>
some will say that the fact that SCO continues to provide Linux source code means that GPL is involved. But once again, that issue would be valid if the released source code were BSD
Actually not. The (reformed) BSD license and friends do allow anyone to do anything but stripe out the copyrights notice. So if Linux was BSD-licensed, SCO could have gotten away with its reasoning, and IBM wouldn't have been able to point SCO's distribution of Linux as hypocritical and infringing.
>
it would be possible to release a proprietary program
which takes a GPL'd source program, patches it and links it with
independent binaries to make a new proprietary program
How practical would that be? Probably not enough to make a dent at the original program community, as any patches if useful would probably be small enough to be quickly replicated, and would be a pain to follow up with new versions.
Now, suppose these patches contain the implementation of an essencial patent... well, this are just like binary modules for the kernel and codecs for mplayer; they exist, but they tend to become roadkill to further development, as most patents are interesting but liable to workarounds.
Good: Once for all this "not yet tested in court" FUD will go away, and future violators might be deterred.
Bad: Media circus. 'nough said.
Worst: A bad result might make thousands of talented people loose faith in justice. This is actually good, as would be any acts of civil disobedience in consequence. The actual evil has been done in the last several decades in the slow, generalised erosion of morals... final defeat would be ugly, even the civil disobedience struggle itself would be honourable but containing scenes of unthinkable ugliness, stupidity and cruelty.
>
it's always going to be an issue when applying a discrete concept such as that of species to what is in reality a continuum
You are probably right, but not only... there are lots of biologists who have evolution as a working hypothesis for changes, yet think evidence is not strong for it being the sole mechanism of the creation of life and personality.
>
I'm really referring to the gap between us and lesser primates, rather than the purely evolutionary chain
This is one issue, how many gaps there are. But remember to consider also the other issue: how wide the gaps are. Lots of gaps, not only between us and lesser primates but also among other primates and in several other points of the evolutionary 'family tree', are big enough that not only one or two fossiles or species are needed to fill up, but several.
It's always temerary to hold this as definitive, but from what I talked with a microbiologist, they (biologists) have next to no hope of filling any significant gap. Basically they think that, like petrol, most has already been discovered; there is potential for new discoveries, but nothing proportional to the quantity and size of the gaps.
>
or it's that missing link anthropologists have been searching for
This would be more like evolutionary biologists than anthropologists; the later are concerned with man as man, not as an animal.
Anyway, tall order. It is not a specie that will fill the gap. There would need to be a big number of fossiles and (or) living species discovered to fill the multiple gaps in evolutionary evidence, and not only near man but all over the classification of animals and vegetables.
>
it wasn't always the best solution for the users
At the time servers were incredibly expensive, and proprietary. Users, in typical self-defeating short-sightedness, preferred to buy their own PCs instead of funding IT to give them professional services and request really open systems. IT ended up supporting a incredibly more costly and complex structure.
>
others are suited for "fat" clients
For example? I can think now only of portables and some really, really high-end technical workstations. Perhaps I'm forgetting some, but it seems to me that nowadays people try to fit all and any situations to a client-server solution...
>
it's not always the holy grail of users getting stuff done
Not always, but almost. I did that, I know. Even with inefficient Terminal Server, users were wooed by the incredible performance they won, not to mention reliability and being able to work from anywhere.
>
There is a major limitation of host/terminal environments-- mobile laptops.
I believe portable computers should be supported, even with more sophistication than now -- stuff like Intermezzo and Coda come to mind. But this extra complexity should have a price: IT should charge the department of the users accordingly. The price differente from supporting a X terminal to a portable would be enough dissuasion for most people...
>
host/terminal has its place, but so does client/server
Yes, and I believe host and terminals should be the core of most organisations' networks, and client-server should be reserved for exceptional cases.
To tell you the truth, I believe each apartment building or houses block should have its small server room with two or three machines hosting all apps for a X terminal at each room...
