I must use ClearCase in my job every day, and I hate it more than any source code control system I have used before (CVS, VSS, StarTeam) - albeit, not always because of ClearCase itself. ClearCase is VERY VERY powerful. This leads to two major disadvantages:
1) With power comes the opportunity for abuse; hence it is easy to really screw things up.
2) With power comes complexity; it is not only difficult to administer, but also difficult to use.
For example, because directories are treated as elements to be managed just like files, you must check out and check in directories before you can change any property of a file.
There is a rule engine [config spec evaluation] for evaluating whether you are going to access a particular set of branches or labels. The branching and merging mechanisms have a steep learning curve, and expose a great deal of complexity that would not lend itself very easily to open source development.
Lastly, dynamic views - a mechanism whereby you mount a filesystem that where "updates" occur instantanteously - have two big problems:
* it's great that you're always up-to-date, but it also means that you can't be sure that interfaces to code you depend on can't change from under you
* performance: dynamic views are terribly slow when there are large numbers of people, it tends to saturate your network, and remote users/remote projects are unbearably slow.
ClearCase, while it has some unique and powerful options, is NOT a good model for Open Source projects.
It seems to me that the most obvious use of this technology is to build it into all CDRW drives and require that the drive "Tattoo" it's serial number on to any discs that it burns. I relealize that even if they went to the length of requiring you to "personalize" your CDRW drive, you could still fake it out, it does seeem exactly like the kind of thing content companies might try to mandate (and it wouldn't have any implications viz Red Book compliance)
Regarding your last, OS X dev toosl [/Developer/Tools ] include things like CpMac, MvMac which are command line versions of cp and mv that do handle resource forks. There are also tools for splitting/joining the resource fork.
Microsoft appears to believe that the GNU Public License (GPL) presents some kind of danger to "intellectual property" rights, and that in particular Microsoft is trying to warn other companies [and not just software companies] of the danger.
With the exception of software companies, what danger does the use of the GPL really pose to most businesses, given that for them software is means to end and not the thing which defines them as a company? And how does Microsoft shared source offer GPL like benefits without the "IP Liability"?
I would like to offer a simple but specific example to clarify the context of the question: a small widget manufacturer has a production line, which he wishes to automate to stamp the logo of his customer on each widget. He has a database of customers, and a program developed by a large software company which controls the stamping press.
He wants to connect the database and the stamping press program to change the stamp pattern according to his purchase orders in the database. This is a very simple adjustment. He cannot however modify the program, because the company which developed it has gone out of business, and he has no access to the source code. GPL software exists which he could use, but at the price of [potentially] sharing the that change with his competitors, but Microsoft also has a stamp press controller which he could buy, and get a Shared Source license for, enabling him to make the change. The question is, why should he choose MS and Shared Source over the GPL code? Is this ability really the core competency of his business, or merely a tool to help him in his real compentency which is making the best widgets? Presumably, if all that distinguished this manufacturer was his ability to stamp logos, how would the MS shared source license accomplish his goal of obscurity? Would MS not be privy to incorporate changes he made, and sell it at later date to his competitors or to make it a feature of MS StampPress? Would the low cost and low overhead of using GPL outweigh his concerns about IP competition? Furthermore, even supposing his competitors did end up using the "custom" code, does the fact that he can now benefit from any improvements THEY make change his decision - and can he benefit from changes other's make if he goes with MS Shared Source?
Uhh... Do these numbers make sense?
on
Benchmark Madness
·
· Score: 2
The numbers listed for XFS make it seem terribly slow in almost every category compared with either ReiserFS or Ext2, yet the writeup seems to conclude that XFS and Reiser offer comparable performance. What gives? Did something get lost in translation?
Gracenote fails to mention that when referring to their "intellectual property", they are mostly referring to the patent they have on the generating the unique key's for identifying a particular CD by segment count and length.
Which leads me to ask the question: if this is their whole IP claim, couldn't FreeDB simply use a different algorithm (say MD5) to generate a unique CD key ???
