One of the big myths about Microsoft's stuff is that it is ``easier'' to administer or somehow less complex than *nix.
When the point-N-click works, it's great. It all begins to turn to pooh when one of these things happens:
You want to do something that the interface designers didn't think of; or
The interface designers think differently to you (one generally can't grep GUI menus, dump them as a resource, or ``man -K'' for keywords in them); or (meltdown time...!)
the GUI doesn't actually work (I have raves elsewhere on/. with detailed examples)
Never confuse a plethora of choices with real freedom. Never confuse pretty with useful. Never confuse slick packaging with thorough testing.
Where OSS such as Linux wins is:
Everything is a tool and can be used in ways their designers never imagined; and
Everything can be redesigned as needed (sometimes the nice package does almost what you need); and
There are usually a whole passel (maybe even a slew) of ways of doing any one thing
Let's not even bother exploring the ugliness represented by layer upon layer of legacies and idealogically incompatible subsystems in the various Microsoft products.
About year ago, a different fellow consultant, an NT expert who also does Linux, was losing hair over a network connection port-forwarded from a Linux gateway box to a masqueraded NT 4.0 server running a Pick app.
An aside: the company concerned (call them BB) had been told to buy an NT box for their app, even though it cost several thousands of dollars more than running it under Linux. Two weeks after commissioning, the provider was bought by a Linux shop, technician from which promptly asked MD of BB why he hadn't bought the Linux version, because it was more reliable and used less resources. Gngnngnngnngn!
Anyway, problem turned out to be that the NT box was gatewayed to... 127.0.0.1! No problem, change gateway to point to the Linux box. Still no go. At my insistence, NT guru does ROUTE PRINT from the command line; system is still gatewayed to localhost. Hmm. Reboot (this is Windows, after all). No change.
Redo the route from the command line... viola! Life and happiness! We wound up running a BAT file on startup. Yerk, but it works.
More recently, similar story with an NT 4.0 DHCP server, changed to nail down a server to allow port-forwarding, fix gatewaying and generally coalesce the bizarre and disparate settings on all of the Windows boxes on that LAN around a sane concensus. No matter what we changed, the DHCP server still allocated the wrong settings. Amongst other things, DNS service was aimed out through the gateway and twice across the Nullarbor to an ISP that this LAN hadn't been connected to for at least 2 years...
Finally, we had to (1) create a new subnet; (2) empty the DHCP server config completely, starting with the old subnet; (3) reboot (surprise: stopping and restarting the service wasn't enough); (4) make a new subnet (and yes, the settings defaulted to the abberrant ones so had to be overridden by hand); and (5) stop and restart the DHCP service.
The usual Linux equivalent of the NT graphical route editing tool (linuxconf) pulls its config from the text tools, so they cannot disagree (and has the additional advantage of a low-bandwidth/low-hardware 2D non-graphic mode if you want it); the usual Linux DHCP server (daemon) keeps config in a text file, and reliably reconfigures itself on a hangup signal. Of course, you have a choice about (and source for) both of these services.
Reading any Microsoft release notes (no matter how carefully or how many times) would have left us entirely unprepared for either of these eventualities. Neither would have assuming or not assuming any amount of stuff.
In short, you're talking out of your fundment. Poor, stupid Anonymous Coward, Microsoft don't provide either config for the filter to temporarily or permanently reduce its enthusiasm, or any means of backing out the patch.
Fellow-consultant of mine spent about two years assembling a large system using MS-only development tools. Four weeks before commissioning, MS announce that (1) said tool won't run on next version of OS, which final client is currently rolling out; (2) next version of tool will run on new version of OS; (3) next version of tool is incompatible with current version; and finally (4) there will be no upgrade/migration tools. Consultant had not written events like this into contract; I doubt client would have stood for it anyway.
Bottom line: consultant had to rebuild the entire app for free (meanwhile somehow continuing to eat, pay off house, run car etc), which took him just over a year, after paying for a complete new set of tools.