>
Microsoft's advantages in areas such as security, feature-completeness and total cost of ownership
This is typical arguing from the conclusions. They already know what they want to find; now it is just a matter of crafting the studies that will find exactly that, not the reality.
But why be serious?
>
Microsoft's advantages in [...] security
Like being so incredibly complex no one can properly manage permissions, hiding information so no one can actually understand what's going on, and making it nearly impossible to log stuff? Not to mention hiding source code so it can't possibly be audited? Now, that's security for me...
>
Microsoft's advantages in [...] feature-completeness
Like there will be a Microsoft Debian distribution with everything and the kitchen sink installable from CDs or network servers, following a coherent policy? Wow... can't wait for that... will take too long!
>
Microsoft's advantages in [...] total cost of ownership
Like no more incredibly expensive MSCE that cost even more by always choosing the most inefficient solution? No more incredibly complex, expensive licensing that charges double for Terminal Server usage? Suddenly MS SysAdmins will be as efficient as their Unix counterparts? Toto, we're not in Kansas...
>
ActiveDirectory has many features that are currently lacking in open source implimentations, including multimaster replication
Perhaps I'm misunderstading you, but isn't that something NIS and NIS+ have been doing for ages?
>
And network administration in a server/client network is easier with Windows, I think.
If both servers and clients are uniformly GNU/Linux, it can be easier than MS Windows. Not to mention that easiness is not the ultimate measure: there are more important issues like security, performance, reliability, efficiency...
But most important, client/server sucks. The holy grail of systems administration is host and terminals, and there X Windows exceeds by enabling X terminals. Just wait until we replicate SunRay functionality...
Interestingly enough, the press release says OpenSource Now is for GNU GPL code. Is it excluding other licenses, say public domain, BSD, OSL or even the GNU LGPL and FDL? I can understand focusing on copyleft, but not exclusively GPL.
>
Gnome is the only one that isn't so much a cross-platform application
Granted it's not an application, it is a programming and user interface framework. It is portable, being runnable mostly everywhere there's POSIX; Mono can potentially make it even more portable, but that remains to be seen.
>
But does Novell have the cash to continue development of all these?
I just hope Novell doesn't let them die on the vine.
Money isn't easy to come by nowadays. I guess having a real company behind may be beneficial to Ximian's business plans, but in any case it is mostly GNU software the community will either take in case of a Ximian debacle, or the FSF will.
OTOH, I don't like big companies. Usually merges and buyouts diminish their ability to compete.
>
It wasn't some sort of high ranking conspiracy that kept Bibles away from the possession of the common people
This was a nice excuse until the invention of the press and the Reformation. After that, the Bible "had" to be put on the Index Librorum Prohibitorum, and stayed there for four long centuries. One needed the bishop's written authorisation to read it. Even classical texts were bowdlerised to use in the Jesuit schools. Protestant Bible colportors were expulsed from Catholic cities and villages until the 1.970's in Northeastern Brazil, in some instances with the help of the cangaceiros bandits; are still being expulsed from Mexican pueblos. The priests slandered the Protestants with mistranslations, and wouldn't allow their own translations to be issued with extensive notes contradicting the sense of damning passages with out-of-context quotes and controversial opinions by later Fathers of the Church.
>
a lawyer says that what he just said doesn't constitute legal advice, what message do you take from that?
That he is wise enough to not let his words be misused out of context as free advice by people who can't get their clues. Like you.
>
Make sure the list the ones that the EFF don't pursue
FSF, not EFF. And everytime someone discovers a GNU GPL violation, it gets fixed. The burden of proof is on you to discover an instance of unpursued violation.
>
We need a test case.
Why do we need? I am far more assured by violators' prudence in complying to avoid a probable defeat.
>
Why are they letting SCO pick the terms?
SCO is not picking the terms. In fact, in countries when this is possible GNU/Linux user groups have already ask the courts to shut them up -- Germany & Australia. But this has nothing to do with a GNU GPL test case, it is just preventing slandering. SCO hasn't yet provided any evidence, so there is no way of the FSF doing anything now. Even when they do, the Linux kernel has little code in it that has been assigned to the FSF, and in general it uses the GNU GPL with the famous binary modules exception. So the FSF has no stake in it. Now if IBM or the Linux developers decide they need FSF's help, I'm sure Eben and co. will be eager to help, but somehow I feel IBM lawyers need no help...
