Which is more important for individuals? Which is bigger in terms of national economics, basic goods or communication infrastructure?
So if you are against free markets as in not-regulated and not politically-controlled markets, why don't you require, that gocery stores everywhere should provide price and quality neutral access to all goods for all good men?
We Europeans are supposed to be more leftist than U.S.A., but we just don't get the fuss over "net neutrality". We don't have "universal access" or free-as-in-communims local calls either, and we could not care less.
Did you know for example, that he says fancy cooling is only necessary, if your components are in a box. Wrap them to racks with velcro, drop the boxes, and you need no active cooling.
FRM seems like an interesting research topic, but while addressing the limitations of now fashionable Single Source Multicast (SSM), it is an overkill for P2P. P2P could do fine with very simple SSM.
But in the end I doubt the feasibility of IP (layer 3) multicast. SSM solves the multicast routing problem, and source/group discovery and advertising can now be easily done manually, but other multicast problems remain: synchronicity (everyboday has to recieve at the same time), least common denominator bandwith requirement (everybody has to recieve at the same bandwith), inflexibility of UDP traffic (does not adapt to network congestion like TCP, which is scaring the ISPs) and above all, the need to have unbroken chain of supporting routers in order to work at all.
P2P already is application level multicast. It can work without support from the routers, with flexible ISP friendly TCP, and with no synchronized bandwith or time synchronization of clients. Besides, application level allows all kinds of innovative optimizations like compression, caching, security, anonymity, billing or whatever. And multicast for video streaming really might require different approach than multicast for file delivery, it is not a good idea to force them to use the same low level mechanism.
Besides, market fragmentation of P2P protocols is not such a bad thing, because f.ex. all my stuff is there in my iTunes or video directories, and I share it on all my active P2P networks anyway.
First find a good distro spesific IRC channel. (Learn IRC if you are an old fart like me who originally hated this kid stuff.) Whatever you have in your mind it is always a good idea to utter something to IRC. For example the Ubuntu default IRC client automatically goes to Ubuntu channel when you open it, and so does bitchx IRC client on Debian.
WHILE waiting for answers/flames read man pages/docs or google, and also the Google Groups. How to balance you problem solving between "official" docs of the SW at hand and more free googling? It depends on the case. If you have an error message, definitely do google with it, with and w/o quotes. AT THE SAME TIME, when you learn anything new with man/docs/google, you can augment or rephrase your question on IRC. On the other hand, if someone answers on IRC, pick and feed better keywords for further googling. In the end, even if nobody answered on IRC, it is an altrustic idea to write your final solution there, because nobody seemed to know it. And they can maybe verify it.
Only when all else fails, ask on a distro specific forum. That is so much slower and rarely gets you any better results.
In difficult application spesific cases you of course have to move the issue to application spesific IRC channels or forums, but it is always a better idea to start on distro channels and forums, because you might be using the sub-optimal application in your environment in the first place. And the help in distro channels/forums is more accessible, especially for newbies. And everybody is a newbie when starting with a new application. And the answers might be distro specific.
Alko known as "movies". Can be distributed on Steam. Or an a Web page. Two fifths of Episode One lenght, two fifths of the price.
Or how about something in between? A movie where you can interact a little here and there, but you don't have to, if you are lazy. Or there can be a plot that is fixed or has a limited set of variations, and action is done by AI NPCs with random variation on each playtime, where a random result of some action scene chooses a plot variation.
You know, me and my 3 year old son love to watch the Monaco Grand Prix in TV mode in GP4. We usually do that when he goes to bed so that he falls a asleep though, so it is not SO exiting. But it is as good as the real one, except that you can control the camera IF you want;-)
Seriously, could this become a new art movement? Unauthorized, pirated "remixes" of movies, with dull parts edited out, added dialog to clear up confusion/contradictions in plot, better effects, new scenes without the original characters, or in the extremely ambitious case, new scenes with stunts of the characters? (Or aunauthorized appearances of bribable original actors;-)
In the 80's they still said that everybody is buying IBM equipment because nobody was ever fired for that. I guess these days "nobody is ever fired for building a Windows based system".
The article talks about backup. The idea could be, that instead of managing incremental backups you just optimize compression of data that is similar to old data. In that way you can do "full" backups, but actually save only incremental backup worth of data.