But, instead of reading and understanding the release notes, the programmer scanned the notes and assumed a great deal.
Ah. So you understand every word of the US Federal Register, do you? As a US citizen, you are obliged to know it all and consequently can be jailed for rules you have no hope of knowing (enough pages are added every day that you would have to do nothing but read (no sleep, probably no meals) to keep up).
But even more pointed, OutLook is so fundamentally insecure and badly structured internally that any ``real'' security patch absolutely has to be a bastard. Ditto for Word, Excel, name it. One reason for this is that Windows, on which these all rely, is fundamentally a single-user system.
Even derived-from-VMS NT has been knackered down to a fancy kind of single-userness. This (and a good deal of the excess baggage in W2k/XP) has been necessary because their premiere apps have depended on it. Remember the brouhaha about MS apps ``cheating'' by using undocumented OS calls...? Well, the biter has now been bit.
Not surprising for something that has never moved.
I've seen a fellow consultant moved by them - to tears, in fact - when they obseleted the development tool that he had just finished a $Oz100K contract with, and the next version was not compatible so he had to rewrite from scratch.
The easiest and sfaest way of achieving (1) and, circumstances permitting, some of (2) is to make sure that at least 60% of what you eat is raw. This fixes and/or helps to fix a heck of a lot more than back pain.
...purportedly for doing good things, which in the case of, say, the Salvation Army is at least believable. Many churches make immense profits and hoard staggering wealth (statues in St Paul's Basilica, for example, are worth $billions each). So why can't Mandrake, who have been doing good things (including, recently, giving a trendy Yankee marketing-bedazzled CEO the bum's rush), not be a target for donations?
And to relate to you more directly, BeCool, WWJD? Well, not go to school for starters - Jesus trained at home to be a responsible citizen and a carpenter. He didn't suffer the damage that factory-school-based ``socialisation'' (regimentation) inflicts.
When Jesus switched careers at maturity (age 30) to full-time public speaker, recruiting officer, pharmacopoeia-less healer and occasional caterer, the new job description included giving good stuff away (life, healing, salvation, assurance, information) for free and living on donations. Is there a parallel with Mandrake? (-:
count yourself bloody lucky that these companies give you their work for free.
Hear, hear!
Go and have a look back at, for example, the old Mandrake Cooker archives (I'm a list member there). Time and time again, people wrote to say ``Hey, I really like your distro but downloaded it instead of buying a boxed set because the box was [too slow/unavailable in my area/Had extras I didn't want/etc]. Is there some way I can give you some money to offset the cost of providing that free download?'' Mandrake caved in after about a year of this and provided the donations link.
I am left with the impression that Mandrake management didn't quite believe it, until the money started flowing...
Even if they're not up the instant you type, they might well be a minute later, and certainly will be by the time most people get to read your article later in the day. It would help kernel.org to not get slashdotted.
This ``probably not updated'' rubbish just doesn't cut it, either. How long would it take you to check? In seconds? Whatever happened to responsible reporting, the kind so often bemoaned on his very site for its lack?
The only question raised here is, am I really going to trust this "helpful" worm or others like it to fully patch up my box properly?
So what are you going to do? Put your unpatched antique box on the net and hope Cheese finds it before Ramen? Ahuk, ahuk, ahuk...
The bottom line is: if your security sucks, you default to trusting every Tom, Dick and Harry out there with your box. The usual term for this is ``data suicide''.
immediately and globally engage in mass copying, uploading, downloading and distributing of all copyrighted and patented materials.
In this case organisations like the Seed Savers Network are protecting examples of prior art by ``mass copying, uploading, downloading and distributing''. Kind of outdoes RFC 1149 or RFC 2549.
The saved seeds are far superior collections of genetic material, in that the patented seeds are closely bred (ie, ``thin'' genetic material, won't breed true) and/or genetically modified, so almost always require special (expensive, proprietary) fertilisers, pesticides etc ad nauseum in order to produce their huge yields.