All things considered, this is not a copyright case, so it has no value to validate the GNU GPL except as a side effect of Caldera GNU/Linux distribution. Given that, what do you suggest the FSF do exactly? There is no case at hand that I can see.
>
CISC chips are a RISC chip with a built-in ROM and microcode decoder
Hmm... not. CISC and RISC define the interface. CISC presents a real problem because the decoding of CISC instructions into the internal architecture, usually RISC but also VLIW, takes chip real state and design efforts, thus delaying development, making the final product more expensive and less efficient, and blocking advancement of the whole architecture.
>
Intel and AMD
I am not interested in Intel and AMD. What do you offend me for?
The real history here is not the foundry, but the architecture. RISC is simpler, not only to run fast but also to design, manufacture and perhaps reverse engineer.
>
All the stuff I see says "MIPS Like" [...] http://www.most.gov.cn/english/newletter/q274.htm
Interestingly, the very first link says fully MIPS compatbile.
I praise them from breaking free from CISC and avoiding the VLIW pitfall. But I'd like to see the reasoning behind not using one of the living architectures, like ARM, POWER or SPARC. AFAIK MIPS is dead for general-purpose computing and will live on embedded only, like the PowerPC.
In which aspect? Actually for playing CDs I dumped XMMS for Totem
>
ARTS is independant of KDE
Doesn't it use Qt?
>
Koffice was started, IIRC, before OpenOffice ever existed.
The whole of KDE was started before Gnome ever existed. Even Gnome Office was started before OpenOffice ever existed. OpenOffice isn't even a proper Gnome app, it is just a foreign port, like Mozilla; but it is the most mature office automation suite at the moment.
That said, I'd love to see it ported natively to Gnome, just as I'd love to see Gecko and Necko too going native.
>
I presume you are referring to PS/2 PORTS, not the PS/2 line of computers
Yes, obviously. My failure not being clear at all. BTW PS/2 ports is a difficulty for me: we should all be in USB nirvana for input peripherals -- not storage and imaging -- but AFAIK IBM Selectronic keyboards (Model M) are only available for PS/2 ports.
>
They also were extremely expensive
Because of both quality and proprietariness
>
producing components for the microchannel architecture was much more expensive than ISA
ISA is also cheaper than PCI, so what? The problem was being proprietary, so that production volumes were divided between open but ugly hack EISA and beautiful but proprietary MCA.
Slashdot Mirror
Actually not. The (reformed) BSD license and friends do allow anyone to do anything but stripe out the copyrights notice. So if Linux was BSD-licensed, SCO could have gotten away with its reasoning, and IBM wouldn't have been able to point SCO's distribution of Linux as hypocritical and infringing.
Moderators, please, +1 Funny.
Now seriously, I mentioned justice, not democracy. They don't necessarily come together, nor is one necessarily caused by the other.
How practical would that be? Probably not enough to make a dent at the original program community, as any patches if useful would probably be small enough to be quickly replicated, and would be a pain to follow up with new versions.
Now, suppose these patches contain the implementation of an essencial patent... well, this are just like binary modules for the kernel and codecs for mplayer; they exist, but they tend to become roadkill to further development, as most patents are interesting but liable to workarounds.
Good:
Once for all this "not yet tested in court" FUD will go away, and future violators might be deterred.
Bad:
Media circus. 'nough said.
Worst:
A bad result might make thousands of talented people loose faith in justice. This is actually good, as would be any acts of civil disobedience in consequence. The actual evil has been done in the last several decades in the slow, generalised erosion of morals... final defeat would be ugly, even the civil disobedience struggle itself would be honourable but containing scenes of unthinkable ugliness, stupidity and cruelty.
Perhaps...
Why that? I'm more on the Creationist camp.
