I'll agree that certain distros are highly insecure, but equally there are hardened distros that will run essentially the same software. Obviously, software that violates security protocols won't run under a hardened distro (if the distro is any good, that is!) so it's not 100% the same. The lead between Windows and Linux should really be measured from "useful best to useful best" rather than "OTS to OTS" (since nobody runs OTS in practice, all systems are tweaked in some fashion) or "worst to worst" (since all OS' in their worst configuration have no security at all). I would dispute the idea that the best Windows configs are orders of magnitude closer to the best Linux configs, in their hardest configurations - GRSecurity + RBACS + Linux Capabilites + Netfilter + L7 Routing is still a very tough combination to beat in terms of the level of granularity of control. It certainly beats Windows' permissions families plus Windows' firewall in terms of what you can do and what you can restrict.
Agreed that Linux is not a magic security bullet. I wouldn't agree that no OS is unbreakable -- no -useful- OS is unbreakable, yes, but it's quite possible to make a useless OS that's unbreakable. Although, as Microsoft has found, it's also possible to make a useless OS that's very breakable indeed.
Put a hardened Windows system and a hardened Linux system in the same room, then run just standard vulnerability scanners over them. I am willing to bet cold hard cash that the Windows system will be flagged as having potential problems and that the Linux system will not.
(How do I define "hardened"? Microsoft has a free tool that tells you if the system is running in a hardened configuration. For Linux, start with a security distro that includes GRSecurity and/or RBACS and essentially follow the same procedures as advised by Microsoft. If you want specific instructions, the NSA have a guide on how to harden Linux - and indeed how to harden Windows. To be fair, you should apply both.)
Define "business applications" in your context. For many years, "Exchange" was considered a "business application" that "could not run under Linux" - which, whilst technically true, was also technically false since oGo and other near-100% Exchange clones have been around for a long time. SQL Server is only marginally acceptable, as tools like "Dezign for Databases" will transliterate SQL Server databases into other database syntaxes and there's really very little that SQL Server can do that cannot be reproduced by any means whatsoever on Ingres, Oracle, DB/2, Informix or PostgreSQL - all of which run perfectly well under Linux. (You only need one server that supports what you need to do, so it is an "or" not an "and".)
For the most part, that leaves web-based applications (which, by definition, will run on any desktop) and "office productivity tools" (aka "office maiming tools"), most of which either have clones or will run directly. In the case of MS Office, both. (Yes, Office can be run under Linux.)
Because anything that runs under OS/X will be IBCS-compliant, anything that runs under OS/X can be made to run natively (not emulated, NATIVELY) under Linux, although that would require that IBCS support be reintroduced.
There are many, many genes involved in different forms of cancer, the most this will do is impact research in a few forms of the disease.
Immortality would get tedious after a while. What you really want is a method to transcribe the contents of the brain plus the original genome of the body, altered to include a flesh-eating component that is normally inactive. When the body inevitably wears out, you make a few adjustments to the genome to prevent that cause of death killing you again. You then make the stem cell "carnivorous", using the raw material of your old body to create a new one, re-inserting "you" into the new brain in the process.
I call this technique "regeneration" and think that, in the interests of population control, people should be limited to 12 of them.
On the other hand, Shakyamuni Buddha lived into his 80s, Jean Manual Fangio only gave up professional motor racing in his 90s, and the Queen Mother was conducting human experiments on the effects of gin past the century mark.
I doubt they'll be unhappy. The wider area means a larger virtual dish and therefore higher accuracy for the scientists. The higher total cost means that the vendors stand to make more money from this, not less. The distribution over two continents means you've twice as many science budgets (so lower cost to each individual tax payer).
Since density is mass per unit volume, the only two ways to increase density are to either replace the atoms in the models' bodies with heavier, highly unstable ones, or push them into a compacter. The first might cause problems due to the cameras suffering radiation damage. The latter would cut into the celebrities they have at those adult expos.
That only helps in Britain, as prescription charges are fixed. In the US, all the insurance companies need to say is that data storage requirements are a pre-existing condition.
So you'd be happy if all the authors, musicians, actors and scriptwriters quit? They CAN choose other careers, right? What these other careers are, given the massive unemployment, is something you don't seem keen to delve into.