See http://en.wikipedia.org/wiki/Venti for similar ideas in a system that easily achives 25x compression for typical archival storage. When a file has been changed only those 512 kbyte blocks that are really new are saved, other blocks are just mapped by their SHA1 hashes to existing blocks. So files with small changes, very similar files and files sharing common parts will all compress very nicely. In a multi-user system the files of different users tend to also have lots of similar parts: same emails, same office documents with perhaps minor changes, same reference material / tools / libraries as personal copies etc.
My guess is TFA refers to a re-invention of this wheel, most likely in an inferior way.
If you have a single application device, it really does not need a comprehensive, separate "OS". If and when you need process management and memory management and sharing of I/O resources between modules/processes, you need SOME libraries/headers to do that. If you want to have layers of abstraction (say, high level I/O or media codecs) to help in development, you need also some libraries for that.
But up to a point you are free to use per application designed common libraries and run-time systems, you don't need to have a general purpose interface/protocol between "applications" and the "operating system". This is what many light-weight embedded toolkits are: just run-time systems and shared libraries with shared data structures. They are far simpler and smaller than any version of Linux, and more tailored to typical needs here.
But if you want to have a user extendable system and an ecosystem of independent code modules and libraries all running smoothly on a variety of hw versions, with standard programming interfaces, then Linux probably is a reasonable choice. (Though I would go with Plan 9 or Inferno.) This kind of system needs more expensive hw and is more complex, but it is also kind of easier to build, because it is more "common" and more "standard". With the cost of hw falling, this type of system is going to be more and more the most cost effective way to engineer the life cycle of several product generations, when design time and itellectual property costs are bigger than costs of general purpose processors, DSPs and memory.
We use Skype as the official company VoIP and conferencing sw. We are a small company, and we have a persistent Skype chat open with all our globally travelling workers, and we cut global phone costs with Skype. Internal meetings are often done is Skype. The two good things are that it works through many kinds of firewalls, and that it encrypts.
I didn't bother to read the link you provided, because as an IPv6 trainer in our company I already know:
IPv6 is not needed, NAT works. Some people will even insists NATing their IPv6 network to protect the internal addresses.
You don't need IPv6 to have global reachability for VoIP and P2P. Teredo actually proves, that you can contact anyone with a private address if you really want. You just need some form of global addressing, and an active "NAT circumvention" server like Teredo. Skype and other P2P networks that work with private addresses also prove, that it is easy to contact anyone globally, know matter how private addressing they use.
Transition to IPv6 is huge work. Upgrading all software is huge work. Running two parallel network infrastructures is huge work. Nobody wants to do that.
Re-checking your security infrastructure (firewalls, IDS, whatever) for IPv6, and for half a dozen clever tunneling tricks is something your security guys will NEVER do. And thus, they will never allow IPv6.
Mobile IP (MIP) would be nice and the only real reason to insist everybody having a public IPv6 address. But after Skype and P2P networks have been made to work on private addresses, you don't really need MIP so much. The cost of having IPv6 just for that is too much.
But the current problem is just that no one can agree on a common hierarchy, and never will. And because the dispute is global, it is difficult, and not very well handled by a monopoly of any sort, not even the U.N. (God forbid not!).
Besides, nobody actually reads or expects to read any info from relation of the sub-domain names in "namex.namey.namez".
On the other way, in my system you would be free to set up any "logical" hierarchy or hierachies you wish, and anybody could register there, if you make it your business service to provide registration to your neat and tidy logical branch of the name tree.
And one thing about squatting disputes: note that I could try to squat "coke." ("c.o.k.e."), due to possible legal challenges, but the company could just use coca.cola or coke.drink or coke.ltd or coke.adds.life or whatever. Squatting is only meaningful while the current TLDs remain a scarce, monopoly controlled resource.
Basic squats ("coke."/"c.o.k.e.") would anyway be solved exactly like now: anything can be registered, but if someone later can show a legitimate interest in the domain and if your interest is definitively squatting only, you loose it to the better owner for free. It is not a big problem today, and it would be less a problem in my system.
This domain squatting problem would be no worse than today. And squatting a single mid-level letter would give you no benefit, just costs, because you would be required to allow (free) delegation of any sub-letter domains.
