There are others, but that's the version I'm familiar with. The last time I tried it, over 5 years ago now, it couldn't successfully run KDE unside of MSWind98. Close though. Even then if you were satisfied with native MS windowing, it was quite good.
Other people have other reasons. Mine begin with and revolve around their EULA. I tend to take agreements seriously, so I read the bloody things (unless they purport to be standard, e.g. I always assume that something which claims to be GPL v.2 actually IS GNU General Public License version 2 as released by the FSF).
As a result of reading the EULAs, and a bit of thought, around 1999 I started looking for alternatives. Around 2000 I decided that Linux was the best available option, and began moving. Now I have one MSWind95 computer. It's about to die a slow death, due to lack of support for more recent peripherals. If it's replaced by a computer, that computer will run Linux. It may become a free desk space (which would be, perhaps, more valuable at this time). It WON'T be upgraded to a more abusive license. I don't think I need another Mac. Another Linux is plausible...but un-allocated flat surfaces are also quite valuable.
This is going to cause me considerable hassle, even though I've known this day was coming for quite awhile. One never seems to prepare sufficiently. Their are still captive files, created by some application that didn't document it's file format, and which aren't readily exportable in more than a minimally usable manner. Perhaps I'll get through this bottleneck (i.e., finding a color ink-jet printer that will print to MSWind95 via a centronics port). If I do though, this is just a warning signpost. Obsolescence nears. (The computer isn't obsolete yet...but MSWind95 doesn't handle USB connections. And doesn't handle the CD drive created by VMWare. I can't even re-install the OS in an emulator. [It's GOT to be MSWind95... the application doesn't work properly with MSWind98.])
GPL software goes obsolete just as rapidly...but you can figure out the file formats.
You overestimate the integrity of scientists and the degree of peer-review. Still, the basic point is sound, if somewhat over-stated.
It needs to be remembered that government isn't the only, or the most flagrant, abuser of scientific research. Commercial firms are, if anything, worse (on the average).
Also, there's a culture against the reporting of negative findings. These are just as important as positive findings, but they don't tend to qualify for publication OR for alternate forms of public exposure and preservation.
Things aren't very rosy. Computer science is, perhaps, one of the purest forms of science around. This is partially because of it's strong footing in mathematics, but even more strongly because it's easy and cheap to check out revealed algorithms and procedures. The GPL is one solid foundation here. It ensures the publication of significant results. (Negative results are still not recorded or revealed.) I tend to think of the GPL as the scientific ethos solidified into a legal structure.
Were I to chose this approach, Erlang would be the "slow but flexible interpretive language", so it would need to be a replacement for Python rather than connecting to it. The back end language would need to be D (Digital Mars D), and since that's still in development I'm rather sure I'd have to work out how to make the connection (via C mimicry). Unfortunately, this would mean that NEITHER of the projects two languages would have a decent collection of libraries...so now I'd need to add in a third language, probably C. (I don't like to do much in C, so it can't be one of the major languages, but it does have a large selection of libraries.)
This has gotten too complex. Since I haven't yet really learned Erlang, that's the piece that needs to go. So I've got to find another way to solve the problems that it would have solved (basically only the concurrency aspect).
This, of course, is dependent on Pyd becoming a stable project. Otherwise it will need to be Python/Pyrex and C, which would be slower, as more of the work would be done in Python. (I don't object to doing lots of the code in D...but D isn't flexible at run time. It concentrates its flexibility at compile time. Otherwise I'd JUST use D...though then I'd run up against the problem of libraries...but that's probably soluble, though with effort, for each library as needed.)
These days it's quite a gamble to go to college at all. To pick a field where it takes a lot of extra work to get good grades just increases the ante. The payoff comes if you get a good job when you graduate.
If most of those who graduate don't get good jobs, the payoff is seen as worse. Thirty years ago I saw a study that claimed that college was of only marginal economic utility to those who went to it. (Sorry, this was a global estimate that included Sculpture majors with Electrical Engineering majors with Business School majors.) I know that since that time college has gotten more expensive. OTOH, it's quite plausible that without college one's chances have gotten worse.
Were I starting today, would I think college was a good investment of time, effort, and money? If I did would I be right?
Prizes seems like an attempt to get a cheap labor pool by trickery and fraud. If you want to have good students commit time, effort, and money, then show them that they can expect a decent payoff for their investment.