Mundie is using tricky doublespeak
on
Mundie Responds
·
· Score: 1
We believe that one of these mechanisms is intellectual property rights. Without intellectual property protection,
neither innovation nor a healthy commercial software industry is sustainable....
From this, the reader is to infer that the GPL threatens intellectual property rights. That is patently (forgive the pun) untrue. There is a semantic ambiguity about what one's right's to intellectual property are, which might include:
1. The right to charge others for the use of your IP. (note: not what MicroSoft generally does)
2. The right to give it away freely. id est , public domain.
3. The right to make up an abitrary set of restrictions about the conditions with which someone can use your IP.
Option 3 is what Microsoft does (think about restrictions on how many copies, which machine it's installed on, etc. [in a word, licensing]. Option 3 is what the GPL does.
The difference is that Microsoft heavily stipulates what can be done to / with their IP and charges money based on those stipulations. The GPL stipulates that anything can be done with GPL protected IP as long as you share it with the rest of the world.
This is not a loss of IP Right Protections, it is a choice about HOW you allow others to use your IP, not whether you retain the rights to it.
Mundie is using tricky doublespeak
on
Mundie Responds
·
· Score: 1
We believe that one of these mechanisms is intellectual property rights. Without intellectual property protection,
neither innovation nor a healthy commercial software industry is sustainable....
From this, the reader is to infer that the GPL threatens intellectual property rights. That is patently (forgive the pun) untrue. There is a semantic ambiguity about what one's right's to intellectual property are, which might include:
1. The right to charge others for the use of your IP. (note: not what MicroSoft generally does)
2. The right to give it away freely. id est , public domain.
3. The right to make up an abitrary set of restrictions about the conditions with which someone can use your IP.
Option 3 is what Microsoft does (think about restrictions on how many copies, which machine it's installed on, etc. [in a word, licensing]. Option 3 is what the GPL does.
The difference is that Microsoft heavily stipulates what can be done to / with their IP and charges money based on those stipulations. The GPL stipulates that anything can be done with GPL protected IP as long as you share it with the rest of the world.
This is not a loss of IP Right Protections, it is a choice about HOW you allow others to use your IP, not whether you retain the rights to it.
There is no technology solution that will "protect" children from the world into which they were born.
Firstly, machines lack intelligence: they cannot distinguish between wholesome material and "vile" material at all; any so-called experts who make that claim are telling people what they want to hear - that there is a simple technology that will save us from having to be responsible parents. Moreover, these technologies always block access to completely unoffensive material due to the vagaries of the censor (websites about Essex, England, for example).
Secondly, there is no such thing as an objective standard of what material is offensive. Where censorware is employed, some unknown person's judgement is made about what constitutes offense or, even more likely, some lest common denominator of decency will define what children can see and hear. It is one thing to offer ratings of sites, and let parents or teachers be forewarned; it is another thing entirely to restrict access to materials that perhaps most citizens would find acceptable for their children, and certainly it is not permissable that that totally unoffensive material be abolished for a dubious "protection" against offensive material.
I do agree though that an interesting possible alternative would be a website rating system something like the moderation system of slashdot, which is effectively a distributed rating system. Though it would be important for the default rating to be reasonable; i.e. by default a site without ratings should be considered OK. It's more important to promote the good than to filter out the few bad things that not moderated down.
I think you miss the point of Microsoft's strategy: the next version of office will be based upon.NET, but consumers won't be forced to resubscribe at first. It is only after users have come to rely on storing all their documents in MS Word 9.x format that they will activate subscription features.
Microsoft products are backwards but never forwards compatible, and the version/format of documents is one of the ways that Microsoft ensure s that whatever new things they decide to add to their products, be it a component technology or a web browser, you are forced to accept them.
At a party not too long ago, a bunch of geeks (myself included) were dicussing the best way to try to ensure that important knowledge of our civilization survives us, particularly in the case that humans may be wiped out, but the rest of the planet exists for alien exobiologists to study.