I choose to live in a world where there is new art. The day we lose art is the day the world is no longer worth living in.
The labels, the brokers, the headhunters - those are expendable and the sooner we're rid of them the better. They produce nothing, they create nothing, they do nothing except sponge money out of the system and into the jets and flash cars. Artists, for the most part, are on the edge of starvation. More than a few work when homeless and malnourished. If you were in those conditions, you would rightly object. But you find it fine for the labels to treat their artists this way. Interesting. Have you seen a doctor about this?
The scribes didn't create anything, though. They merely copied. Here, the people being put out of work are NOT the scribes (the record labels), but the artists who actually produce the stuff. The scribes are making MORE profit than they did in the old system.
Nobody is saying the old way of doing things was good - it's almost universally accepted that the labels would gouge the artists of every penny they could (and more).
What IS being said here is that the new method is NOT helping the artist and is really not helping anyone else either (since it takes money to make art).
The solution is NOT to help the 99.9% (you're still helping the 1%, the record labels, via this new approach). It is better to stick with supporting the artists AND the consumers. If this means that it will be at the expense of the labels, so be it, but it is the labels that should suffer, NOT the artists.
I've worked with X.25 myself. It's complicated, yes, but if you look at the rapidly increasing complexity of modern networks, X.25 is not as evil as it might first appear.
The problem with formal specifications is that they have to handle future cases virtually seamlessly as formal approaches are not really amenable to gradual improvements and refinements. This leads to a specification either being too much of a special case to be generally useful (so you need lots of specs) or so general that it covers those future needs that can be realistically anticipated (which gives you one very complicated specification).
X.25 seems to have gone for the latter, which in many respects is a sensible approach. If you were to say they missed the mark in some area, I'd agree. If you were to say that TCP/IP is better because you've loads of specs and myriads of specialist layer 4 protocols, I'd say that it depends on your definition of "better" but would agree that it has a lot of advantages.
Hundreds of parameters, many esoteric - true, very true, but there's no shortage of socket options and other tweakables for TCP/IP. The Web100 (now Web10G) project exists purely because it isn't trivial to set them correctly. Linux also supports a vast number of TCP schemes and traffic engineering functions in order to allow you to do further manipulation of windowing, buffering and traffic shaping.
If you could imagine every tweakable on a single control panel, it would be no better - and maybe a lot worse - than X.25. They aren't on a single control panel, they're in a multitude of different APIs, where no one API is that heavy. In other words, X.25 is probably lighter on the total number of parameters but TCP/IP in general (and Linux' implementation in particular) encapsulates these in a vastly more manageable way, uses more defaults and uses better defaults.
If X.25 were revised today (not that anyone is going to), it would likely be segmented rather than trying to do everything in one place and would end up very similar to modern TCP/IP. Which, of course, is why nobody would revise X.25 - TCP/IP already exists so there's no need for what would be an incompatible substitute. It's why Infiniband is unlike either (but learns some from TCP/IP on proper segmenting).
Although I disagree with your sentiment on the UN, I have no problem with agreeing with the idea of a wholly independent political entity (as opposed to the quasi-independent one I suggested). There's no serious question that it needs a high level of independence and that it needs to be impervious to national or corporate pressures to manipulate things.
My argument for quasi-independence is purely that it would be very hard to get any nation to sign up to such a body - the US especially and that's the one that's absolutely needed because it has near-totalitarian control right now. On the other hand, a UN Charter would not require all nations to agree. I think you need 2/3rds before something can become International Law, at which point non-signatories are out of luck. Their disagreement simply doesn't matter at that point.
2/3rds of all nations is still a hell of a lot, it might well not be achievable, but it's a much smaller number than 100% of both nations and international ISPs combined and it'd be easier to persuade nations that currently have little-to-no say in things that no other nation should have a monopoly than to persuade the ISPs (who have control over their own links) that they should relinquish some of that control.
However, that's just the means to the end. The end itself is for a group that is outside of the political power games that go on and independent of those who play them. Since we both seem to agree on the end, we can leave how to get there for some other time.