This would let everyone run their own pseudo "top level domain" without stressing the global system too much. The goal here is especially that the DNS cache hit ratios would not drop dramatically because of multitude of global domains.
I once tried to discuss this on the DNSO mailing list, but finally could not find a right place:
Let's first set up top level domains for all single letters a., b.,...z. Then if I want to have a name "anssi.abc." for my machine. I rewrite it (first in my head) to be "anssi.a.b.c." I go to the "c." TLD authority. I ask them, if they have already delegated the 2nd level domain "b.c.". If such domain exists, I go to its authority, and ask, if they have a domain "a.b.c." already.
If the domain "b.c." did exist, but the domain "a.b.c." did not, then I register my "a.b.c." servers with the domain "b.c." authority, to set up new delegation. There should be a standard IETF rule, that single letter domains must allow delegation of any unregistered single letter subdomains!
So now I can set up names like "anssi.a.b.c." in my own name server, serving the domain "a.b.c.", registered at "b.c.". NOW I MODIFY MY RESOLVER and ask everybody to do the same, if they please! The new versions of resolvers would be standardized to TRANSLATE ALL UNRECOGNISED multi-letter TLDs into a sequence of one letter domains before the query. So "anssi.abc." will be resolved as "anssi.a.b.c.".
The classical, now existing TLDs can be handled as a special case. But now anybody is also free to set up whatever TLDs they want to!!! You greatly benefit from an updated resolver, but all resolvers will still work with the old TLDs, and old resolvers will also kind of work with the new names, but only if you translate manually, writing "anssi.a.b.c" instead of "anssi.abc".
You don't have to touch applications, but of course a Web browser might be aware of the new system, and compensate for non-upgraded resolvers. Anyway, writing HTML anchors like "anssi.a.b.c" would work also in old resolvers.
Otto and Clockwurk are right, but Linux is still "good" for many things, and getting better faster than Windows is.
Otto's anyway has two very good main points:
Point 1:
Linux has not reached the "it 'just works' for everyday people on everyday needs" state. Today, n people in the world are capable of doing m everyday things on Windows (or a Mac) without trying very hard. For Linux the n x m is just orders of magnitude less than for Linux. Be it lack of standards in UI, lack of standards in package management and distribution, bugs, limitations in applications (I am switching back to IE on Windows because Firefox offline features suck), hardware interfacing, lack of vendor support, lack of helpful colleagues or whatever, but the end result is that for some 6 billion people and for a half of everyday computing needs (how do I enable WLAN on this laptop using Linux, in Windows I activate it by a hardware button?) work, not at least without major pain.
Point 2:
The most visible Linux people like the Slashdot crowd are unfriendly and fanatic: they don't want to consider a view of a person who hates computers and just wants them to work with as little as possible hassle.
Not very POSIX, but look at FOSS (unless you are picky) Inferno at www.vitanuova.com. A bit sidetracked these days, but designed at Bell Labs by Rob Pike, Dennis Ritchie and friends.
The fast pseudocode interpreter is as portable as it gets. It is available e.g as an ActiveX plugin, executing inside IE, running whatever code you need from the network.
Distribution is based on the 9P "object access" protocol. (Which is now supported by the Linux kernel.) It makes even the device driver interfaces network mountable. You can run an Inferno machine and mount the TCP/IP stack from another machine. Or stack-mount an "inheritable" combination of two different TCP/IP stacks (as 9p/Styx objects) from two different remote machines and at run-time force an unknowing application to use TCP through one remote machine and UDP through another.
Some anecdotes:
- Bell Labs negotiated with Sony about using Inferno as the OS for the Playstation 1, sometime in 1995. The deal failed. (But it resulted in a marriage between a couple from the two companies.)
- Microsoft beta-tested Inferno, but could not agree on a licensing deal. "The problem is that Microsoft has this view of the universe, where they are in the center..." said Dennis Ritchie. Bill Gates was asked to defend his company in the U.S. Congress and mentioned among other things Inferno, when he was trying to prove, that Microsoft does not have a safe monopoly position with no competing innovations.
