Heh. Very good question to which we do not yet know the answer. MS hasn't yet announced the licensing terms, now has it?
The best I can figure is that 1) MS will want to make this useful right out-of-box, but 2) they want to make sure that those who find it really useful (like 50 instances useful) contribute a little bit more.
The closest thing to this situation in existing MS licensing would probably be Terminal Server. XP Pro ships with the ability to handle 1 session at a time. The Betas of.Net Server handle 2 sessions without additional licenses required (IIRC). My personal guess is that a similar situation would exist with MS's VPC software -- if it is included with the client OS, probably 1 instance would be allowed. If it is included with the server OS, 1 or 2 instances would be allowed without additional licensing. I also suspect that the licensing would not be quite as expensive as purchasing the equivalent number of copies of Windows.
However, this is only my guess. My guess and $1.00 will get you a 20 minute long distance call...
Well, that kind of header caching isn't the idea behind precompiled headers. First, note that (absent the concept of precompiled headers) nobody directly compiles headers:
gcc niftystuff.h -o niftystuff.o
I suppose it might work, but it is meaningless. You can't access the declarations in niftystuff.h by linking with niftystuff.o, since the declarations are a compile-time feature.
The best you can do is 1) minimize the headers included in each.C file to the minimum needed (possibly even organizing your program according to the headers needed) and 2) minimize the need to recompile.C files.
As far as I can tell, you are talking about part 2. If properly set up, make can compile a.C file only when it or an included.h file has changed. But when you have to recompile the.C file, you have to compile the included.h files, too.
The point of precompiled headers (.pch) is to save the time spent parsing the declarations of headers. That way, even when the.C file changes, you don't have to reparse the.h files it includes.
This is primarily useful for headers that come from outside of the current project, e.g. OS headers, CRT headers, X11 toolkit headers, libraries, etc. You probably never change these, so you precompile these headers once when they are installed (or upgraded). Now you never have to wait for stdlib.h to parse.
I'm a bit confused -- the excuse for not caching headers is because that is what make would do? Make doesn't cache precompiled headers, which is what we want to do. So what if the precompiled headers are recompiled using make's dependency rules. That is ideal, if you ask me.
No. "Fair use" is not a right. "Fair use" is a limitation on the copyright law. You are not guilty of copyright infringement if your use fell under the "Fair use" provision. Unfortunately, that does not mean either of the following:
* That your "fair use" will not be illegal under some other law (i.e. DMCA).
* That the copyright holder has to allow "fair use" if they can prevent it by other means.
This (the com.com article you linked) in fact is the article I did read, and while you claim that I am completely "ignorant" and/or lying through my teeth, the article clearly states that the NSA had some trouble with the GPL license, did not understand all of the consequences of the license, and "won't be doing anything like this again" (i.e. working on something GPLed). And MS wasn't the only one to complain.
Good point. I meant that Microsoft wasn't complaining that GPL code in general isn't available for them to steal. I meant that Microsoft is concerned that code that should be available to all (government sponsored research) wouldn't be available for them to use and examine. Which is definitely a legitimate concern. If you think otherwise, you need to study the real meaning of GPL and also study previous lawsuits regarding license infringement of source code. A BSD license allows a coder who will potentially use the results of research to examine the actual sample code when trying to figure out the specs (specs are never complete, in case you didn't know). A GPL license forces a clean-room setup where the specs are provided, and then whenever the programmer has a question, he/she can ask (in writing, recorded for all lawyers who ever might get involved) a "tainted" person to find the answer, and the tainted person can then examine the GPL code, answer the question (and never go beyond directly answering the question) and record the answer. What a major pain! It increases development cost by an order of magnitude or more. Sounds like MS has a real beef.
When people make comments that "Microsoft wants to steal all that GPLed code," the inferred meaning is that Microsoft wishes it could have the code to Linux, GnuCash, and XBill, which is bogus. However, it is somewhat true that Microsoft does want/need access to certain code, and GPLed code is pretty much out of reach, since it is really incompatible with any business model that is centered on selling software (and by the way, Red Hat doesn't count, since selling software is most definitely not the center of their business mode). Since I am still not convinced that selling software is morally wrong, I think Microsoft has a point.
Clean room sucks rocks. Have you ever been involved? Anybody who thinks it is fair to release something GPL because "they can always just do a clean-room implementation of it" has obviously never been involved in a clean-room operation. A real pain in the neck.