Ok, agreed TCP/IP uses four and agreed that those four don't entirely agree with the OSI model (you have to talk of fractions of a level to line things up, bleagh!).
Applications increasingly add more on top, and yes technically you are correct that from a purist network perspective they aren't important. However, there's generally 2-3 of them and that matches the OSI prediction on how many application-level layers are needed, so I'm going to say that the OSI model got the ratio of total level of functionality needed to total complexity affordable just about right even if experience tells us that the extra layers needed aren't really network at all. If you therefore re-state OSI as a 4+3 layer model (network+application-level), rather than as a 7 layer model (all network) it's an impressive fit for something that is absolutely ancient in computing terms.
So, yes, in the technical sense you are absolutely correct. It is the basis for teaching about networks, rather than the reality of networks. No question about it. It is in the totality of modern applications and how they work -in the most abstract sense- that you find that a 4+3 design nicely balances flexibility with overhead.
(On this basis, I would argue that those teaching OSI should not go on to say that it's not used in practice, but should actually say that the OSI model correctly describes network activities but incorrectly lumps all the complexity and functionality together into a single box, where it's the box that doesn't exist in practice.)
The ITU came up with a hell of a lot more than those. X.25 was the wire protocol used by Europe for a very long time - worked extremely well and was highly robust, compared to its US contemporary which was IPv0.
X.400 was probably heavier than necessary, but 99% of all work to improve on the limitations of SMTP have basically been reinventions of features X.400 had from the start.
X.500 exists today in the form of LDAP + ASN.1 + Digital Certificates + Federated methods of authentication. All these combined still don't cover the full spectrum of X.500 capabilities, but most of what's left wasn't really needed. However, there's nothing done today that wasn't in the standard. Not bad going.
Their other work includes little-known standards like JPEG, MPEG-2, MPEG-4, H.323 and ISDN, along with developing and standardizing technology that you can't possibly have heard of like wavelength-division multiplexing for optic fibre and DSL (yes, it's an ITU product as well).
Yes, they did the OSI model (which is still the basis for most networking) and SDL, but nobody's perfect.
The US did not create the Internet. DARPA (quasi-independent) created the protocols, the NSF (quasi-independent) installed the US infrastructure and research institutions in other countries (wholly independent) installed the rest. The US (as in the elected government, or the corporate entities within it) did absolutely bugger all apart from make life difficult.
I didn't say equal in free speech, I said equal in MEANINGFUL free speech. The freedom to speak in the US is absolute provided it can't actually achieve anything. The total amount of influence you, I, or any other individual has is no greater than that of any North Korean. We can do nothing that has any substance, any real material worth, we can only play-act that what we say will do any good. Being able to uselessly scream in the dark might well make you feel better and make you feel like you've done something worthwhile, but you haven't.
Look at the "Occupy" protests. They were massive, they were long-lasting, they were crushed through excessive violence and the political leaders responsible for that crushing have seen their poll numbers RISE as a result of using terror and intimidation on those who were asking only that the 1% pay fair dues. How's a protester getting his head smashed in and left in a vegetative state in Oakland that different from a protester getting his head blown off by a North Korean guard? Both are functionally dead, neither has done any good and neither had any free speech worth a damn.
Sure, the US protestor was completely free to set up a website and colour it pink to protest, because that would have done bugger all. Sure, you can march in protest all day in the middle of Death Valley, where nobody can see or hear you, but protestors in Seattle for the WTO talks there got blasted with tear gas, water canons and probably shot at with the occasional live round. Speech where it matters isn't free in the slightest.
Joe the Plumber only had any impact at all in the US precisely because he wasn't Joe and wasn't a plumber. If he had been, he would have been ignored utterly. It's because he was well-to-do and had connections that he was groomed, positioned and then hyped. How's that different from the way the Chinese government works?
The US control over territories it illegally seized from Mexico is not legally or morally distinct from the Chinese illegally seizing Tibet. Both were criminal acts at gunpoint, both have been supplemented by attempting to replace the indigenous population with imports together with destruction of indigenous traditions and attempts to eliminate the indigenous languages. Both have been justified by the occupying power as having civilized barbaric regions.
To me, that says that perspective isn't on the side of the US on this one. Once a criminal, always a criminal.