Response to Everphilski proves his point. Linux hackers think that the "value" of Linux is in its utility for them as an elegant art object, or as an idealistic cry for freedom. That's ok, but the value of Linux to the general population or to the mankind is another story. It could be much bigger, if "convenience" would be appreciated over "quality". Joe Blow has his life to configure, he is not interested in configuring his computer.
The saddest thing is, that I can test a dozen Linux 2.6 based distros on my Fujitsu Amilo V2000 laptop, and my USB devices and my WLAN card may work on 6 and not work on 6 others.
It should definitely be, that if a working driver is available for "Linux", I should be able to download and use it on all (roughly same age) Linux versions, without re-compiling the kernel. Working "Linux 2.6 drivers" should be easily downloadable from one well known place to all "Linux 2.6" installations, no matter what the distro is. This is what Joe Blow has energy to do, not more.
Note that I don't ask the drivers to be in the kernel space, as long as they have a standard binary interface. I do encourage people, who want to provide closed source drivers right now, to provide them with syscalls to user space. Or better yet, use local the (new in kernel) 9P server interface as the driver interface!
Actyally I think the whole problem of Linux starts to be, that it is so hard to produce and market closed source application software for Linux, and make a living on it. The GPL looms there always as the scarecrow, and if you try to artificially avoid using any open source components, you will have to re-invent sqare wheels to your proprietary "cars" for the Linux highway. And the community is more or hostile or at least lukewarm to all proprietary software, no matter how good it is.
Well, being a Slashdotter and a hacker/geek is supposed to mean passion for new computer technology. New is supposed to be better than old. Implementing standard service with standard money is not good enough. The whole point (or the one rational point) of Slashdot is to discuss ways to do better with less money.
Pardon me, Finland was not occupied by Russia. We fought on two occasions against the Soviets. We were never occupied though we finally surrended the war and gave parts of Eastern Finland to the Soviet government.
Note that from a libertarian perspective having just any "democracy" is not enough, and at the same time may be too much. Democracy can be tyranny of the majority. The power that a democracy may wield must be limited.
Regardless of democracy, and according to libertarian ideals, everybody should have sovereignty over his personal matters. You should have sovereignty over your body (drugs, abortion), your possessions, your income (taxes are dubious), and your free contracts with sellers, buyers, employers and employees (no regulation of businesses).
If you agree with me, that the current democracy uses two much tyrannial majority powers in a socialist and collectivist fashion, you don't want to make that power more agile and dangerous with "direct" democracy. And the less power that is left at the public democracy, the less need there is for any direct democracy.
There is a chinese wisdom about arts: "If it looks real, it is not art." I think the "mimesis paradox" is also known in western art philosophy: that striving for realism is kind of futile, because absolute realism would be in no way more beautiful or fun than the reality already is.
Of course, it might be better to differ from reality by the ways of the artists all-powerful mind, and not because of limitations of our tools. So photorealism here and some fantasy somewhere else makes sense. But if you insists on photorealism, realistic chracter AI, a working realistic environment and complete freedom of storyline, what is there left as the "art"?
In his famous studies about "emotional intelligence" Daniel Goleman argues that statistical correlation between salary and subjective happiness studies is strongest with internal emotional abilities like:
- being aware of your own feelings - self-discipline - self-esteem - self-respect and being able to stand-up for your rights
And intra-personal emotional skills like:
- being able to notice other people's feelings - communication skills - "niceness" - honesty
Traditional IQ tests at young age have weaker correlation to future income than these emotional intelligence test.
I think it is important to undertand that success is a) about producing something for _people_ and b) getting something nice from _people_ in return. It makes you poor to concentrate on too individual goals or skills, which are not "marketable" in a society, no matter have clever you are. And the worst thing for your happiness is to be blinded by some intellectually challenging factual and rational things which don't respond even to your own deep emotional needs, which you may not even recognize. Or which you may even be reluctant to explore, like if you have some childhood trauma, or ethical conflicts inside yourself.
Slashdot Mirror
Which is more important for individuals? Which is bigger in terms of national economics, basic goods or communication infrastructure?
So if you are against free markets as in not-regulated and not politically-controlled markets, why don't you require, that gocery stores everywhere should provide price and quality neutral access to all goods for all good men?
We Europeans are supposed to be more leftist than U.S.A., but we just don't get the fuss over "net neutrality". We don't have "universal access" or free-as-in-communims local calls either, and we could not care less.