(I'm sure that anybody that remembers my name in the posts think that I'm avidly pro-Microsoft. I'm not. I've been ticked off by Microsoft every once in a while since 1985, longer than most people on SlashDot have been using computers. I'm just avidly anti-illogical-argument and avidly anti-hypocrisy. And since the most illogical arguments seem to be Microsoft-bashing, that is where I speak up most often.)
Right, but once there are a lot of contributors (such as if you didn't start from scratch on your project), YOU can't relicense it either. So when the government does research and bases it on GPL stuff, the result it pretty useless to industry.
Huh? So now you accuse me of lying. To tell you the truth, I don't remember exactly the reasons why the NSA was upset about the GPL, but they WERE. They had some plans for their secure distro, and they realized that those plans were going to be incompatible with the GPL. I'm not making the story up. So you tell me what they were upset about, and if it doesn't apply here, you have won the argument. But I think it does (from what I do remember), and thus my argument stands.
Right off, I can think of one situation where they might have been upset by the GPL. If they planned on making a secure distro available to the world, and needed to put some proprietary or sensitive code into it to make it meet their needs, they would be blocked by the "aggregate product" clauses of the GPL.
I'm mind stuck on the idea that computer "users" in most companies are EMPLOYEES that are contributing something useful. If you screw with their productivity, you cost the company something big time.
Yeah, but what I'm saying is that the journals can never have enough detail to be worth anything when it comes to computer science stuff (small advances, nifty algorithms - yes, but important new technologies such as TCP/IP - no). So Academia(TM) releases some sample code. If Academia GPLs the code, then companies are going to have to clean-room their implementation to avoid the possibility of lawsuits. If Academia puts a BSD license on it, then the companies can look at the code freely to understand the algorithms involved, even if they choose not to use the code directly.
(By the way, everybody keeps talking about BSD code in NT. First, that is unsubstantiated. Second, if true, it is in no way wrong. Third, according to my contacts, my understanding is that "if that was ever true, it is certainly not true in currently shipping code.")
Is it all working properly? Then you don't need new software!
Do you need software that won't run on your existing system? Then it isn't working properly!
OS X hasn't been around long enough to be sure, but they have already broken some compatibility with early software. And look at OS 9 -- OS X transition for additional history (not pretty!).
FreeBSD is great (I use it at home), but new software DOES often require OS updates.
Re:United Nations -- Iraq -- Weapons Inspections
on
Linux Is Cheaper
·
· Score: 1
(First, it is pore over books, not pour -- one involves intense scrutiny, the other involves wet books and angry librarians.)
Books = Not enough information to properly train someone in charge of an important system.
Books + experimentation + fiddling around (on a test system, of couse) + some time actually using the product = a decent course of study, whether it be for MS or Linux or Solaris.
Re:That's because Linux admins are self-taught
on
Linux Is Cheaper
·
· Score: 1
Not everybody is like that on either end. There are Linux morons and MS morons (and Sun morons), Linux gurus and MS gurus (and Sun), and Linux experimenters and MS experimenters and Sun experimenters. It is the person, not the platform, that makes the difference.
So what, users are just expenses, and nobody gets fired for messing with their productivity? FALSE. Small adjustments that don't really affect productivity may be one thing, but if a change reduces productivity, it costs money. Changing a mail/calendar/etc. program will never be transparent.
You can't just take research results out of a paper or journal and use them. There are always IP issues to worry about. So what you are saying is that if a university releases something under GPL, but I don't want to be bound by the GPL restrictions, I can just magically lift the same information from some other source? Nope. First, there isn't always another source. Second, if there is, it probably is missing a lot of important information since that information wasn't worth discussing in the journal (aimed at those who need just the facts, ma'am) but is necessary for a proper implemenation.
Umm, I didn't. I recognized that this was not in line with the exact nature of the GPL. Guess why I clarified things right before the statement, and went on to make a more realistic example.
Three cheers to taking things out of context! Good job, you get a cookie.
Photo 21: You can't get much text onto a slide and keep it readable. The context can either be supplied by the person doing the presentation, or by informed guessing. You guessed, but it seems that you left the "informed" part out of it. Lets see where context needs to be supplied.
"includes _________________ the benefits of open source."