Finally, the whole idea is an open source/sharing kind of thing, very much in tune with the current software revolution.
...and consequently, SlashDot is as good a venue as any to look for explanations.
Perhaps the stars really are embedded in a crystal sphere, and the perturbations are the gravitic effects of the sphere beginning to manifest. Watch for a sudden end to transmissions and soon afterwards the mother of all cracks in the sky... (-:
``But seriously, folks''
NASA don't know how our own sun works, so how can they be expected to predict behaviour even further away from the sun? How is it that a (ghasp) lone scientist with no resources can bullseye planetary magnetic fields before the fact, but NASA (besides many other large and well-equipped organisations) are several orders of magnitude wide of the mark?
Ah, yes, it all becomes clear now. At least four orders of magnitude in error is ``100% reliable''.
So... I've got a real nice bridge here, hardly used, previous owner (a little old lady) only ever drove over it on Sundays(*); it's got a good, steady revenue stream from the tollgates; no liability for suicides; magnificent outlook; as pictured on millions of postcards; easy terms available. Interested?
I've rescued a couple of dead sites out of it already, and been able to rip stuff from ``obsolete'' pages when it had disappeared from the original site.
I'm sure there's a reason it's making the news now
Because a lot of people (and with good reason) aren't entirely sure about the Big Bang (-: or as another poster phrased it, ``the Horrendous Space Kablooie'':-). People don't like being unsure that something foundational to their worldview might actually be wrong, so they either stuff it down your throat at every conceivable opportunity, or avoid the topic altogether.
For another example of this effect in science, consider the number of ``birdosaurs'' or of ``pre-humans'' that have been discovered, proclaimed at full volume, and then quietly denied in small print as every single one is discovered to be a furphy of one kind or another. Some people have a need to believe that a dinosaur is where we got birds from, so anything that threatens this belief becomes a target for their zealotry. If someone convincingly ``proved'' that birds evolved from amphibians, and this ``proof'' was widely proclaimed and accepted as true, these zealots would be at a total loss for words.
If that evidence remains consistent with the theory, then we consider it a good theory. If it doesn't, we might still consider it a good theory, but work on it some more.
OK, a nice tongue-in-cheek statement. So when is a theory ever considered to be a bad theory?
Some of Velikovsky's ideas seem to be classed as ``bad theories'' despite being rather good predictors of observation. Is this a matter of politics or sociology rather than science?
Some of the plasma cosmologies also seem to be excellent predictors of observation, and in some cases the only sane explanation of observations, yet they are still not treated as a ``good theory'' - so what needs to happen in order for them to be considered ``good?''
If the answer is ``a quorum of acceptance'' can you then go on to explain how that is science and not politics? (-:
They dealt with this long ago, even AiG have an answer for this latest incidence of notionally feathery megalizards.
ICU[^HR] will probably laugh their butts off
on
Feather Dino Fossil
·
· Score: 2
No, wait, they already have...
The ratio is at least thousands of provably damaging mutations to one notionally helpful mutation. Sickle cell anaemia is not an example of a helpful mutation, as it is 25% lethal and 25% ineffective. The duds (if collectively fertile they are inheritable) are called ``genetic burden'' and one of the consequences of discovering this was a conference to decide whether mutation plus natural selection cold be sufficient to drive evolution as she is conventionally understood. The answer was ``No''. Not maybe, not at long odds, not in some circumstances, just ``No''.
I can dig out refs if required, but basically evolution is once again in the embarrassing position of saying, ``we don't know how it happened, but we believe that it did''. Now that's blind faith!
One of the biggest ideals of free software is CHOICE.
Boy, would Richard Stallman bounce you hard for that one! Choice is the poor, crippled cousin of freedom. What the Argentinian gummint is doing is curtailing the freedom of suppliers a little (not completely: nothing stoppping them from Opening their products and soldiering on) in order to protect their own freedom. Which is a very good idea.