I have looked at it with great interest...but doesn't it require all concurrent instances to be running on the same virtual machine? (It's been awhile.) This is one of the things that I would need to avoid if I want to get any advantage out of multiple processors.
OK. I didn't do my own timing, so I don't know for certain that the times are valid. (But you might want to update the language examples on that test site.)
Second, I've used Lambda. I don't like it. I don't find it a reasonable substitute for classes with inheritance....Strike "I don't like it". It's fine in small simple cases, and I occasionally use it.
Note: I am stating this as a personal preference. I know people who are enthusiastic over Scheme. I've occasionally considered Lisp. (It would have a lot going for it...fast, simple syntax, etc.) I don't use Lisp, because I really don't like the syntax. This is similar to the way I don't use C because I don't like pointers. OF COURSE everything is really done with pointers anyway, that's not the point. So if it's at all possible I will use D rather than C or C++, or I'll even use Eiffel. I'm not certain about Ada...that starts getting too verbose and finicky. (Multiple incompatible types of strings of exactly the same characters is a bad idea! At least when they only have syntactic meaning rather than semantic meaning.) Sometimes I consider Java. It's got a lot going for it. (Libraries, mostly.) But I prefer D, and if it's a reasonable choice I'll use it.
Actually, I really prefer Python, and even more Ruby...but both of those are slow, so I need a fall-back position. Somehow D ends up testing faster than C (which shouldn't be possible), so it's a fair choice for when things get time-critical. NONE of the languages that I seriously consider make heavy use of Lamba expressions...except internally. I'm sure that internally they make heavy use of assembler, too, but I don't want to.
FWIW, I once wrote a primitive Lisp and it did use lambda expressions. This did not cause me to love them. (It wasn't a great Lisp, and I'm rather glad I never made the code public, but I did learn about Lambda expressions. They aren't a reasonable substitute for objects and classes. Prototypes might be...but Io is slow compared to Python, and Prothon is dead. So is that Smalltalk derivative that was going to be based around prototypes. If I need them, I can do them in Python without too much grief.)
The phone company was the object of scorn for being high-handed and arbitrary. I never, however, heard anyone say that Bell Labs wasn't innovative. The phone company may have tried to stamp out modems, but Bell Labs wasn't blamed for that, and didn't deserve blame for that.
OTOH, can anyone say that MS has an equivalent organization? If so, they definitely hide it.
Another problem that I have with it is the lack of any object model, or any decent tutorial.
It does do concurrency well, but that, by itself, isn't sufficient. (I just checked it out for a couple of days recently. Not long enough to really know the language, but long enough to get a high-level feel of it.)
One nice thing about Erlang is it's good connection to a database. (Well, not just ANY database, but a particular one, Mnestra or something close. It's FOSS.) Unfortunately, this doesn't suffice. (The database probably does, but database + concurrency doesn't suffice to make Erlang a viable choice for me.)
My current idea for a hack is connection via TCP connections and pickled data. I'll need to nail down the details when I get closer to implementation, but for now it looks like Python is the high level language I'll choose, with either pyrex + C or pyd + D for low level modules. (D has LOTS of advantages over C, but C has the libraries...OTOH, so does Python.)
...fool me three times - that's enemy action.I think you got that part right. I stopped considering them reliable quite some while ago, though, and missed getting burned through trusting them. It was happenstance, and good fortune rather than good sense, but I wasn't ever that please by SuSE, and didn't like SUSE. (I bought a copy of 9.1 because I liked something that they had done, and wanted to support them, but it didn't stay installed very long.)
It's interesting, because I have 5 different partitions each with a different copy of Linux installed, but the only "commercial" ones are the two Ubuntu partitions, and I spend most of my time in Debian Etch. This isn't because I haven't tried lots of distributions...I could see 10 different boxed versions before I cleaned off my shelves earlier this year, but they aren't the ones I stayed with.
I have a suspicion that companies are basically untrustworthy. Red Hat will probably remain (relatively) trustworthy just a long as Bob Young remains in control. Ubuntu as long as Shuttleworth stays in charge. So the only answer is to have a strong license. GPL3 may suffice. Something that means that we WILL continue to be able to fork the code base, as it's only by frequent forks that FOSS can remain viable, and companies be prevented from being obnoxiously predatory.
Step 10. (You did say "Quick frozen", and that requires a blizzard of sub-Siberian temperature.)
OTOH, if you just mean the Thames freezing in autumn and thawing late in spring, that's step "The great conveyor shuts down".