You seem to have placed "all of" in the blank, then ridiculed the result. In reality, you should be red in the face since it is your own addition that you are making fun of. More sensible additions that come to mind (although I have no way to know if they are correct, just they seem more sensible to me) include "some of" and ", as a compromise without all of the commercial disadvantages of open source, some of".
Photo 22: Again, you are making fun of your own [lousy] interpretation of the caption. You jumped to a conclusion for Microsoft, effectively putting extra (stupid) words in MS's mouth, then once again, laughed at their stupidity. Perhaps you forget that MS does a LOT of work with open standards. But they most certainly do produce a lot of confusion, as can be seen by the way the "open" HTML standard has evolved over the years.
No, IP rights are retained by the copyright holder, and certain (but not all) IP rights irrevocably sublicensed to users of the GPL software.
Somehow, I find this argument to be very lawyer-istic.
Retained, yet irrevocably sublicensed. Cute.
Irrevocably sublicensed is pretty much "commons," at least as far as the target audience for this slide goes.
Bad for the industry...
You need to go back to English class and learn about the ambiguities of the English language, and then learn about how to properly fill them in.
There are two alternate interpretations here. One, which as you have explained, makes no sense whatsoever. Which means it is obviously not the correct one (if the second interpretation makes any sense). The second interpretation is that it is bad for the industry to choose. That is the point of the whole slide show. It is not normally a problem for an individual to choose GPL (unless he/she wants to try to make a lot of money from the software). It is often a problem when a company chooses GPL, since the results sometimes come back to haunt them later on.
Anyway, reading the rest of your complaints, they mostly come down to a difference of ideology between you and the author of the slide (and who is to say that there doesn't exist the same difference in ideology between you and the target audience for the slide? in which case, there is nothing wrong with what MS is saying) and your picking at the tiniest possible misalignment between what the slide says and what your perception of the truth is. Remember -- this is a slide show, where the objective is "get a simple reminder or bullet point onto the slide, then explain verbally in more depth." You assume that what is on the slide is exactly what was meant or all that was said. Very bad assumption.
Slashdot Mirror
Hint: This message redirects to somewhere else, to a website that I, for one, do not want to visit.
Heh. Very good question to which we do not yet know the answer. MS hasn't yet announced the licensing terms, now has it?
.Net Server handle 2 sessions without additional licenses required (IIRC). My personal guess is that a similar situation would exist with MS's VPC software -- if it is included with the client OS, probably 1 instance would be allowed. If it is included with the server OS, 1 or 2 instances would be allowed without additional licensing. I also suspect that the licensing would not be quite as expensive as purchasing the equivalent number of copies of Windows.
The best I can figure is that 1) MS will want to make this useful right out-of-box, but 2) they want to make sure that those who find it really useful (like 50 instances useful) contribute a little bit more.
The closest thing to this situation in existing MS licensing would probably be Terminal Server. XP Pro ships with the ability to handle 1 session at a time. The Betas of
However, this is only my guess. My guess and $1.00 will get you a 20 minute long distance call...
Don't worry -- I checked it for you. I can guarantee that there was some spelling in your posting.
Note that many blacklists work at the domain level. You're still toast.
Well, that kind of header caching isn't the idea behind precompiled headers. First, note that (absent the concept of precompiled headers) nobody directly compiles headers:
.C file to the minimum needed (possibly even organizing your program according to the headers needed) and 2) minimize the need to recompile .C files.
.C file only when it or an included .h file has changed. But when you have to recompile the .C file, you have to compile the included .h files, too.
.C file changes, you don't have to reparse the .h files it includes.
gcc niftystuff.h -o niftystuff.o
I suppose it might work, but it is meaningless. You can't access the declarations in niftystuff.h by linking with niftystuff.o, since the declarations are a compile-time feature.
The best you can do is 1) minimize the headers included in each
As far as I can tell, you are talking about part 2. If properly set up, make can compile a
The point of precompiled headers (.pch) is to save the time spent parsing the declarations of headers. That way, even when the
This is primarily useful for headers that come from outside of the current project, e.g. OS headers, CRT headers, X11 toolkit headers, libraries, etc. You probably never change these, so you precompile these headers once when they are installed (or upgraded). Now you never have to wait for stdlib.h to parse.
I'm a bit confused -- the excuse for not caching headers is because that is what make would do? Make doesn't cache precompiled headers, which is what we want to do. So what if the precompiled headers are recompiled using make's dependency rules. That is ideal, if you ask me.
