(talking x86 here, it's usually quite similar on other arches)
by default the inb/outb family of instructions is privileged - so you need supervisor/ring0 (ie. kernelmode) to execute them and there's a trap otherwise (which delegates to kernel, which can _check_ your arguments against some tables and grant your call, or not)
unfortunately in OSS unices there's a workaround in that root (uid 0) gets those calls for free to make xfree86 (and derivatives) happy. this workaround doesn't check if you're actually writing to the video card (so reprogramming the IDE controller would work, if you're fast enough to fit your data between two attempts of the kernel to manage IDE)
it has some annoying defaults inside realplayer (which you can change in the options) and doesn't affect anything else (except installing browser plugins, installing icons - ie. standard stuff)
realplayer these days is a different beast from realplayer in 2001
I'm not so sure what monotone should learn from git..
the speed problem was a bug and lack of optimization in "fresh" code (that extensive sanity checking didn't exist in the version before the one linus tried) - the bug is gone (0.18 release), optimizations are on their way in (0.19 release, planned for monday or so)
monotone is fast these days.. maybe not as fast as git (yet), but otoh monotone does merging, auth and optimized p2p data exchange
no, IBM just uses its own MPL derivative (eclipse's license) or the CPL - which have basically the same terms - for most of its projects (except those it didn't start, where it has to abide to the original license)
fbsd5 might be "fixed" somehow these days, but it's still a pretty complex piece of code.
dfly attempts to remove this complexity (instead of routing around its problems, introducing more complexity), making it easier to work on productive code instead of fighting with thousands of locks.
because european companies are subjected to US fines, too, if they do business in the US and violate american laws - it's normal, and it's good that way, otherwise all those companies would incorporate on some pacific island without any law and export to all other countries avoiding the rules the local economy has to abide to.
oh, and btw. the EU fines european companies all the time, it just happens that we don't have a high-profile monopolist who breaks the rules in software business over here, so you probably just didn't notice it.
in that case, microsoft better settles by abiding the _real_ deal (ie. opening the specs for everyone without silly restrictions) on at least some cases
the car manufacturers don't have to disclose the codes, it's just that reverse engineering of the whole process can't be prevented using the DMCA (assuming that the lexmark case can be considered to be comparable).
after all, lexmark isn't forced to open the spec to the cartridges either
so where does the CDDL (or MPL, where CDDL comes from) not match the criteria that OSI, the FSF and debian apply to licenses?
(essentially to be a "free software" license, such as GPL, MPL or CDDL, you have to match the same rules for all three entities, it's just that there are different reasons behind those rules)
from the horse's mouth: (http://www.fsf.org/licenses/license-list.html)
Mozilla Public License (MPL)
This is a free software license which is not a strong copyleft; unlike the X11 license, it has some complex restrictions that make it incompatible with the GNU GPL. That is, a module covered by the GPL and a module covered by the MPL cannot legally be linked together. We urge you not to use the MPL for this reason.
However, MPL 1.1 has a provision (section 13) that allows a program (or parts of it) to offer a choice of another license as well. If part of a program allows the GNU GPL as an alternate choice, or any other GPL-compatible license as an alternate choice, that part of the program has a GPL-compatible license.
----------------------
so, MPL itself is free, but GPL-incompatible, but explicitely allows for other licenses..
btw, CDDL is based on MPL 1.1, so it might just be the same wrt "freedom" and "GPL compatibility"
CDDL is just as "free" as GPL, just incompatible (same situation as with MPL and a whole bunch of other "accepted" free, GPL-incompatible licenses)
it's just the decision whether you prefer the details of gpl, mpl, cddl, or one of the other hundreds free licenses out there. and yes, it effectively prevents opensolaris code to enter linux, boohoo - the same happened with gnupg being gpl (incl. all interface libraries) which made enigmail (gnupg for mozilla mail - which is MPL'd) pretty complicated
the CPL is a Free (caps-F, libre, GNU, FSF, "hippie", "commie", you name it) license, too (according to the FSF), and still not GPL compatible. being "GPL incompatible" is not a problem in itself (it might even be a strength)
Re:Another "Innovation" from Microsoft?
on
A .Net CPU
·
· Score: 1
shared source would be equivalent to SCSL.. SUN already upgraded to the "java research license" with jdk1.5 (still not OSI, but definitely nicer than SCSL)
so, the race for the first official VM and DK for one of the languages under an OSI approved license is still undecided
Slashdot Mirror
(talking x86 here, it's usually quite similar on other arches)
by default the inb/outb family of instructions is privileged - so you need supervisor/ring0 (ie. kernelmode) to execute them and there's a trap otherwise (which delegates to kernel, which can _check_ your arguments against some tables and grant your call, or not)
unfortunately in OSS unices there's a workaround in that root (uid 0) gets those calls for free to make xfree86 (and derivatives) happy.
this workaround doesn't check if you're actually writing to the video card (so reprogramming the IDE controller would work, if you're fast enough to fit your data between two attempts of the kernel to manage IDE)
what's the issue with real10?
it has some annoying defaults inside realplayer (which you can change in the options) and doesn't affect anything else (except installing browser plugins, installing icons - ie. standard stuff)
realplayer these days is a different beast from realplayer in 2001
well, the law - as is - only states that over time, the probability that any discussion involves a comparison with hitler or nazis approaches 1.
any conclusions (as in, that the discussion can be stopped afterwards) is a later addition (and imho a faulty one)
Jargon entry
interesting, how CPL is a "very nice license", while its brother, the CDDL is usually regarded pure evil over here..