Well, no, the US didn't control the Internet when it was any good, the NSF did. (No, the DARPA days weren't better, DARPA's screwball decisions are why the Internet protocols are as messed up as they are.) The NSF isn't run by Congress or Corporations. One option would be for the NSF to claim eminent domain, seize all fibre (lit and dark) in the US or owned by US organizations, and run the lot on rational principles.
However, that would only cover the US. The Internet is very global. Even CERN is primarily European. We need a UN body for the global reach, but it would need to meet the following criteria to actually work:
* It needs to be quasi-independent * Members should be elected purely on merit, not on grounds of money or territory covered * Officials should be 75% from the academic community and 25% from the InfoSec community, NOTHING from the political or corporate communities * The organization should be primarily concerned with research, collaborative projects and the information demands of science * The Internet should be a means to achieve the desired end results, not an end in itself * Since this limits direct law-enforcement options, it would need to have significant muscle (eg: veto powers in the IMF and WTO) to ensure nations complied
However, let's assume the GA wants to take over and not create a meaningful NSF-like body. Actual gangsters and dictators hold onto power because they know what they can take and when not to push too hard. The KKK was well-known for charity work, not because they gave a crap but because it's by far the easiest way to manipulate the hearts and minds of those peons and fools they needed to be compliant. Corporations hold onto power through smoke, mirrors and legislation. They take it all and don't give a crap about pushing too hard because customers are expendable. I have zero faith in the mob, but that's still far more faith than I'll ever have in a megacorp.
I'd also point to Japan where actual mobsters and criminal gangs ARE in charge of many areas of law enforcement -- the nation has better Internet than the US (eg: gigabit to the home), better medical care, lower levels of (unlicensed) crime, lower levels of overt violence and far better sushi. It's an actual real-life embodiment of Terry Pratchett's Thieves' Guild. (I would not be surprised if Terry Pratchett got the idea from them, since many of his books are sourced in real-world ideas.)
That's far from ideal, and I repeat I have zero faith in it, but my faith in the current system is so far in the negative that zero is a definite improvement.
They're not providing you with a lower speed just to be dicks.
Actually, I'm confident that that IS the reason, since their line speed varies so much. The ISP has probably installed the cheapest DSL equipment in the teleco that they can get away with and still claim to provide a service, along with the lowest-rated line onto the backbone.
It's the traditional way to run a profit-oriented business -- supply the least and get the most. ISPs are not, as a rule, out there to serve you, they are out there to serve their bank account. They don't care about losing customers - there'll be another sucker along soon and once they run out of suckers they'll change their name or advertise they're upgrading. The fools they sponge off will try again, and this can be kept going for a very long time. AT&T and Microsoft have built huge empires out of this practice. Facebook is about to be given billions for doing the same. Enron so very nearly got away with what is truly grand theft but the market burped.
The only reason the free market exists is because consumers have figured out that being robbed blind and left in the gutter is better than the previous approach to markets which was to rape you, beat you within an inch of your life, THEN rob you blind and leave you in the gutter. (Survivalists, on the other hand, maim and torture themselves before robbing themselves blind and throwing themselves in the gutter.)
As crazy as it is, it's infinitely more believable than US' Congress saying that (and the Supreme Court has said the FCC can't regulate the Internet unless Congress DOES pass the law you describe).
The US is about equal to all those you list in terms of civil rights, meaningful free speech (not just the playstuff that's actually allowed), levels of corruption and levels of actual democracy.
I agree that none of those listed should have a voice, but by the same standard neither should the US. At present, the US has very near absolute power. The GA may have depraved and corrupt elements, but on aggregate it's no worse than the US on any metric and at times is a whole lot better.
Ideally, the Internet would be run by a meritocratic UN group, with all nations recognizing and respecting a group that chooses members by merit and acts on merit. There have been *cough* enough incidents where nations (US included) have actively sought to cripple meritocratic groups that I do not believe such a group could function. It would lack the teeth necessary to impose its decisions and to work it would need Predator X-like teeth.