No. "Fair use" is not a right. "Fair use" is a limitation on the copyright law. You are not guilty of copyright infringement if your use fell under the "Fair use" provision. Unfortunately, that does not mean either of the following:
* That your "fair use" will not be illegal under some other law (i.e. DMCA).
* That the copyright holder has to allow "fair use" if they can prevent it by other means.
There can be no such program. I have discovered a truly remarkable proof which this margin is too small to contain.
Ok, now I'm a bit ticked.
This (the com.com article you linked) in fact is the article I did read, and while you claim that I am completely "ignorant" and/or lying through my teeth, the article clearly states that the NSA had some trouble with the GPL license, did not understand all of the consequences of the license, and "won't be doing anything like this again" (i.e. working on something GPLed). And MS wasn't the only one to complain.
Good point. I meant that Microsoft wasn't complaining that GPL code in general isn't available for them to steal. I meant that Microsoft is concerned that code that should be available to all (government sponsored research) wouldn't be available for them to use and examine. Which is definitely a legitimate concern. If you think otherwise, you need to study the real meaning of GPL and also study previous lawsuits regarding license infringement of source code. A BSD license allows a coder who will potentially use the results of research to examine the actual sample code when trying to figure out the specs (specs are never complete, in case you didn't know). A GPL license forces a clean-room setup where the specs are provided, and then whenever the programmer has a question, he/she can ask (in writing, recorded for all lawyers who ever might get involved) a "tainted" person to find the answer, and the tainted person can then examine the GPL code, answer the question (and never go beyond directly answering the question) and record the answer. What a major pain! It increases development cost by an order of magnitude or more. Sounds like MS has a real beef.
When people make comments that "Microsoft wants to steal all that GPLed code," the inferred meaning is that Microsoft wishes it could have the code to Linux, GnuCash, and XBill, which is bogus. However, it is somewhat true that Microsoft does want/need access to certain code, and GPLed code is pretty much out of reach, since it is really incompatible with any business model that is centered on selling software (and by the way, Red Hat doesn't count, since selling software is most definitely not the center of their business mode). Since I am still not convinced that selling software is morally wrong, I think Microsoft has a point.
Clean room sucks rocks. Have you ever been involved? Anybody who thinks it is fair to release something GPL because "they can always just do a clean-room implementation of it" has obviously never been involved in a clean-room operation. A real pain in the neck.
Yeah, too bad.
(I'm sure that anybody that remembers my name in the posts think that I'm avidly pro-Microsoft. I'm not. I've been ticked off by Microsoft every once in a while since 1985, longer than most people on SlashDot have been using computers. I'm just avidly anti-illogical-argument and avidly anti-hypocrisy. And since the most illogical arguments seem to be Microsoft-bashing, that is where I speak up most often.)
Right, but once there are a lot of contributors (such as if you didn't start from scratch on your project), YOU can't relicense it either. So when the government does research and bases it on GPL stuff, the result it pretty useless to industry.
Huh? So now you accuse me of lying. To tell you the truth, I don't remember exactly the reasons why the NSA was upset about the GPL, but they WERE. They had some plans for their secure distro, and they realized that those plans were going to be incompatible with the GPL. I'm not making the story up. So you tell me what they were upset about, and if it doesn't apply here, you have won the argument. But I think it does (from what I do remember), and thus my argument stands.
Right off, I can think of one situation where they might have been upset by the GPL. If they planned on making a secure distro available to the world, and needed to put some proprietary or sensitive code into it to make it meet their needs, they would be blocked by the "aggregate product" clauses of the GPL.
I'm mind stuck on the idea that computer "users" in most companies are EMPLOYEES that are contributing something useful. If you screw with their productivity, you cost the company something big time.
Yeah, but what I'm saying is that the journals can never have enough detail to be worth anything when it comes to computer science stuff (small advances, nifty algorithms - yes, but important new technologies such as TCP/IP - no). So Academia(TM) releases some sample code. If Academia GPLs the code, then companies are going to have to clean-room their implementation to avoid the possibility of lawsuits. If Academia puts a BSD license on it, then the companies can look at the code freely to understand the algorithms involved, even if they choose not to use the code directly.