Through the course of discussion, the best approach we could think of was to encode important knowledge (in a redundant fashion, a la Hamming coding, etc) into the "junk" DNA of several species likely to survive and slow to mutate. I believe redwoods, cockroaches, and sharks were all suggested as candidates.
So the real trick is to work out a good system for genetically engineering species to carry our important information for us.
This could perhaps be applied to the problem of loss of computers, except that we currently require computers to read DNA! So the other thing we need is a way to sequence DNA without computers.
The Macintosh has arguably even better support for top-down or right-to-left foreign alphabets than Windows. Also, you can run the MS applications (Word and whatnot, including IE) on it.
And if you would like to have the power of UNIX, OS X will enable you to run things the proper *nix way. True, you'll have to wait maybe 6 months until Hebrew support is finished, and of course that you'll have to buy new hardware, but by that time you may find that you're in the market for a new machine anyway.
In a relatively short time, there will be a BSD based OS with applications relevant to the documents you are trying to process, and support for Hebrew.
The advantages of Mac OS X are not likely to be of significance to PC based Linux users. So assuming that we are talking about people who run LinuxPPC or YellowDog on their Macs, there are some good reasons to switch.
1) Mac OS X will be able to run almost any software available for Linux. [you get to run lots of great free software- gcc, apache, you name it.]
2) Fundamentally, it has most of the power of linux with a GUI that is going to be supported by a fairly large chunk of commercial developers. [you get to run lots of great commercial software]
Think of how many times you run across QuickTime in Linux and you're hosed. No problem with OSX.
3) Performance wise, it's going to be a good UN*X variant, and everyone familiar and happy with the power of Un*X will be able to work in it just fine.
4) X windows sucks. It is no great feat for Apple to come up with an alternative that people prefer to use (or even to code to - though GTK mitigates the problem).
It may seem bizarre given how often this argument has been used against the Mac, but the essence of the argument to use OS X over Linux is that you can run a lot more software on OS X. While that can certainly be said of Windows, it also means that you have to inherit the pain of the windows OS. OS X will for most purposes be un*x.
Darwin notwithstanding, the reason to stay with Linux is that it is FREE and truly open source. For this reason, I think that many people will stay with Linux, but I suspect that a lot of cross pollination may occur. What Apple intended to do with Darwin (IMHO) is offer an appeal to Linux users by only maintaining a proprietary stranglehold on the gui.
Slashdot Mirror
I must use ClearCase in my job every day, and I hate it more than any source code control system I have used before (CVS, VSS, StarTeam) - albeit, not always because of ClearCase itself.
ClearCase is VERY VERY powerful. This leads to two major disadvantages:
1) With power comes the opportunity for abuse; hence it is easy to really screw things up.
2) With power comes complexity; it is not only difficult to administer, but also difficult to use.
For example, because directories are treated as elements to be managed just like files, you must check out and check in directories before you can change any property of a file.
There is a rule engine [config spec evaluation] for evaluating whether you are going to access a particular set of branches or labels. The branching and merging mechanisms have a steep learning curve, and expose a great deal of complexity that would not lend itself very easily to open source development.
Lastly, dynamic views - a mechanism whereby you mount a filesystem that where "updates" occur instantanteously - have two big problems:
* it's great that you're always up-to-date, but it also means that you can't be sure that interfaces to code you depend on can't change from under you
* performance: dynamic views are terribly slow when there are large numbers of people, it tends to saturate your network, and remote users/remote projects are unbearably slow.
ClearCase, while it has some unique and powerful options, is NOT a good model for Open Source projects.
It seems to me that the most obvious use of this technology is to build it into all CDRW drives and require that the drive "Tattoo" it's serial number on to any discs that it burns. I relealize that even if they went to the length of requiring you to "personalize" your CDRW drive, you could still fake it out, it does seeem exactly like the kind of thing content companies might try to mandate (and it wouldn't have any implications viz Red Book compliance)
Regarding your last, OS X dev toosl [ /Developer/Tools ] include things like CpMac, MvMac which are command line versions of cp and mv that do handle resource forks. There are also tools for splitting/joining the resource fork.