/. isn't hive mind - but hey, you weren't even modded as troll for that praise of the CPL!
I know,
I'm not so sure what monotone should learn from git..
the speed problem was a bug and lack of optimization in "fresh" code (that extensive sanity checking didn't exist in the version before the one linus tried) - the bug is gone (0.18 release), optimizations are on their way in (0.19 release, planned for monday or so)
monotone is fast these days.. maybe not as fast as git (yet), but otoh monotone does merging, auth and optimized p2p data exchange
no, IBM just uses its own MPL derivative (eclipse's license) or the CPL - which have basically the same terms - for most of its projects (except those it didn't start, where it has to abide to the original license)
fbsd5 might be "fixed" somehow these days, but it's still a pretty complex piece of code.
dfly attempts to remove this complexity (instead of routing around its problems, introducing more complexity), making it easier to work on productive code instead of fighting with thousands of locks.
sun jre can be built on freebsd (been there, done that), and most likely on netbsd and openbsd, too.. after all, you can get the source.
ppc or arm are different issues, as the JIT would have to be adapted
because european companies are subjected to US fines, too, if they do business in the US and violate american laws - it's normal, and it's good that way, otherwise all those companies would incorporate on some pacific island without any law and export to all other countries avoiding the rules the local economy has to abide to.
oh, and btw. the EU fines european companies all the time, it just happens that we don't have a high-profile monopolist who breaks the rules in software business over here, so you probably just didn't notice it.
in that case, microsoft better settles by abiding the _real_ deal (ie. opening the specs for everyone without silly restrictions) on at least some cases
I wasn't aware that the original vi (by sun) was released as open source. are you sure that gentoo's vi isn't just nvi or elvis?
the car manufacturers don't have to disclose the codes, it's just that reverse engineering of the whole process can't be prevented using the DMCA (assuming that the lexmark case can be considered to be comparable).
after all, lexmark isn't forced to open the spec to the cartridges either
that was back in qt-2.3 times - ie. ancient stuff
so where does the CDDL (or MPL, where CDDL comes from) not match the criteria that OSI, the FSF and debian apply to licenses?
(essentially to be a "free software" license, such as GPL, MPL or CDDL, you have to match the same rules for all three entities, it's just that there are different reasons behind those rules)
http://www.opensource.org/docs/definition.php
so it's still a free software license? ask RMS to fix the GPL then, after all, _that_ can be changed, right?
(oh, and ask linus to adopt gpl3 for his code then, as he dropped the "or later" option)
you couldn't implement the concepts on your own before already, so that they offer the patents for CDDL-projects can only make the situation better.
well, implicit crosslicensing - nice way to get rid of patents, or not?
from the horse's mouth: (http://www.fsf.org/licenses/license-list.html)
Mozilla Public License (MPL)
This is a free software license which is not a strong copyleft; unlike the X11 license, it has some complex restrictions that make it incompatible with the GNU GPL. That is, a module covered by the GPL and a module covered by the MPL cannot legally be linked together. We urge you not to use the MPL for this reason.
However, MPL 1.1 has a provision (section 13) that allows a program (or parts of it) to offer a choice of another license as well. If part of a program allows the GNU GPL as an alternate choice, or any other GPL-compatible license as an alternate choice, that part of the program has a GPL-compatible license.
----------------------
so, MPL itself is free, but GPL-incompatible, but explicitely allows for other licenses..
btw, CDDL is based on MPL 1.1, so it might just be the same wrt "freedom" and "GPL compatibility"
CDDL is just as "free" as GPL, just incompatible (same situation as with MPL and a whole bunch of other "accepted" free, GPL-incompatible licenses)
it's just the decision whether you prefer the details of gpl, mpl, cddl, or one of the other hundreds free licenses out there.
and yes, it effectively prevents opensolaris code to enter linux, boohoo - the same happened with gnupg being gpl (incl. all interface libraries) which made enigmail (gnupg for mozilla mail - which is MPL'd) pretty complicated
they apply the same rules to licenses to be considered "open source" (OSI) or "free" (FSF).
the only difference is philosophy
similar to the way they got to 2:
java1.2=java2
java1.5=java5
marketing stuff (and yes, 1.5 would actually warrant a new major number, for the generics are quite some change)
the CPL is a Free (caps-F, libre, GNU, FSF, "hippie", "commie", you name it) license, too (according to the FSF), and still not GPL compatible.
being "GPL incompatible" is not a problem in itself (it might even be a strength)
java is open source for years now, and java5 got a pretty neat license, too.
http://java.sun.com/j2se/jrl_download.html
it's a private research institute, thus a company
shared source would be equivalent to SCSL.. SUN already upgraded to the "java research license" with jdk1.5 (still not OSI, but definitely nicer than SCSL)
so, the race for the first official VM and DK for one of the languages under an OSI approved license is still undecided