If they mandated this for their citizens as well, that would not be so crash hot. It would be better to mandate open interfaces (file formats, APIs etc) and penalise corporations who fail to comply with their own APIs.
Open-sourced and free software follow the process of evolution
I very much agree. Unlike biological evolution, Open Source clearly admits an intelligent design mechanism and directed feedback.
Biological evolution postulates mutation as an information-generating mechanism, against all common sense, mathematical principles and every shard of experimental evidence. This is so absurd a practice that bizarre ideas like entelechy are offered as alternatives.
Mutation is about as useful for providing or improving biological design as a high-powered rifle is for providing or improving Lego layouts.
To make the US gummint purchasing system work for software, all you need add is Free (libre) Software. They may get shafted once, but only once, in each segment of the market.
the CIA [...] won't even _touch_ GPL'd software, because they have to modify everything
I hate to come across all Jesuit about this, but it seems from your statements that the CIA don't understand the GPL: the modified source only has to be distributed as much as the modified binary. You could comply by shipping your GPL-software-containg grey box with a source CD inside it (taped/clamped inside the cover).
can't make anything known publicly about what exact modifications they have made
This is either perfectly normal paranoia (``every being in the universe has that'') or very bad software design. If your software becomes insecure when the source for it becomes available, you have no business doing security in the first place.
If the device the CIA are working with shouldn't exist, or if they're exceeding their mandate in dealing with something, then what they are already involved in is far more problematic to them than a mere software licence. To put this into perspective: if you are ignoring laws about dealing with drugs, murdering people, prohibited weapons etc then you're not going to give two pinches of gecko poo about a mere software licence, are you?
And I might suggest to any enemies of the CIA that it may be safer to rely on a black-box analysis of a dodgy CIA program than on the source.
There's nothing preventing Microsoft from opening their software and continuing to not be paid for it by Argentinians. (-:
While this government regulation is selling some freedom (per se) it is maintaining at least a broad range of choices as a consequence. And in this case the tradeoff seems to be a good one. The best program for the job should win out, but that presumes the absence of salesmen and back-room deals (leaving aside proprietary lock-in effects), which in the case of a normal South American country would be an extremely stupid presumption to make.
Seriously, this is not a law, it is an internal gummint regulation. The Argentinian gummint departments simply abide by it. If the director of a department refuses to comply, he promptly stops being the director, and the next director complies.
By tradition this is a pro-M$ and anti-Un*x country, mostly because in the 80's we had a prohibition to import anything
I wasn't aware that Microsoft was headquartered in Argentina (I've always heard of them as being ``in Redmond, Washington''), so I'm a bit confused about the mechanism by which Windows is deemed native and Unix is deemed foreign.
Linux, in particular, has Argentinian contributors (Quiz question: is there a country from which nobody has contributed to the Linux kernel?) so would by any normal classification system be regarded as less importado than Windows.
Microsoft must have slipped one past the Argentine populace ``tal como mantequilla en un chango pelon'' (-: BTW, since I guess you speak fluent Argentine, commentary on the language abuse in the above link would be welcome:-)
Slashdot Mirror
When the point-N-click works, it's great. It all begins to turn to pooh when one of these things happens:
Never confuse a plethora of choices with real freedom. Never confuse pretty with useful. Never confuse slick packaging with thorough testing.
Where OSS such as Linux wins is:
Let's not even bother exploring the ugliness represented by layer upon layer of legacies and idealogically incompatible subsystems in the various Microsoft products.
About year ago, a different fellow consultant, an NT expert who also does Linux, was losing hair over a network connection port-forwarded from a Linux gateway box to a masqueraded NT 4.0 server running a Pick app.
An aside: the company concerned (call them BB) had been told to buy an NT box for their app, even though it cost several thousands of dollars more than running it under Linux. Two weeks after commissioning, the provider was bought by a Linux shop, technician from which promptly asked MD of BB why he hadn't bought the Linux version, because it was more reliable and used less resources. Gngnngnngnngn!