Microsoft appears to believe that the GNU Public License (GPL) presents some kind of danger to "intellectual property" rights, and that in particular Microsoft is trying to warn other companies [and not just software companies] of the danger.
With the exception of software companies, what danger does the use of the GPL really pose to most businesses, given that for them software is means to end and not the thing which defines them as a company? And how does Microsoft shared source offer GPL like benefits without the "IP Liability"?
I would like to offer a simple but specific example to clarify the context of the question: a small widget manufacturer has a production line, which he wishes to automate to stamp the logo of his customer on each widget. He has a database of customers, and a program developed by a large software company which controls the stamping press. He wants to connect the database and the stamping press program to change the stamp pattern according to his purchase orders in the database. This is a very simple adjustment. He cannot however modify the program, because the company which developed it has gone out of business, and he has no access to the source code. GPL software exists which he could use, but at the price of [potentially] sharing the that change with his competitors, but Microsoft also has a stamp press controller which he could buy, and get a Shared Source license for, enabling him to make the change. The question is, why should he choose MS and Shared Source over the GPL code? Is this ability really the core competency of his business, or merely a tool to help him in his real compentency which is making the best widgets? Presumably, if all that distinguished this manufacturer was his ability to stamp logos, how would the MS shared source license accomplish his goal of obscurity? Would MS not be privy to incorporate changes he made, and sell it at later date to his competitors or to make it a feature of MS StampPress? Would the low cost and low overhead of using GPL outweigh his concerns about IP competition? Furthermore, even supposing his competitors did end up using the "custom" code, does the fact that he can now benefit from any improvements THEY make change his decision - and can he benefit from changes other's make if he goes with MS Shared Source?
The numbers listed for XFS make it seem terribly slow in almost every category compared with either ReiserFS or Ext2, yet the writeup seems to conclude that XFS and Reiser offer comparable performance. What gives? Did something get lost in translation?
Gracenote fails to mention that when referring to their "intellectual property", they are mostly referring to the patent they have on the generating the unique key's for identifying a particular CD by segment count and length.
Which leads me to ask the question: if this is their whole IP claim, couldn't FreeDB simply use a different algorithm (say MD5) to generate a unique CD key ???
That is patently (forgive the pun) untrue.
There is a semantic ambiguity about what one's right's to intellectual property are, which might include:
1. The right to charge others for the use of your IP. (note: not what MicroSoft generally does)
2. The right to give it away freely. id est , public domain.
3. The right to make up an abitrary set of restrictions about the conditions with which someone can use your IP.
Option 3 is what Microsoft does (think about restrictions on how many copies, which machine it's installed on, etc. [in a word, licensing].
Option 3 is what the GPL does.
The difference is that Microsoft heavily stipulates what can be done to / with their IP and charges money based on those stipulations.
The GPL stipulates that anything can be done with GPL protected IP as long as you share it with the rest of the world.
This is not a loss of IP Right Protections, it is a choice about HOW you allow others to use your IP, not whether you retain the rights to it.
That is patently (forgive the pun) untrue.
There is a semantic ambiguity about what one's right's to intellectual property are, which might include:
1. The right to charge others for the use of your IP. (note: not what MicroSoft generally does)
2. The right to give it away freely. id est , public domain.
3. The right to make up an abitrary set of restrictions about the conditions with which someone can use your IP.
Option 3 is what Microsoft does (think about restrictions on how many copies, which machine it's installed on, etc. [in a word, licensing].
Option 3 is what the GPL does.
The difference is that Microsoft heavily stipulates what can be done to / with their IP and charges money based on those stipulations.
The GPL stipulates that anything can be done with GPL protected IP as long as you share it with the rest of the world.
This is not a loss of IP Right Protections, it is a choice about HOW you allow others to use your IP, not whether you retain the rights to it.
There is no technology solution that will "protect" children from the world into which they were born.
Firstly, machines lack intelligence: they cannot distinguish between wholesome material and "vile" material at all; any so-called experts who make that claim are telling people what they want to hear - that there is a simple technology that will save us from having to be responsible parents. Moreover, these technologies always block access to completely unoffensive material due to the vagaries of the censor (websites about Essex, England, for example).