This is more like it. Thank you for enlightening me.
You are probably right, but not only... there are lots of biologists who have evolution as a working hypothesis for changes, yet think evidence is not strong for it being the sole mechanism of the creation of life and personality.
This is one issue, how many gaps there are. But remember to consider also the other issue: how wide the gaps are. Lots of gaps, not only between us and lesser primates but also among other primates and in several other points of the evolutionary 'family tree', are big enough that not only one or two fossiles or species are needed to fill up, but several.
It's always temerary to hold this as definitive, but from what I talked with a microbiologist, they (biologists) have next to no hope of filling any significant gap. Basically they think that, like petrol, most has already been discovered; there is potential for new discoveries, but nothing proportional to the quantity and size of the gaps.
This would be more like evolutionary biologists than anthropologists; the later are concerned with man as man, not as an animal.
Anyway, tall order. It is not a specie that will fill the gap. There would need to be a big number of fossiles and (or) living species discovered to fill the multiple gaps in evolutionary evidence, and not only near man but all over the classification of animals and vegetables.
Yes. NIS+ would be more similar to AD.
At the time servers were incredibly expensive, and proprietary. Users, in typical self-defeating short-sightedness, preferred to buy their own PCs instead of funding IT to give them professional services and request really open systems. IT ended up supporting a incredibly more costly and complex structure.
For example? I can think now only of portables and some really, really high-end technical workstations. Perhaps I'm forgetting some, but it seems to me that nowadays people try to fit all and any situations to a client-server solution...
Not always, but almost. I did that, I know. Even with inefficient Terminal Server, users were wooed by the incredible performance they won, not to mention reliability and being able to work from anywhere.
I believe portable computers should be supported, even with more sophistication than now -- stuff like Intermezzo and Coda come to mind. But this extra complexity should have a price: IT should charge the department of the users accordingly. The price differente from supporting a X terminal to a portable would be enough dissuasion for most people...
Yes, and I believe host and terminals should be the core of most organisations' networks, and client-server should be reserved for exceptional cases.
To tell you the truth, I believe each apartment building or houses block should have its small server room with two or three machines hosting all apps for a X terminal at each room...
This is typical arguing from the conclusions. They already know what they want to find; now it is just a matter of crafting the studies that will find exactly that, not the reality.
But why be serious?
Like being so incredibly complex no one can properly manage permissions, hiding information so no one can actually understand what's going on, and making it nearly impossible to log stuff? Not to mention hiding source code so it can't possibly be audited? Now, that's security for me...
Like there will be a Microsoft Debian distribution with everything and the kitchen sink installable from CDs or network servers, following a coherent policy? Wow... can't wait for that... will take too long!
Like no more incredibly expensive MSCE that cost even more by always choosing the most inefficient solution? No more incredibly complex, expensive licensing that charges double for Terminal Server usage? Suddenly MS SysAdmins will be as efficient as their Unix counterparts? Toto, we're not in Kansas...
Perhaps I'm misunderstading you, but isn't that something NIS and NIS+ have been doing for ages?
If both servers and clients are uniformly GNU/Linux, it can be easier than MS Windows. Not to mention that easiness is not the ultimate measure: there are more important issues like security, performance, reliability, efficiency...
But most important, client/server sucks. The holy grail of systems administration is host and terminals, and there X Windows exceeds by enabling X terminals. Just wait until we replicate SunRay functionality...
And pray, how is this different from media contamination? This is all about them choosing their facts...
Interestingly enough, the press release says OpenSource Now is for GNU GPL code. Is it excluding other licenses, say public domain, BSD, OSL or even the GNU LGPL and FDL? I can understand focusing on copyleft, but not exclusively GPL.
Granted it's not an application, it is a programming and user interface framework. It is portable, being runnable mostly everywhere there's POSIX; Mono can potentially make it even more portable, but that remains to be seen.
Money isn't easy to come by nowadays. I guess having a real company behind may be beneficial to Ximian's business plans, but in any case it is mostly GNU software the community will either take in case of a Ximian debacle, or the FSF will.