Anyway, problem turned out to be that the NT box was gatewayed to... 127.0.0.1! No problem, change gateway to point to the Linux box. Still no go. At my insistence, NT guru does ROUTE PRINT from the command line; system is still gatewayed to localhost. Hmm. Reboot (this is Windows, after all). No change.
Redo the route from the command line... viola! Life and happiness! We wound up running a BAT file on startup. Yerk, but it works.
More recently, similar story with an NT 4.0 DHCP server, changed to nail down a server to allow port-forwarding, fix gatewaying and generally coalesce the bizarre and disparate settings on all of the Windows boxes on that LAN around a sane concensus. No matter what we changed, the DHCP server still allocated the wrong settings. Amongst other things, DNS service was aimed out through the gateway and twice across the Nullarbor to an ISP that this LAN hadn't been connected to for at least 2 years...
Finally, we had to (1) create a new subnet; (2) empty the DHCP server config completely, starting with the old subnet; (3) reboot (surprise: stopping and restarting the service wasn't enough); (4) make a new subnet (and yes, the settings defaulted to the abberrant ones so had to be overridden by hand); and (5) stop and restart the DHCP service.
The usual Linux equivalent of the NT graphical route editing tool (linuxconf) pulls its config from the text tools, so they cannot disagree (and has the additional advantage of a low-bandwidth/low-hardware 2D non-graphic mode if you want it); the usual Linux DHCP server (daemon) keeps config in a text file, and reliably reconfigures itself on a hangup signal. Of course, you have a choice about (and source for) both of these services.
Reading any Microsoft release notes (no matter how carefully or how many times) would have left us entirely unprepared for either of these eventualities. Neither would have assuming or not assuming any amount of stuff.
In short, you're talking out of your fundment. Poor, stupid Anonymous Coward, Microsoft don't provide either config for the filter to temporarily or permanently reduce its enthusiasm, or any means of backing out the patch.
Bottom line: consultant had to rebuild the entire app for free (meanwhile somehow continuing to eat, pay off house, run car etc), which took him just over a year, after paying for a complete new set of tools.
Ah. So you understand every word of the US Federal Register, do you? As a US citizen, you are obliged to know it all and consequently can be jailed for rules you have no hope of knowing (enough pages are added every day that you would have to do nothing but read (no sleep, probably no meals) to keep up).
But even more pointed, OutLook is so fundamentally insecure and badly structured internally that any ``real'' security patch absolutely has to be a bastard. Ditto for Word, Excel, name it. One reason for this is that Windows, on which these all rely, is fundamentally a single-user system.
Even derived-from-VMS NT has been knackered down to a fancy kind of single-userness. This (and a good deal of the excess baggage in W2k/XP) has been necessary because their premiere apps have depended on it. Remember the brouhaha about MS apps ``cheating'' by using undocumented OS calls...? Well, the biter has now been bit.
Sand supplies available here.
The easiest and sfaest way of achieving (1) and, circumstances permitting, some of (2) is to make sure that at least 60% of what you eat is raw. This fixes and/or helps to fix a heck of a lot more than back pain.
...purportedly for doing good things, which in the case of, say, the Salvation Army is at least believable. Many churches make immense profits and hoard staggering wealth (statues in St Paul's Basilica, for example, are worth $billions each). So why can't Mandrake, who have been doing good things (including, recently, giving a trendy Yankee marketing-bedazzled CEO the bum's rush), not be a target for donations?
And to relate to you more directly, BeCool, WWJD? Well, not go to school for starters - Jesus trained at home to be a responsible citizen and a carpenter. He didn't suffer the damage that factory-school-based ``socialisation'' (regimentation) inflicts.
When Jesus switched careers at maturity (age 30) to full-time public speaker, recruiting officer, pharmacopoeia-less healer and occasional caterer, the new job description included giving good stuff away (life, healing, salvation, assurance, information) for free and living on donations. Is there a parallel with Mandrake? (-:
Hear, hear!