Still, do remember that I said I didn't know how long any step took. I can't rule out several of them happening within the same year. (OTOH, the ocean temperatures change rather slower than air and land temperatures. Allow at least a few decades for most of these. The great conveyor shutting down would likely be due to Greenland shedding lots of fresh water, which would decrease the salinity of the surface, resulting in the surface water not being denser than the deeper water even when it was colder. [So floating buoys measuring temperature would be much use. Instead they need to measure BOTH temperature and salinity.])
N.B.: Current results indicate that ice sheets can "melt" a lot faster than previously supposed, due to surface water tunneling down to the base rock and lubricating the ice flow. I haven't heard of any good current estimates of the maximum expectable rate of "ice shedding" that include this effect in the models. I also don't know just how much ice needs to be shed the shut down the great conveyor. It's happened before, however. (But WE weren't around then, so we didn't care.) The last time it's reported to have happened was towards the end of the last ice age...but then the ocean was already cool, and so water wasn't being evaporated.
P.S.: If I'm remembering things correctly, the great conveyor shutting down would, indeed, make Europe VERY cold, but since the ocean wouldn't be circulating warm water up along the coast there wouldn't be much precipitation. Thus you wouldn't get an ice age, because the winter would be too dry. Perhaps summer would be also...in which case Europe would, instead, turn into a cold desert...think steppes. (Why do the models focus on Europe instead of the US? My guess: Europe is simpler to model.) If this should be correct, then could I point you to the recent development of a wave powered desalinization device? It's not yet ready for engineering trials, but perhaps someone should invest in it? (Sorry, I forget the source. Check New Scientist for the last couple of months.)
Do you know that it's easier to predict the seasons than the weather?
Why should you be able to predict the seasons when you can't tell whether it will rain tomorrow?
I'm sorry, but your argument "Why wasn't 2004 worse than 2003 which was worse than 2002 and so on? " appears silly. Your general argument would be worth addressing, except that your tone implies you won't understand the replies.
However, I consider planetary albedo (generally ice coverage) more significant than hurricanes. Frequent and intense hurricanes are a predicted end result from certain weather models. Ice is directly measurable, and directly causative.
Also, global changes are more significant than local changes. Rising sea levels are more significant that melting glaciers. One may contribute in a minor way to the other.
P.S.: "The Day After Tomorrow" was based on an absurd story line, and it compressed (hopefully) centuries worth of change into a year or less, but the basic concept is reasonable. (Except that Mexico would welcome folk from the US in such an event.) It bears as much relation to science as Shakespeare Histories bear to history. I.e., there are several points of match-up, but it's been adapted to suit supporters more interested in political propaganda than facts.
P.P.S.: Also there's currently a question of whether the shutdown of the great conveyor would actually cause an ice age in Europe. AFAIK this hasn't yet been decided. But at the time that the movie was made it was the orthodox belief, where now it's disputed. (I haven't heard any report on the effect on the US. One can presume that this is also in dispute.)
N.B.: IF the shutdown of the great conveyor leads to an ice age, then a "big melt" can be expected to be the normal prequel to a new ice age. The mechanism would work sort of like this: 1) Temperatures rise 2) Ice melts 3) The earth starts absorbing more sunlight, causing temperatures to rise more quickly. 4) Enough ice melts to shut down the great conveyor. (Say, much of Greenland.) 5) The great conveyor shuts down. 6) The poles freeze and the equator heats up. 7) The new ice on the poles reflects lots of sunlight, so the earth cools down, but the oceans are still warm, so there's lots of moisture evaporating. 8) Lots of precipitation, cloud coverage increases. In the north snow starts piling up. 9) More clouds & snow means more sunlight reflected. Things cool more. 10) The glaciers begin walking south. 11) Finally the oceans cool off. There's now less precipitation. 12) Cloud coverage decreases, but there's all that new snow...so things keep getting cooler for awhile....... I don't know what's supposed to happen here. Something eventually causes the glaciers to retreat.
Note that an important part of this cycle is the thermal mass of the oceans. When they are warm, they evaporate a lot of water. When they are cool, the amount of evaporation is a lot less.
N.B.: I don't say anything here about how long any particular step lasts. I don't know. Some indications seem to show that the glaciers may begin a rapid walk south in as little as one year. (Quick frozen Mammoths with tender green vegetation still in their stomachs, e.g. Perhaps there are other explanations.)
This think reads like it was scripted by the PR department.
