the state's job in a free society) and prosecuting crime (which is the state's job).
I have no real issues with PPPoE for those who don't mind using it--my Mom uses it because she doesn't need a static IP. The connection time isn't that big a deal, and there's no measurable throughput loss.
My problem is that for security purposes I require a static IP (for example, I only allow FTP connections to my colo'ed server machine from certain IPs), and I purchased service from my value-added ISP under certain terms of service that I pretty much expected would last forever until something better came out.
Now both I and my ISP are getting the rug pulled out under us because of SBC's greediness. None of us would be complaining about the change in TOS if something better was coming along, but instead SBC is fucking us over because they want to push their own content over a line that is supposed to be contain only data I send or request between me and my ISP.
This isn't about address allocation (my ISP has plenty of IP addresses for its customers), this is about corporate greed and taking advantage of a monopoly. I'll say it once or a thousand times; "he who owns the infrastructure will never compete fairly."
Mr. Mundie and the rest of Microsoft have most people chasing a red herring.
The issue is not whether or not Microsoft's software is open source. The issue is not whether or not they like or will ever use the GPL. Being a capitalist entity, neither publishing open source software nor being GPL-friendly should be expected or demanded of them. It's just not logical.
The problem that everyone is concerned about, that prohibits true competition between Microsoft and Open Source developers, is not whether Microsoft's source code is available, but rather whether interoperability can be achieved. What the world needs, and what open source strives to provide, is something that can effectively compete with Microsoft's "standards."
Currently open source software has to compete in a crippled fashion: The developers don't have access to Microsoft's file formats (Word, Excel, etc). They don't have access to the network protocol documentation (e.g. SMB). They don't have access to many hardware forums and manufacturers, because they're not willing to sign NDAs or pay money for the privilege.
What people should be asking for is a level playing field. Making Microsoft open their source code is one way to do it, but it may take away their competitive advantage in other areas (for example, if they build a faster matrix solver) and arguably isn't fair to them. A reasonable alternative, however, would be to ask them to make all of their communication protocols transparent. Once you can interact with other Microsoft users as well as you could if you used Microsoft software yourself, the game is over.
Without such a concession from Microsoft (forced or otherwise), Microsoft's monopoly power will remain unchecked: the Internet will consist of Microsoft clients connecting to Microsoft servers running undocumented application protocols over TCP/UDP. (And you thought forcing HTML email on everyone was bad enough.) If you're smart, you'll either lobby Microsoft (or better, your gov't rep) to make them open their protocols, or you'll buy Microsoft stock.
This is only true if your filesystem was created with a 4Kb block size -- otherwise, you won't be able to do this.
What the Gartner Group doesn't understand about TCO (or at least fails to recognize in this report) is that the price of the OS license is only a tiny fraction of TCO. Here's what I think really matters in the calculation of TCO:
Primary documentation. Is the operating system well-documented? Are the manuals (or man pages) accurate, well-written, clear and concise? Linux doesn't score particularly well in the man page department, unfortunately--I think it loses points here. However...
Secondary documentation. How good is the secondary, user-contributed documentation out there relative to other operating systems? Are there well-written, current HOWTOs, user guides, tips and tricks, etc out there? If so, how well are they organized? Linux scores very high here. Other UNIXes are not so lucky. Windows has a lot of users out there--there are a lot of helpful tips on the Net but the documentation is not organized well.
Source code availability. It's an old adage that "the source code is the best documentation," and it's hard to argue otherwise. If I really want to know how a certain function call works, or how the kernel is talking to the hardware, I can dig out the glibc or kernel sources and see for myself, and be 100% certain of the accuracy of my conclusions (well, as long as my understanding of the code is correct). Once again, Linux and other free OSes score high here. With commercial UNIXes and Windows, you have to trust the documentation -- which is often not 100% accurate. In addition, if there's just a bug that needs to be fixed, it's often much easier to fix it yourself if you have access to the source code. Waiting for Microsoft or a commercial UNIX vendor to get a patch out to you can be painfully costly if your product or service's success is dependent on your software vendor's turnaround capability.
In summary, what I think really counts the most towards TCO is the relative understanding of the managers who plan system rollouts and the administrators who maintain them once they're out there. I've seen way too many Windows NT admins out there banging their heads on their desks because something doesn't work the way the documentation says it will, and they have no hope of getting it right because Tech Support doesn't know either. This leads to a lot of costly experimentation (labour is much more expensive than hardware) to make things work.
On the other hand, while Linux documentation isn't always perfect, it's plentiful, it's reasonably organized, and if you just can't find what you need from the HOWTOs or the man pages, you can always fall back on the source code.
rm -rf /* is a poor deletion technique anyways. There are several undocumented 'unrm' utilities out there (I've seen one myself) that are basically fancy implementations of 'dd' that can easily recover data from a UFS or ext2 filesystem.
If you really want to wipe your data clean you should use sdelete (Windows) or secure delete (Unix).
1.5. Did the RIAA really sue Pan?
- ask why a new version of a package was released?
- see a list of changes between old and new versions?
You can:rpm -qp --changelog rpm-file
There already is a very cool site that does this called MongoMusic. The "Sounds Like" feature is eerily accurate.
Interestingly, CNet recently reported that Microsoft is acquiring them...
I'm not sure I understand why people are calling it a publicity stunt. As far as I can tell, items placed for auction on eBay are legally binding; pud has to sell his site to the highest bidder. What makes anyone think otherwise--that he's not in fact selling it?
Anyways, good for him! At least someone is making money off his Web endeavours.
Neither a freemail account nor a free web hosting account is truly anonymous. They do retain HTTP access logs, and your ISP will be required to turn over their dialup logs pointing the activity to you personally if they are subpoenaed to do so. Remember Yahoo being required to positively identify the poster of the Apple leaks?
If you want true privacy, sign up at Zero Knowledge Systems and/or use a Mixmaster remailer.
Disclaimer: I am not a lawyer. See a competent attorney in your jurisdiction if you need legal advice.
Beware, folks. This opinion is from someone who apparently is employed by a company under the Universal umbrella. Do a 'whois' on naughtydog.com to see for yourself (they made the Crash Bandicoot series of games for Sony PlayStation).
Besides, whether or not MP3.com really broke the law still seems like a gray area to me. It seems odd to me that the judge didn't even bother to examine the scheme closely and ponder the unintended legal consequences that would result from such a hasty decision like that.
I hope MP3.com appeals. I really want to see a better clarification of what the law allows and what it doesn't, because the District Court is a rather low place on the totem pole.