Slashdot Mirror
I am Rassilon! Besides, if prior art isn't a problem for patents.... :)
I'll agree that certain distros are highly insecure, but equally there are hardened distros that will run essentially the same software. Obviously, software that violates security protocols won't run under a hardened distro (if the distro is any good, that is!) so it's not 100% the same. The lead between Windows and Linux should really be measured from "useful best to useful best" rather than "OTS to OTS" (since nobody runs OTS in practice, all systems are tweaked in some fashion) or "worst to worst" (since all OS' in their worst configuration have no security at all). I would dispute the idea that the best Windows configs are orders of magnitude closer to the best Linux configs, in their hardest configurations - GRSecurity + RBACS + Linux Capabilites + Netfilter + L7 Routing is still a very tough combination to beat in terms of the level of granularity of control. It certainly beats Windows' permissions families plus Windows' firewall in terms of what you can do and what you can restrict.
Agreed that Linux is not a magic security bullet. I wouldn't agree that no OS is unbreakable -- no -useful- OS is unbreakable, yes, but it's quite possible to make a useless OS that's unbreakable. Although, as Microsoft has found, it's also possible to make a useless OS that's very breakable indeed.
Put a hardened Windows system and a hardened Linux system in the same room, then run just standard vulnerability scanners over them. I am willing to bet cold hard cash that the Windows system will be flagged as having potential problems and that the Linux system will not.
(How do I define "hardened"? Microsoft has a free tool that tells you if the system is running in a hardened configuration. For Linux, start with a security distro that includes GRSecurity and/or RBACS and essentially follow the same procedures as advised by Microsoft. If you want specific instructions, the NSA have a guide on how to harden Linux - and indeed how to harden Windows. To be fair, you should apply both.)
Define "business applications" in your context. For many years, "Exchange" was considered a "business application" that "could not run under Linux" - which, whilst technically true, was also technically false since oGo and other near-100% Exchange clones have been around for a long time. SQL Server is only marginally acceptable, as tools like "Dezign for Databases" will transliterate SQL Server databases into other database syntaxes and there's really very little that SQL Server can do that cannot be reproduced by any means whatsoever on Ingres, Oracle, DB/2, Informix or PostgreSQL - all of which run perfectly well under Linux. (You only need one server that supports what you need to do, so it is an "or" not an "and".)
For the most part, that leaves web-based applications (which, by definition, will run on any desktop) and "office productivity tools" (aka "office maiming tools"), most of which either have clones or will run directly. In the case of MS Office, both. (Yes, Office can be run under Linux.)
Because anything that runs under OS/X will be IBCS-compliant, anything that runs under OS/X can be made to run natively (not emulated, NATIVELY) under Linux, although that would require that IBCS support be reintroduced.
I thought it was cowbell.
There are many, many genes involved in different forms of cancer, the most this will do is impact research in a few forms of the disease.
Immortality would get tedious after a while. What you really want is a method to transcribe the contents of the brain plus the original genome of the body, altered to include a flesh-eating component that is normally inactive. When the body inevitably wears out, you make a few adjustments to the genome to prevent that cause of death killing you again. You then make the stem cell "carnivorous", using the raw material of your old body to create a new one, re-inserting "you" into the new brain in the process.
I call this technique "regeneration" and think that, in the interests of population control, people should be limited to 12 of them.
On the other hand, Shakyamuni Buddha lived into his 80s, Jean Manual Fangio only gave up professional motor racing in his 90s, and the Queen Mother was conducting human experiments on the effects of gin past the century mark.
Yeah, but that's how you can regulate anything.
Agreed. Who has the patent on Soylent Green?
Politics is involved. That makes uncertainty infinite. Triple infinity is still infinity.
I doubt they'll be unhappy. The wider area means a larger virtual dish and therefore higher accuracy for the scientists. The higher total cost means that the vendors stand to make more money from this, not less. The distribution over two continents means you've twice as many science budgets (so lower cost to each individual tax payer).
Yes, but the platters are round, not square.
Since density is mass per unit volume, the only two ways to increase density are to either replace the atoms in the models' bodies with heavier, highly unstable ones, or push them into a compacter. The first might cause problems due to the cameras suffering radiation damage. The latter would cut into the celebrities they have at those adult expos.