Also, I notice that they had things rigged so that they could censor any questions they didn't like. (Reasonable, an open forum would have been a mad house, but not exactly a process that builds trust.)
They also didn't say anything about which of their customers could redistribute what. The short answer appears to be "We aren't interested in developers."
I'll accept that GNU/Linux is "non-sense". (It isn't really, but it's a bad idea because short words tend to be more acceptable for high usage.) So I mean something besides that.
Nothing that we know of obligate Novell to publish MS IP. But they've just bought insurance that allows them to do so without worrying about getting sued. Also there are many parts of the deal that have not been made public.
There is no *proof* that we have any more to worry about than before. But given patent law, if you wait for proof your business will be toast. The question is "Can you trust them", and they've just bought insurance that allows them to grossly violate your trust without worrying about getting sued. It's clearly not proof...but it removes most grounds for trusting them. And if you don't trust them, then public statements that aren't legally binding are worthless.
They do have valid reasons for declaring the contract void. Both sides have publicly declared that there has been "no meeting of minds". But if they were to void the contract, then they would need to return the money, so they don't want to.
And they it is proper that they be despised and turned into pariahs. They should be seen as the carriers of a loathsome infectious disease, and folk who value their own well-being should refuse to approach them, or to accept merchandise from them.
The thing that bothers me about Gnome is their lack of a decent menu editor. This appears trivial, but I find it so annoying that I've never looked much further.
This isn't quite true, I used Gnome by preference back around KDE1...but not since they removed the menu editor.
I believe that BSD uses the gcc. I could be mistaken. as I'm not really that well acquainted with it.
But note that Darwin uses the gcc under the GPL, and doesn't try to impose non-GPL acceptable conditions on the use of GPL code. Novell doesn't appear to have that option.
I don't think that's going to be optional. Rejecting code from Novell is going to be a necessary defensive act. I've already stopped accepting updates from them to old software (so I'm looking for alternatives when the applications are necessary...fortunately I'd only just started investigating Mono, as I'd been worried about IP of SOME [undetermined] nature all along).
If Novell has been making significant contributions, then this will be a sad loss. But necessary.
If you think this is purely an emotional argument, then you have read the publicly released portions of the agreement. Or you haven't analyzed what actions could be taken on the basis of those agreements.
"IBM thinks its a good agreement"? This bothers me, and if the arguments presented made sense, then I'd re-think my position. When I read them, though, they don't make sense unless you're a large company that is already maintaining a stable of lawyers. I can clearly see why the agreement wouldn't significantly bother or threaten IBM, but IBM's situation is not at all similar to that of the average FOSS developer.
You must not be an experienced programmer. ALL code develops incompatibilities and bugs appear as the environment changes. Only "short-time" projects can even appear stable. I've had code last for five years. It wasn't doing much that was environment facing.
Well, I suppose that if it's simple enough.... "hello, world" still works, and that's been decades, but once things start getting at all complex things change.
You've got the point, but you don't understand it.
My wife teaches privately, and can select here students. She has noticed that when the mother is working outside the home, the kids are less developed. Her belief (based on evidence, but I don't know it all) is that when the mother's come home they are too tired to do a decent job of being a mother, also. There's a lot of variation here, depending partially on what job the mother has, and what shape she's in, and, of course, on both her skill level, and the willingness of the child to be trained. But usually she can work with a child awhile and tell whether or not his mother is employed outside the home. (This is not a fair test, because she usually already knows, but she believes that this information tends to be expressed in the child's behavior.)
So... what's a fair answer, that doesn't have a cost to society that isn't excessively high?
P.S.: Wages have fallen since the 1980's, relative to prices, so that now it's usually mandatory for both parents to be employed. The two salary family now has no more money than the one salary family did in the 1980's, when measured by value of residual discretionary funds.
OK, I'm no expert, and my information is 2-3 decades old...however...
I have this strange suspicion that the stuff they enter into cell-phones is the rough equivalent to "texting" (or whatever that set of peculiar abbreviations is called).
Slashdot Mirror
Red Hat supports (used to support?) it.
There are others, but that's the version I'm familiar with. The last time I tried it, over 5 years ago now, it couldn't successfully run KDE unside of MSWind98. Close though. Even then if you were satisfied with native MS windowing, it was quite good.
Other people have other reasons. Mine begin with and revolve around their EULA. I tend to take agreements seriously, so I read the bloody things (unless they purport to be standard, e.g. I always assume that something which claims to be GPL v.2 actually IS GNU General Public License version 2 as released by the FSF).