Much more Google techie info is on Rob Pike Usenix 04 talk: "Cheap Hardware + Fault Tolerance = Web Site" http://www.usenix.org/events/usenix04/tech/thurs.h tml
Did you know for example, that he says fancy cooling is only necessary, if your components are in a box. Wrap them to racks with velcro, drop the boxes, and you need no active cooling.
Manhole (http://en.wikipedia.org/wiki/The_Manhole) was better. And much more arty. It was even B/W. Highbrow as anything, highbrow as Lewis Carroll.
FRM seems like an interesting research topic, but while addressing the limitations of now fashionable Single Source Multicast (SSM), it is an overkill for P2P. P2P could do fine with very simple SSM.
But in the end I doubt the feasibility of IP (layer 3) multicast. SSM solves the multicast routing problem, and source/group discovery and advertising can now be easily done manually, but other multicast problems remain: synchronicity (everyboday has to recieve at the same time), least common denominator bandwith requirement (everybody has to recieve at the same bandwith), inflexibility of UDP traffic (does not adapt to network congestion like TCP, which is scaring the ISPs) and above all, the need to have unbroken chain of supporting routers in order to work at all.
P2P already is application level multicast. It can work without support from the routers, with flexible ISP friendly TCP, and with no synchronized bandwith or time synchronization of clients. Besides, application level allows all kinds of innovative optimizations like compression, caching, security, anonymity, billing or whatever. And multicast for video streaming really might require different approach than multicast for file delivery, it is not a good idea to force them to use the same low level mechanism.
Besides, market fragmentation of P2P protocols is not such a bad thing, because f.ex. all my stuff is there in my iTunes or video directories, and I share it on all my active P2P networks anyway.
First find a good distro spesific IRC channel. (Learn IRC if you are an old fart like me who originally hated this kid stuff.) Whatever you have in your mind it is always a good idea to utter something to IRC. For example the Ubuntu default IRC client automatically goes to Ubuntu channel when you open it, and so does bitchx IRC client on Debian.
WHILE waiting for answers/flames read man pages/docs or google, and also the Google Groups. How to balance you problem solving between "official" docs of the SW at hand and more free googling? It depends on the case. If you have an error message, definitely do google with it, with and w/o quotes. AT THE SAME TIME, when you learn anything new with man/docs/google, you can augment or rephrase your question on IRC. On the other hand, if someone answers on IRC, pick and feed better keywords for further googling. In the end, even if nobody answered on IRC, it is an altrustic idea to write your final solution there, because nobody seemed to know it. And they can maybe verify it.
Only when all else fails, ask on a distro specific forum. That is so much slower and rarely gets you any better results.
In difficult application spesific cases you of course have to move the issue to application spesific IRC channels or forums, but it is always a better idea to start on distro channels and forums, because you might be using the sub-optimal application in your environment in the first place. And the help in distro channels/forums is more accessible, especially for newbies. And everybody is a newbie when starting with a new application. And the answers might be distro specific.
Alko known as "movies". Can be distributed on Steam. Or an a Web page. Two fifths of Episode One lenght, two fifths of the price.
;-)
Or how about something in between? A movie where you can interact a little here and there, but you don't have to, if you are lazy. Or there can be a plot that is fixed or has a limited set of variations, and action is done by AI NPCs with random variation on each playtime, where a random result of some action scene chooses a plot variation.
You know, me and my 3 year old son love to watch the Monaco Grand Prix in TV mode in GP4. We usually do that when he goes to bed so that he falls a asleep though, so it is not SO exiting. But it is as good as the real one, except that you can control the camera IF you want
Seriously, could this become a new art movement? Unauthorized, pirated "remixes" of movies, with dull parts edited out, added dialog to clear up confusion/contradictions in plot, better effects, new scenes without the original characters, or in the extremely ambitious case, new scenes with stunts of the characters? (Or aunauthorized appearances of bribable original actors ;-)
In the 80's they still said that everybody is buying IBM equipment because nobody was ever fired for that. I guess these days "nobody is ever fired for building a Windows based system".
Well, it works for me, with the Firefox Coralize extension. And Coral always worked for me when pages were slasdotted.