OTOH, I don't like big companies. Usually merges and buyouts diminish their ability to compete.
How bringing a suit against IBM and spreading FUD violates the GNU GPL? What has the FSF to do with that?
This was a nice excuse until the invention of the press and the Reformation. After that, the Bible "had" to be put on the Index Librorum Prohibitorum, and stayed there for four long centuries. One needed the bishop's written authorisation to read it. Even classical texts were bowdlerised to use in the Jesuit schools. Protestant Bible colportors were expulsed from Catholic cities and villages until the 1.970's in Northeastern Brazil, in some instances with the help of the cangaceiros bandits; are still being expulsed from Mexican pueblos. The priests slandered the Protestants with mistranslations, and wouldn't allow their own translations to be issued with extensive notes contradicting the sense of damning passages with out-of-context quotes and controversial opinions by later Fathers of the Church.
That he is wise enough to not let his words be misused out of context as free advice by people who can't get their clues. Like you.
FSF, not EFF. And everytime someone discovers a GNU GPL violation, it gets fixed. The burden of proof is on you to discover an instance of unpursued violation.
Why do we need? I am far more assured by violators' prudence in complying to avoid a probable defeat.
SCO is not picking the terms. In fact, in countries when this is possible GNU/Linux user groups have already ask the courts to shut them up -- Germany & Australia. But this has nothing to do with a GNU GPL test case, it is just preventing slandering. SCO hasn't yet provided any evidence, so there is no way of the FSF doing anything now. Even when they do, the Linux kernel has little code in it that has been assigned to the FSF, and in general it uses the GNU GPL with the famous binary modules exception. So the FSF has no stake in it. Now if IBM or the Linux developers decide they need FSF's help, I'm sure Eben and co. will be eager to help, but somehow I feel IBM lawyers need no help...
All things considered, this is not a copyright case, so it has no value to validate the GNU GPL except as a side effect of Caldera GNU/Linux distribution. Given that, what do you suggest the FSF do exactly? There is no case at hand that I can see.
Can you expand on this?
He has been doing this for years. Proof is, there have been scores of GNU GPL violations, and not one of the violators has taken his chances on court.
Hmm... not. CISC and RISC define the interface. CISC presents a real problem because the decoding of CISC instructions into the internal architecture, usually RISC but also VLIW, takes chip real state and design efforts, thus delaying development, making the final product more expensive and less efficient, and blocking advancement of the whole architecture.
I am not interested in Intel and AMD. What do you offend me for?
The real history here is not the foundry, but the architecture. RISC is simpler, not only to run fast but also to design, manufacture and perhaps reverse engineer.
Interestingly, the very first link says fully MIPS compatbile.
I praise them from breaking free from CISC and avoiding the VLIW pitfall. But I'd like to see the reasoning behind not using one of the living architectures, like ARM, POWER or SPARC. AFAIK MIPS is dead for general-purpose computing and will live on embedded only, like the PowerPC.
In which aspect? Actually for playing CDs I dumped XMMS for Totem
Doesn't it use Qt?
The whole of KDE was started before Gnome ever existed. Even Gnome Office was started before OpenOffice ever existed. OpenOffice isn't even a proper Gnome app, it is just a foreign port, like Mozilla; but it is the most mature office automation suite at the moment.
That said, I'd love to see it ported natively to Gnome, just as I'd love to see Gecko and Necko too going native.
No one ever said it, even if the bigger you are, the more responsibility you have.
It is. Why not?
I care about freedom, sad thing you don't.
Yes, obviously. My failure not being clear at all. BTW PS/2 ports is a difficulty for me: we should all be in USB nirvana for input peripherals -- not storage and imaging -- but AFAIK IBM Selectronic keyboards (Model M) are only available for PS/2 ports.
Because of both quality and proprietariness
ISA is also cheaper than PCI, so what? The problem was being proprietary, so that production volumes were divided between open but ugly hack EISA and beautiful but proprietary MCA.