That only helps in Britain, as prescription charges are fixed. In the US, all the insurance companies need to say is that data storage requirements are a pre-existing condition.
So you'd be happy if all the authors, musicians, actors and scriptwriters quit? They CAN choose other careers, right? What these other careers are, given the massive unemployment, is something you don't seem keen to delve into.
I choose to live in a world where there is new art. The day we lose art is the day the world is no longer worth living in.
The labels, the brokers, the headhunters - those are expendable and the sooner we're rid of them the better. They produce nothing, they create nothing, they do nothing except sponge money out of the system and into the jets and flash cars. Artists, for the most part, are on the edge of starvation. More than a few work when homeless and malnourished. If you were in those conditions, you would rightly object. But you find it fine for the labels to treat their artists this way. Interesting. Have you seen a doctor about this?
The scribes didn't create anything, though. They merely copied. Here, the people being put out of work are NOT the scribes (the record labels), but the artists who actually produce the stuff. The scribes are making MORE profit than they did in the old system.
Nobody is saying the old way of doing things was good - it's almost universally accepted that the labels would gouge the artists of every penny they could (and more).
What IS being said here is that the new method is NOT helping the artist and is really not helping anyone else either (since it takes money to make art).
The solution is NOT to help the 99.9% (you're still helping the 1%, the record labels, via this new approach). It is better to stick with supporting the artists AND the consumers. If this means that it will be at the expense of the labels, so be it, but it is the labels that should suffer, NOT the artists.
I've worked with X.25 myself. It's complicated, yes, but if you look at the rapidly increasing complexity of modern networks, X.25 is not as evil as it might first appear.
The problem with formal specifications is that they have to handle future cases virtually seamlessly as formal approaches are not really amenable to gradual improvements and refinements. This leads to a specification either being too much of a special case to be generally useful (so you need lots of specs) or so general that it covers those future needs that can be realistically anticipated (which gives you one very complicated specification).
X.25 seems to have gone for the latter, which in many respects is a sensible approach. If you were to say they missed the mark in some area, I'd agree. If you were to say that TCP/IP is better because you've loads of specs and myriads of specialist layer 4 protocols, I'd say that it depends on your definition of "better" but would agree that it has a lot of advantages.
Hundreds of parameters, many esoteric - true, very true, but there's no shortage of socket options and other tweakables for TCP/IP. The Web100 (now Web10G) project exists purely because it isn't trivial to set them correctly. Linux also supports a vast number of TCP schemes and traffic engineering functions in order to allow you to do further manipulation of windowing, buffering and traffic shaping.
If you could imagine every tweakable on a single control panel, it would be no better - and maybe a lot worse - than X.25. They aren't on a single control panel, they're in a multitude of different APIs, where no one API is that heavy. In other words, X.25 is probably lighter on the total number of parameters but TCP/IP in general (and Linux' implementation in particular) encapsulates these in a vastly more manageable way, uses more defaults and uses better defaults.
If X.25 were revised today (not that anyone is going to), it would likely be segmented rather than trying to do everything in one place and would end up very similar to modern TCP/IP. Which, of course, is why nobody would revise X.25 - TCP/IP already exists so there's no need for what would be an incompatible substitute. It's why Infiniband is unlike either (but learns some from TCP/IP on proper segmenting).
Although I disagree with your sentiment on the UN, I have no problem with agreeing with the idea of a wholly independent political entity (as opposed to the quasi-independent one I suggested). There's no serious question that it needs a high level of independence and that it needs to be impervious to national or corporate pressures to manipulate things.
My argument for quasi-independence is purely that it would be very hard to get any nation to sign up to such a body - the US especially and that's the one that's absolutely needed because it has near-totalitarian control right now. On the other hand, a UN Charter would not require all nations to agree. I think you need 2/3rds before something can become International Law, at which point non-signatories are out of luck. Their disagreement simply doesn't matter at that point.
2/3rds of all nations is still a hell of a lot, it might well not be achievable, but it's a much smaller number than 100% of both nations and international ISPs combined and it'd be easier to persuade nations that currently have little-to-no say in things that no other nation should have a monopoly than to persuade the ISPs (who have control over their own links) that they should relinquish some of that control.