The article talks about backup. The idea could be, that instead of managing incremental backups you just optimize compression of data that is similar to old data. In that way you can do "full" backups, but actually save only incremental backup worth of data.
See http://en.wikipedia.org/wiki/Venti for similar ideas in a system that easily achives 25x compression for typical archival storage. When a file has been changed only those 512 kbyte blocks that are really new are saved, other blocks are just mapped by their SHA1 hashes to existing blocks. So files with small changes, very similar files and files sharing common parts will all compress very nicely. In a multi-user system the files of different users tend to also have lots of similar parts: same emails, same office documents with perhaps minor changes, same reference material / tools / libraries as personal copies etc.
My guess is TFA refers to a re-invention of this wheel, most likely in an inferior way.
If you have a single application device, it really does not need a comprehensive, separate "OS". If and when you need process management and memory management and sharing of I/O resources between modules/processes, you need SOME libraries/headers to do that. If you want to have layers of abstraction (say, high level I/O or media codecs) to help in development, you need also some libraries for that.
But up to a point you are free to use per application designed common libraries and run-time systems, you don't need to have a general purpose interface/protocol between "applications" and the "operating system". This is what many light-weight embedded toolkits are: just run-time systems and shared libraries with shared data structures. They are far simpler and smaller than any version of Linux, and more tailored to typical needs here.
But if you want to have a user extendable system and an ecosystem of independent code modules and libraries all running smoothly on a variety of hw versions, with standard programming interfaces, then Linux probably is a reasonable choice. (Though I would go with Plan 9 or Inferno.) This kind of system needs more expensive hw and is more complex, but it is also kind of easier to build, because it is more "common" and more "standard". With the cost of hw falling, this type of system is going to be more and more the most cost effective way to engineer the life cycle of several product generations, when design time and itellectual property costs are bigger than costs of general purpose processors, DSPs and memory.
We use Skype as the official company VoIP and conferencing sw. We are a small company, and we have a persistent Skype chat open with all our globally travelling workers, and we cut global phone costs with Skype. Internal meetings are often done is Skype. The two good things are that it works through many kinds of firewalls, and that it encrypts.
I didn't bother to read the link you provided, because as an IPv6 trainer in our company I already know:
IPv6 is not needed, NAT works. Some people will even insists NATing their IPv6 network to protect the internal addresses.
You don't need IPv6 to have global reachability for VoIP and P2P. Teredo actually proves, that you can contact anyone with a private address if you really want. You just need some form of global addressing, and an active "NAT circumvention" server like Teredo. Skype and other P2P networks that work with private addresses also prove, that it is easy to contact anyone globally, know matter how private addressing they use.
Transition to IPv6 is huge work. Upgrading all software is huge work. Running two parallel network infrastructures is huge work. Nobody wants to do that.
Re-checking your security infrastructure (firewalls, IDS, whatever) for IPv6, and for half a dozen clever tunneling tricks is something your security guys will NEVER do. And thus, they will never allow IPv6.
Mobile IP (MIP) would be nice and the only real reason to insist everybody having a public IPv6 address. But after Skype and P2P networks have been made to work on private addresses, you don't really need MIP so much. The cost of having IPv6 just for that is too much.
But the current problem is just that no one can agree on a common hierarchy, and never will. And because the dispute is global, it is difficult, and not very well handled by a monopoly of any sort, not even the U.N. (God forbid not!).
Besides, nobody actually reads or expects to read any info from relation of the sub-domain names in "namex.namey.namez".
On the other way, in my system you would be free to set up any "logical" hierarchy or hierachies you wish, and anybody could register there, if you make it your business service to provide registration to your neat and tidy logical branch of the name tree.
And one thing about squatting disputes: note that I could try to squat "coke." ("c.o.k.e."), due to possible legal challenges, but the company could just use coca.cola or coke.drink or coke.ltd or coke.adds.life or whatever. Squatting is only meaningful while the current TLDs remain a scarce, monopoly controlled resource.
Basic squats ("coke."/"c.o.k.e.") would anyway be solved exactly like now: anything can be registered, but if someone later can show a legitimate interest in the domain and if your interest is definitively squatting only, you loose it to the better owner for free. It is not a big problem today, and it would be less a problem in my system.