(By the way, everybody keeps talking about BSD code in NT. First, that is unsubstantiated. Second, if true, it is in no way wrong. Third, according to my contacts, my understanding is that "if that was ever true, it is certainly not true in currently shipping code.")
Make up your mind:
Is it all working properly? Then you don't need new software!
Do you need software that won't run on your existing system? Then it isn't working properly!
OS X hasn't been around long enough to be sure, but they have already broken some compatibility with early software. And look at OS 9 -- OS X transition for additional history (not pretty!).
FreeBSD is great (I use it at home), but new software DOES often require OS updates.
(First, it is pore over books, not pour -- one involves intense scrutiny, the other involves wet books and angry librarians.)
Books = Not enough information to properly train someone in charge of an important system.
Books + experimentation + fiddling around (on a test system, of couse) + some time actually using the product = a decent course of study, whether it be for MS or Linux or Solaris.
Not everybody is like that on either end. There are Linux morons and MS morons (and Sun morons), Linux gurus and MS gurus (and Sun), and Linux experimenters and MS experimenters and Sun experimenters. It is the person, not the platform, that makes the difference.
So what, users are just expenses, and nobody gets fired for messing with their productivity? FALSE. Small adjustments that don't really affect productivity may be one thing, but if a change reduces productivity, it costs money. Changing a mail/calendar/etc. program will never be transparent.
You can't just take research results out of a paper or journal and use them. There are always IP issues to worry about. So what you are saying is that if a university releases something under GPL, but I don't want to be bound by the GPL restrictions, I can just magically lift the same information from some other source? Nope. First, there isn't always another source. Second, if there is, it probably is missing a lot of important information since that information wasn't worth discussing in the journal (aimed at those who need just the facts, ma'am) but is necessary for a proper implemenation.
Umm, I didn't. I recognized that this was not in line with the exact nature of the GPL. Guess why I clarified things right before the statement, and went on to make a more realistic example.
Three cheers to taking things out of context! Good job, you get a cookie.
Photo 21:
You can't get much text onto a slide and keep it readable. The context can either be supplied by the person doing the presentation, or by informed guessing. You guessed, but it seems that you left the "informed" part out of it. Lets see where context needs to be supplied.
"includes _________________ the benefits of open source."
You seem to have placed "all of" in the blank, then ridiculed the result. In reality, you should be red in the face since it is your own addition that you are making fun of. More sensible additions that come to mind (although I have no way to know if they are correct, just they seem more sensible to me) include "some of" and ", as a compromise without all of the commercial disadvantages of open source, some of".
Photo 22:
Again, you are making fun of your own [lousy] interpretation of the caption. You jumped to a conclusion for Microsoft, effectively putting extra (stupid) words in MS's mouth, then once again, laughed at their stupidity. Perhaps you forget that MS does a LOT of work with open standards. But they most certainly do produce a lot of confusion, as can be seen by the way the "open" HTML standard has evolved over the years.
No, IP rights are retained by the copyright holder, and certain (but not all) IP rights irrevocably sublicensed to users of the GPL software.
Somehow, I find this argument to be very lawyer-istic.
Retained, yet irrevocably sublicensed. Cute.
Irrevocably sublicensed is pretty much "commons," at least as far as the target audience for this slide goes.
Bad for the industry...
You need to go back to English class and learn about the ambiguities of the English language, and then learn about how to properly fill them in.
There are two alternate interpretations here. One, which as you have explained, makes no sense whatsoever. Which means it is obviously not the correct one (if the second interpretation makes any sense). The second interpretation is that it is bad for the industry to choose. That is the point of the whole slide show. It is not normally a problem for an individual to choose GPL (unless he/she wants to try to make a lot of money from the software). It is often a problem when a company chooses GPL, since the results sometimes come back to haunt them later on.
Anyway, reading the rest of your complaints, they mostly come down to a difference of ideology between you and the author of the slide (and who is to say that there doesn't exist the same difference in ideology between you and the target audience for the slide? in which case, there is nothing wrong with what MS is saying) and your picking at the tiniest possible misalignment between what the slide says and what your perception of the truth is. Remember -- this is a slide show, where the objective is "get a simple reminder or bullet point onto the slide, then explain verbally in more depth." You assume that what is on the slide is exactly what was meant or all that was said. Very bad assumption.
Huh? No way.
Microsoft just wants to be sure that Academia doesn't start releasing through GPL, cuz that would screw over any potential commercial use.