Go and have a look back at, for example, the old Mandrake Cooker archives (I'm a list member there). Time and time again, people wrote to say ``Hey, I really like your distro but downloaded it instead of buying a boxed set because the box was [too slow/unavailable in my area/Had extras I didn't want/etc]. Is there some way I can give you some money to offset the cost of providing that free download?'' Mandrake caved in after about a year of this and provided the donations link.
I am left with the impression that Mandrake management didn't quite believe it, until the money started flowing...
Even if they're not up the instant you type, they might well be a minute later, and certainly will be by the time most people get to read your article later in the day. It would help kernel.org to not get slashdotted.
This ``probably not updated'' rubbish just doesn't cut it, either. How long would it take you to check? In seconds? Whatever happened to responsible reporting, the kind so often bemoaned on his very site for its lack?
So what are you going to do? Put your unpatched antique box on the net and hope Cheese finds it before Ramen? Ahuk, ahuk, ahuk...
The bottom line is: if your security sucks, you default to trusting every Tom, Dick and Harry out there with your box. The usual term for this is ``data suicide''.
In this case organisations like the Seed Savers Network are protecting examples of prior art by ``mass copying, uploading, downloading and distributing''. Kind of outdoes RFC 1149 or RFC 2549.
The saved seeds are far superior collections of genetic material, in that the patented seeds are closely bred (ie, ``thin'' genetic material, won't breed true) and/or genetically modified, so almost always require special (expensive, proprietary) fertilisers, pesticides etc ad nauseum in order to produce their huge yields.
Finally, the whole idea is an open source/sharing kind of thing, very much in tune with the current software revolution.
...and consequently, SlashDot is as good a venue as any to look for explanations.
Perhaps the stars really are embedded in a crystal sphere, and the perturbations are the gravitic effects of the sphere beginning to manifest. Watch for a sudden end to transmissions and soon afterwards the mother of all cracks in the sky... (-:
``But seriously, folks''
NASA don't know how our own sun works, so how can they be expected to predict behaviour even further away from the sun? How is it that a (ghasp) lone scientist with no resources can bullseye planetary magnetic fields before the fact, but NASA (besides many other large and well-equipped organisations) are several orders of magnitude wide of the mark?
Ah, yes, it all becomes clear now. At least four orders of magnitude in error is ``100% reliable''.
So... I've got a real nice bridge here, hardly used, previous owner (a little old lady) only ever drove over it on Sundays(*); it's got a good, steady revenue stream from the tollgates; no liability for suicides; magnificent outlook; as pictured on millions of postcards; easy terms available. Interested?
No wonder this coward is anonymous!
(*) on her way to the races
I've rescued a couple of dead sites out of it already, and been able to rip stuff from ``obsolete'' pages when it had disappeared from the original site.
Because a lot of people (and with good reason) aren't entirely sure about the Big Bang (-: or as another poster phrased it, ``the Horrendous Space Kablooie''
For another example of this effect in science, consider the number of ``birdosaurs'' or of ``pre-humans'' that have been discovered, proclaimed at full volume, and then quietly denied in small print as every single one is discovered to be a furphy of one kind or another. Some people have a need to believe that a dinosaur is where we got birds from, so anything that threatens this belief becomes a target for their zealotry. If someone convincingly ``proved'' that birds evolved from amphibians, and this ``proof'' was widely proclaimed and accepted as true, these zealots would be at a total loss for words.
Some of Velikovsky's ideas seem to be classed as ``bad theories'' despite being rather good predictors of observation. Is this a matter of politics or sociology rather than science?
Some of the plasma cosmologies also seem to be excellent predictors of observation, and in some cases the only sane explanation of observations, yet they are still not treated as a ``good theory'' - so what needs to happen in order for them to be considered ``good?''