This domain squatting problem would be no worse than today. And squatting a single mid-level letter would give you no benefit, just costs, because you would be required to allow (free) delegation of any sub-letter domains.
This would let everyone run their own pseudo "top level domain" without stressing the global system too much. The goal here is especially that the DNS cache hit ratios would not drop dramatically because of multitude of global domains.
6 8.html
I once tried to discuss this on the DNSO mailing list, but finally could not find a right place:
"A letter hierarchy based scheme for arbitrary TLDs"
http://www.cafax.se/dnsop/maillist/2001-05/msg000
Let's first set up top level domains for all single letters a., b.,...z. Then if I want to have a name "anssi.abc." for my machine. I rewrite it (first in my head) to be "anssi.a.b.c." I go to the "c." TLD authority. I ask them, if they have already delegated the 2nd level domain "b.c.". If such domain exists, I go to its authority, and ask, if they have a domain "a.b.c." already.
If the domain "b.c." did exist, but the domain "a.b.c."
did not, then I register my "a.b.c." servers with the domain "b.c." authority, to set up new delegation. There should be a standard IETF rule, that single letter domains must allow delegation of any unregistered single letter subdomains!
So now I can set up names like "anssi.a.b.c." in my own name server, serving the domain "a.b.c.", registered at "b.c.". NOW I MODIFY MY RESOLVER and ask everybody to do the same, if they please! The new versions of resolvers would be standardized to TRANSLATE ALL UNRECOGNISED multi-letter TLDs into a sequence of one letter domains before the query. So "anssi.abc." will be resolved as "anssi.a.b.c.".
The classical, now existing TLDs can be handled as a special case. But now anybody is also free to set up whatever TLDs they want to!!! You greatly benefit from an updated resolver, but all resolvers will still work with the old TLDs, and old resolvers will also kind of work with the new names, but only if you translate manually, writing "anssi.a.b.c" instead of "anssi.abc".
You don't have to touch applications, but of course a Web browser might be aware of the new system, and compensate for non-upgraded resolvers. Anyway, writing HTML anchors like "anssi.a.b.c" would work also in old resolvers.
Otto and Clockwurk are right, but Linux is still "good" for many things, and getting better faster than Windows is.
Otto's anyway has two very good main points:
Point 1:
Linux has not reached the "it 'just works' for everyday people on everyday needs" state. Today, n people in the world are capable of doing m everyday things on Windows (or a Mac) without trying very hard. For Linux the n x m is just orders of magnitude less than for Linux. Be it lack of standards in UI, lack of standards in package management and distribution, bugs, limitations in applications (I am switching back to IE on Windows because Firefox offline features suck), hardware interfacing, lack of vendor support, lack of helpful colleagues or whatever, but the end result is that for some 6 billion people and for a half of everyday computing needs (how do I enable WLAN on this laptop using Linux, in Windows I activate it by a hardware button?) work, not at least without major pain.
Point 2:
The most visible Linux people like the Slashdot crowd are unfriendly and fanatic: they don't want to consider a view of a person who hates computers and just wants them to work with as little as possible hassle.
Not very POSIX, but look at FOSS (unless you are picky) Inferno at www.vitanuova.com. A bit sidetracked these days, but designed at Bell Labs by Rob Pike, Dennis Ritchie and friends.
The fast pseudocode interpreter is as portable as it gets. It is available e.g as an ActiveX plugin, executing inside IE, running whatever code you need from the network.
Distribution is based on the 9P "object access" protocol. (Which is now supported by the Linux kernel.) It makes even the device driver interfaces network mountable. You can run an Inferno machine and mount the TCP/IP stack from another machine. Or stack-mount an "inheritable" combination of two different TCP/IP stacks (as 9p/Styx objects) from two different remote machines and at run-time force an unknowing application to use TCP through one remote machine and UDP through another.
Some anecdotes:
- Bell Labs negotiated with Sony about using Inferno as the OS for the Playstation 1, sometime in 1995. The deal failed. (But it resulted in a marriage between a couple from the two companies.)
- Microsoft beta-tested Inferno, but could not agree on a licensing deal. "The problem is that Microsoft has this view of the universe, where they are in the center..." said Dennis Ritchie. Bill Gates was asked to defend his company in the U.S. Congress and mentioned among other things Inferno, when he was trying to prove, that Microsoft does not have a safe monopoly position with no competing innovations.