Secondly, there is no such thing as an objective standard of what material is offensive. Where censorware is employed, some unknown person's judgement is made about what constitutes offense or, even more likely, some lest common denominator of decency will define what children can see and hear. It is one thing to offer ratings of sites, and let parents or teachers be forewarned; it is another thing entirely to restrict access to materials that perhaps most citizens would find acceptable for their children, and certainly it is not permissable that that totally unoffensive material be abolished for a dubious "protection" against offensive material.
I do agree though that an interesting possible alternative would be a website rating system something like the moderation system of slashdot, which is effectively a distributed rating system. Though it would be important for the default rating to be reasonable; i.e. by default a site without ratings should be considered OK. It's more important to promote the good than to filter out the few bad things that not moderated down.
There is a great installation guide/diary complete with photographic documentation by Ben Rota at ArsTechnica.
http://arstechnica.com/guide/networking/installat
I think you miss the point of Microsoft's strategy: the next version of office will be based upon .NET, but consumers won't be forced to resubscribe at first. It is only after users have come to rely on storing all their documents in MS Word 9.x format that they will activate subscription features.
Microsoft products are backwards but never forwards compatible, and the version/format of documents is one of the ways that Microsoft ensure s that whatever new things they decide to add to their products, be it a component technology or a web browser, you are forced to accept them.
At a party not too long ago, a bunch of geeks (myself included) were dicussing the best way to try to ensure that important knowledge of our civilization survives us, particularly in the case that humans may be wiped out, but the rest of the planet exists for alien exobiologists to study.
Through the course of discussion, the best approach we could think of was to encode important knowledge (in a redundant fashion, a la Hamming coding, etc) into the "junk" DNA of several species likely to survive and slow to mutate.
I believe redwoods, cockroaches, and sharks were all suggested as candidates.
So the real trick is to work out a good system for genetically engineering species to carry our important information for us.
This could perhaps be applied to the problem of loss of computers, except that we currently require computers to read DNA! So the other thing we need is a way to sequence DNA without computers.
The Macintosh has arguably even better support for top-down or right-to-left foreign alphabets than Windows. Also, you can run the MS applications (Word and whatnot, including IE) on it.
And if you would like to have the power of UNIX, OS X will enable you to run things the proper *nix way. True, you'll have to wait maybe 6 months until Hebrew support is finished, and of course that you'll have to buy new hardware, but by that time you may find that you're in the market for a new machine anyway.
In a relatively short time, there will be a BSD based OS with applications relevant to the documents you are trying to process, and support for Hebrew.
The advantages of Mac OS X are not likely to be of significance to PC based Linux users. So assuming that we are talking about people who run LinuxPPC or YellowDog on their Macs, there are some good reasons to switch. 1) Mac OS X will be able to run almost any software available for Linux. [you get to run lots of great free software- gcc, apache, you name it.] 2) Fundamentally, it has most of the power of linux with a GUI that is going to be supported by a fairly large chunk of commercial developers. [you get to run lots of great commercial software] Think of how many times you run across QuickTime in Linux and you're hosed. No problem with OSX. 3) Performance wise, it's going to be a good UN*X variant, and everyone familiar and happy with the power of Un*X will be able to work in it just fine. 4) X windows sucks. It is no great feat for Apple to come up with an alternative that people prefer to use (or even to code to - though GTK mitigates the problem). It may seem bizarre given how often this argument has been used against the Mac, but the essence of the argument to use OS X over Linux is that you can run a lot more software on OS X. While that can certainly be said of Windows, it also means that you have to inherit the pain of the windows OS. OS X will for most purposes be un*x. Darwin notwithstanding, the reason to stay with Linux is that it is FREE and truly open source. For this reason, I think that many people will stay with Linux, but I suspect that a lot of cross pollination may occur. What Apple intended to do with Darwin (IMHO) is offer an appeal to Linux users by only maintaining a proprietary stranglehold on the gui.