If the answer is ``a quorum of acceptance'' can you then go on to explain how that is science and not politics? (-:
They dealt with this long ago, even AiG have an answer for this latest incidence of notionally feathery megalizards.
No, wait, they already have...
The ratio is at least thousands of provably damaging mutations to one notionally helpful mutation. Sickle cell anaemia is not an example of a helpful mutation, as it is 25% lethal and 25% ineffective. The duds (if collectively fertile they are inheritable) are called ``genetic burden'' and one of the consequences of discovering this was a conference to decide whether mutation plus natural selection cold be sufficient to drive evolution as she is conventionally understood. The answer was ``No''. Not maybe, not at long odds, not in some circumstances, just ``No''.
I can dig out refs if required, but basically evolution is once again in the embarrassing position of saying, ``we don't know how it happened, but we believe that it did''. Now that's blind faith!
Boy, would Richard Stallman bounce you hard for that one! Choice is the poor, crippled cousin of freedom. What the Argentinian gummint is doing is curtailing the freedom of suppliers a little (not completely: nothing stoppping them from Opening their products and soldiering on) in order to protect their own freedom. Which is a very good idea.
If they mandated this for their citizens as well, that would not be so crash hot. It would be better to mandate open interfaces (file formats, APIs etc) and penalise corporations who fail to comply with their own APIs.
I very much agree. Unlike biological evolution, Open Source clearly admits an intelligent design mechanism and directed feedback.
Biological evolution postulates mutation as an information-generating mechanism, against all common sense, mathematical principles and every shard of experimental evidence. This is so absurd a practice that bizarre ideas like entelechy are offered as alternatives.
Mutation is about as useful for providing or improving biological design as a high-powered rifle is for providing or improving Lego layouts.
To make the US gummint purchasing system work for software, all you need add is Free (libre) Software. They may get shafted once, but only once, in each segment of the market.
I hate to come across all Jesuit about this, but it seems from your statements that the CIA don't understand the GPL: the modified source only has to be distributed as much as the modified binary. You could comply by shipping your GPL-software-containg grey box with a source CD inside it (taped/clamped inside the cover).
This is either perfectly normal paranoia (``every being in the universe has that'') or very bad software design. If your software becomes insecure when the source for it becomes available, you have no business doing security in the first place.
If the device the CIA are working with shouldn't exist, or if they're exceeding their mandate in dealing with something, then what they are already involved in is far more problematic to them than a mere software licence. To put this into perspective: if you are ignoring laws about dealing with drugs, murdering people, prohibited weapons etc then you're not going to give two pinches of gecko poo about a mere software licence, are you?
And I might suggest to any enemies of the CIA that it may be safer to rely on a black-box analysis of a dodgy CIA program than on the source.
There's nothing preventing Microsoft from opening their software and continuing to not be paid for it by Argentinians. (-:
While this government regulation is selling some freedom (per se) it is maintaining at least a broad range of choices as a consequence. And in this case the tradeoff seems to be a good one. The best program for the job should win out, but that presumes the absence of salesmen and back-room deals (leaving aside proprietary lock-in effects), which in the case of a normal South American country would be an extremely stupid presumption to make.
Seriously, this is not a law, it is an internal gummint regulation. The Argentinian gummint departments simply abide by it. If the director of a department refuses to comply, he promptly stops being the director, and the next director complies.
I wasn't aware that Microsoft was headquartered in Argentina (I've always heard of them as being ``in Redmond, Washington''), so I'm a bit confused about the mechanism by which Windows is deemed native and Unix is deemed foreign.
Linux, in particular, has Argentinian contributors (Quiz question: is there a country from which nobody has contributed to the Linux kernel?) so would by any normal classification system be regarded as less importado than Windows.
Microsoft must have slipped one past the Argentine populace ``tal como mantequilla en un chango pelon'' (-: BTW, since I guess you speak fluent Argentine, commentary on the language abuse in the above link would be welcome :-)