Response to Everphilski proves his point. Linux hackers think that the "value" of Linux is in its utility for them as an elegant art object, or as an idealistic cry for freedom. That's ok, but the value of Linux to the general population or to the mankind is another story. It could be much bigger, if "convenience" would be appreciated over "quality". Joe Blow has his life to configure, he is not interested in configuring his computer.
The saddest thing is, that I can test a dozen Linux 2.6 based distros on my Fujitsu Amilo V2000 laptop, and my USB devices and my WLAN card may work on 6 and not work on 6 others.
It should definitely be, that if a working driver is available for "Linux", I should be able to download and use it on all (roughly same age) Linux versions, without re-compiling the kernel. Working "Linux 2.6 drivers" should be easily downloadable from one well known place to all "Linux 2.6" installations, no matter what the distro is. This is what Joe Blow has energy to do, not more.
Note that I don't ask the drivers to be in the kernel space, as long as they have a standard binary interface. I do encourage people, who want to provide closed source drivers right now, to provide them with syscalls to user space. Or better yet, use local the (new in kernel) 9P server interface as the driver interface!
Actyally I think the whole problem of Linux starts to be, that it is so hard to produce and market closed source application software for Linux, and make a living on it. The GPL looms there always as the scarecrow, and if you try to artificially avoid using any open source components, you will have to re-invent sqare wheels to your proprietary "cars" for the Linux highway. And the community is more or hostile or at least lukewarm to all proprietary software, no matter how good it is.
Well, being a Slashdotter and a hacker/geek is supposed to mean passion for new computer technology. New is supposed to be better than old. Implementing standard service with standard money is not good enough. The whole point (or the one rational point) of Slashdot is to discuss ways to do better with less money.
d =13874861
I think bernz put up a very good example there in a previous comment http://ask.slashdot.org/comments.pl?sid=166332&ci
Pardon me, Finland was not occupied by Russia. We fought on two occasions against the Soviets. We were never occupied though we finally surrended the war and gave parts of Eastern Finland to the Soviet government.
Note that from a libertarian perspective having just any "democracy" is not enough, and at the same time may be too much. Democracy can be tyranny of the majority. The power that a democracy may wield must be limited.
Regardless of democracy, and according to libertarian ideals, everybody should have sovereignty over his personal matters. You should have sovereignty over your body (drugs, abortion), your possessions, your income (taxes are dubious), and your free contracts with sellers, buyers, employers and employees (no regulation of businesses).
If you agree with me, that the current democracy uses two much tyrannial majority powers in a socialist and collectivist fashion, you don't want to make that power more agile and dangerous with "direct" democracy. And the less power that is left at the public democracy, the less need there is for any direct democracy.
There is a chinese wisdom about arts: "If it looks real, it is not art." I think the "mimesis paradox" is also known in western art philosophy: that striving for realism is kind of futile, because absolute realism would be in no way more beautiful or fun than the reality already is.
Of course, it might be better to differ from reality by the ways of the artists all-powerful mind, and not because of limitations of our tools. So photorealism here and some fantasy somewhere else makes sense. But if you insists on photorealism, realistic chracter AI, a working realistic environment and complete freedom of storyline, what is there left as the "art"?
Dijjer!? Great! I didn't now about that!
In his famous studies about "emotional intelligence" Daniel Goleman argues that statistical correlation between salary and subjective happiness studies is strongest with internal emotional abilities like:
- being aware of your own feelings
- self-discipline
- self-esteem
- self-respect and being able to stand-up for your rights
And intra-personal emotional skills like:
- being able to notice other people's feelings
- communication skills
- "niceness"
- honesty
Traditional IQ tests at young age have weaker correlation to future income than these emotional intelligence test.
I think it is important to undertand that success is a) about producing something for _people_ and b) getting something nice from _people_ in return. It makes you poor to concentrate on too individual goals or skills, which are not "marketable" in a society, no matter have clever you are. And the worst thing for your happiness is to be blinded by some intellectually challenging factual and rational things which don't respond even to your own deep emotional needs, which you may not even recognize. Or which you may even be reluctant to explore, like if you have some childhood trauma, or ethical conflicts inside yourself.