However, that's just the means to the end. The end itself is for a group that is outside of the political power games that go on and independent of those who play them. Since we both seem to agree on the end, we can leave how to get there for some other time.
Ok, agreed TCP/IP uses four and agreed that those four don't entirely agree with the OSI model (you have to talk of fractions of a level to line things up, bleagh!).
Applications increasingly add more on top, and yes technically you are correct that from a purist network perspective they aren't important. However, there's generally 2-3 of them and that matches the OSI prediction on how many application-level layers are needed, so I'm going to say that the OSI model got the ratio of total level of functionality needed to total complexity affordable just about right even if experience tells us that the extra layers needed aren't really network at all. If you therefore re-state OSI as a 4+3 layer model (network+application-level), rather than as a 7 layer model (all network) it's an impressive fit for something that is absolutely ancient in computing terms.
So, yes, in the technical sense you are absolutely correct. It is the basis for teaching about networks, rather than the reality of networks. No question about it. It is in the totality of modern applications and how they work -in the most abstract sense- that you find that a 4+3 design nicely balances flexibility with overhead.
(On this basis, I would argue that those teaching OSI should not go on to say that it's not used in practice, but should actually say that the OSI model correctly describes network activities but incorrectly lumps all the complexity and functionality together into a single box, where it's the box that doesn't exist in practice.)
The ITU came up with a hell of a lot more than those. X.25 was the wire protocol used by Europe for a very long time - worked extremely well and was highly robust, compared to its US contemporary which was IPv0.
X.400 was probably heavier than necessary, but 99% of all work to improve on the limitations of SMTP have basically been reinventions of features X.400 had from the start.
X.500 exists today in the form of LDAP + ASN.1 + Digital Certificates + Federated methods of authentication. All these combined still don't cover the full spectrum of X.500 capabilities, but most of what's left wasn't really needed. However, there's nothing done today that wasn't in the standard. Not bad going.
Their other work includes little-known standards like JPEG, MPEG-2, MPEG-4, H.323 and ISDN, along with developing and standardizing technology that you can't possibly have heard of like wavelength-division multiplexing for optic fibre and DSL (yes, it's an ITU product as well).
Yes, they did the OSI model (which is still the basis for most networking) and SDL, but nobody's perfect.
The US did not create the Internet. DARPA (quasi-independent) created the protocols, the NSF (quasi-independent) installed the US infrastructure and research institutions in other countries (wholly independent) installed the rest. The US (as in the elected government, or the corporate entities within it) did absolutely bugger all apart from make life difficult.
I didn't say equal in free speech, I said equal in MEANINGFUL free speech. The freedom to speak in the US is absolute provided it can't actually achieve anything. The total amount of influence you, I, or any other individual has is no greater than that of any North Korean. We can do nothing that has any substance, any real material worth, we can only play-act that what we say will do any good. Being able to uselessly scream in the dark might well make you feel better and make you feel like you've done something worthwhile, but you haven't.
Look at the "Occupy" protests. They were massive, they were long-lasting, they were crushed through excessive violence and the political leaders responsible for that crushing have seen their poll numbers RISE as a result of using terror and intimidation on those who were asking only that the 1% pay fair dues. How's a protester getting his head smashed in and left in a vegetative state in Oakland that different from a protester getting his head blown off by a North Korean guard? Both are functionally dead, neither has done any good and neither had any free speech worth a damn.
Sure, the US protestor was completely free to set up a website and colour it pink to protest, because that would have done bugger all. Sure, you can march in protest all day in the middle of Death Valley, where nobody can see or hear you, but protestors in Seattle for the WTO talks there got blasted with tear gas, water canons and probably shot at with the occasional live round. Speech where it matters isn't free in the slightest.
Joe the Plumber only had any impact at all in the US precisely because he wasn't Joe and wasn't a plumber. If he had been, he would have been ignored utterly. It's because he was well-to-do and had connections that he was groomed, positioned and then hyped. How's that different from the way the Chinese government works?
The US control over territories it illegally seized from Mexico is not legally or morally distinct from the Chinese illegally seizing Tibet. Both were criminal acts at gunpoint, both have been supplemented by attempting to replace the indigenous population with imports together with destruction of indigenous traditions and attempts to eliminate the indigenous languages. Both have been justified by the occupying power as having civilized barbaric regions.
To me, that says that perspective isn't on the side of the US on this one. Once a criminal, always a criminal.
Well, no, the US didn't control the Internet when it was any good, the NSF did. (No, the DARPA days weren't better, DARPA's screwball decisions are why the Internet protocols are as messed up as they are.) The NSF isn't run by Congress or Corporations. One option would be for the NSF to claim eminent domain, seize all fibre (lit and dark) in the US or owned by US organizations, and run the lot on rational principles.
However, that would only cover the US. The Internet is very global. Even CERN is primarily European. We need a UN body for the global reach, but it would need to meet the following criteria to actually work:
* It needs to be quasi-independent
* Members should be elected purely on merit, not on grounds of money or territory covered
* Officials should be 75% from the academic community and 25% from the InfoSec community, NOTHING from the political or corporate communities
* The organization should be primarily concerned with research, collaborative projects and the information demands of science
* The Internet should be a means to achieve the desired end results, not an end in itself
* Since this limits direct law-enforcement options, it would need to have significant muscle (eg: veto powers in the IMF and WTO) to ensure nations complied
However, let's assume the GA wants to take over and not create a meaningful NSF-like body. Actual gangsters and dictators hold onto power because they know what they can take and when not to push too hard. The KKK was well-known for charity work, not because they gave a crap but because it's by far the easiest way to manipulate the hearts and minds of those peons and fools they needed to be compliant. Corporations hold onto power through smoke, mirrors and legislation. They take it all and don't give a crap about pushing too hard because customers are expendable. I have zero faith in the mob, but that's still far more faith than I'll ever have in a megacorp.
I'd also point to Japan where actual mobsters and criminal gangs ARE in charge of many areas of law enforcement -- the nation has better Internet than the US (eg: gigabit to the home), better medical care, lower levels of (unlicensed) crime, lower levels of overt violence and far better sushi. It's an actual real-life embodiment of Terry Pratchett's Thieves' Guild. (I would not be surprised if Terry Pratchett got the idea from them, since many of his books are sourced in real-world ideas.)
That's far from ideal, and I repeat I have zero faith in it, but my faith in the current system is so far in the negative that zero is a definite improvement.
Actually, I'm confident that that IS the reason, since their line speed varies so much. The ISP has probably installed the cheapest DSL equipment in the teleco that they can get away with and still claim to provide a service, along with the lowest-rated line onto the backbone.
It's the traditional way to run a profit-oriented business -- supply the least and get the most. ISPs are not, as a rule, out there to serve you, they are out there to serve their bank account. They don't care about losing customers - there'll be another sucker along soon and once they run out of suckers they'll change their name or advertise they're upgrading. The fools they sponge off will try again, and this can be kept going for a very long time. AT&T and Microsoft have built huge empires out of this practice. Facebook is about to be given billions for doing the same. Enron so very nearly got away with what is truly grand theft but the market burped.
The only reason the free market exists is because consumers have figured out that being robbed blind and left in the gutter is better than the previous approach to markets which was to rape you, beat you within an inch of your life, THEN rob you blind and leave you in the gutter. (Survivalists, on the other hand, maim and torture themselves before robbing themselves blind and throwing themselves in the gutter.)
As crazy as it is, it's infinitely more believable than US' Congress saying that (and the Supreme Court has said the FCC can't regulate the Internet unless Congress DOES pass the law you describe).
The US is about equal to all those you list in terms of civil rights, meaningful free speech (not just the playstuff that's actually allowed), levels of corruption and levels of actual democracy.
I agree that none of those listed should have a voice, but by the same standard neither should the US. At present, the US has very near absolute power. The GA may have depraved and corrupt elements, but on aggregate it's no worse than the US on any metric and at times is a whole lot better.
Ideally, the Internet would be run by a meritocratic UN group, with all nations recognizing and respecting a group that chooses members by merit and acts on merit. There have been *cough* enough incidents where nations (US included) have actively sought to cripple meritocratic groups that I do not believe such a group could function. It would lack the teeth necessary to impose its decisions and to work it would need Predator X-like teeth.