... the application doesn't work properly with MSWind98.])
As a result of reading the EULAs, and a bit of thought, around 1999 I started looking for alternatives. Around 2000 I decided that Linux was the best available option, and began moving. Now I have one MSWind95 computer. It's about to die a slow death, due to lack of support for more recent peripherals. If it's replaced by a computer, that computer will run Linux. It may become a free desk space (which would be, perhaps, more valuable at this time). It WON'T be upgraded to a more abusive license. I don't think I need another Mac. Another Linux is plausible...but un-allocated flat surfaces are also quite valuable.
This is going to cause me considerable hassle, even though I've known this day was coming for quite awhile. One never seems to prepare sufficiently. Their are still captive files, created by some application that didn't document it's file format, and which aren't readily exportable in more than a minimally usable manner. Perhaps I'll get through this bottleneck (i.e., finding a color ink-jet printer that will print to MSWind95 via a centronics port). If I do though, this is just a warning signpost. Obsolescence nears. (The computer isn't obsolete yet...but MSWind95 doesn't handle USB connections. And doesn't handle the CD drive created by VMWare. I can't even re-install the OS in an emulator. [It's GOT to be MSWind95
GPL software goes obsolete just as rapidly...but you can figure out the file formats.
You overestimate the integrity of scientists and the degree of peer-review. Still, the basic point is sound, if somewhat over-stated.
It needs to be remembered that government isn't the only, or the most flagrant, abuser of scientific research. Commercial firms are, if anything, worse (on the average).
Also, there's a culture against the reporting of negative findings. These are just as important as positive findings, but they don't tend to qualify for publication OR for alternate forms of public exposure and preservation.
Things aren't very rosy. Computer science is, perhaps, one of the purest forms of science around. This is partially because of it's strong footing in mathematics, but even more strongly because it's easy and cheap to check out revealed algorithms and procedures. The GPL is one solid foundation here. It ensures the publication of significant results. (Negative results are still not recorded or revealed.) I tend to think of the GPL as the scientific ethos solidified into a legal structure.
"Truth! Truth! Truth!" crieth the lord of the abyss of hallucinations.
Were I to chose this approach, Erlang would be the "slow but flexible interpretive language", so it would need to be a replacement for Python rather than connecting to it. The back end language would need to be D (Digital Mars D), and since that's still in development I'm rather sure I'd have to work out how to make the connection (via C mimicry). Unfortunately, this would mean that NEITHER of the projects two languages would have a decent collection of libraries...so now I'd need to add in a third language, probably C. (I don't like to do much in C, so it can't be one of the major languages, but it does have a large selection of libraries.)
This has gotten too complex. Since I haven't yet really learned Erlang, that's the piece that needs to go. So I've got to find another way to solve the problems that it would have solved (basically only the concurrency aspect).
This, of course, is dependent on Pyd becoming a stable project. Otherwise it will need to be Python/Pyrex and C, which would be slower, as more of the work would be done in Python. (I don't object to doing lots of the code in D...but D isn't flexible at run time. It concentrates its flexibility at compile time. Otherwise I'd JUST use D...though then I'd run up against the problem of libraries...but that's probably soluble, though with effort, for each library as needed.)
These days it's quite a gamble to go to college at all. To pick a field where it takes a lot of extra work to get good grades just increases the ante. The payoff comes if you get a good job when you graduate.
If most of those who graduate don't get good jobs, the payoff is seen as worse. Thirty years ago I saw a study that claimed that college was of only marginal economic utility to those who went to it. (Sorry, this was a global estimate that included Sculpture majors with Electrical Engineering majors with Business School majors.) I know that since that time college has gotten more expensive. OTOH, it's quite plausible that without college one's chances have gotten worse.
Were I starting today, would I think college was a good investment of time, effort, and money? If I did would I be right?
Prizes seems like an attempt to get a cheap labor pool by trickery and fraud. If you want to have good students commit time, effort, and money, then show them that they can expect a decent payoff for their investment.
I have looked at it with great interest...but doesn't it require all concurrent instances to be running on the same virtual machine? (It's been awhile.) This is one of the things that I would need to avoid if I want to get any advantage out of multiple processors.
OK. I didn't do my own timing, so I don't know for certain that the times are valid. (But you might want to update the language examples on that test site.)
...Strike "I don't like it". It's fine in small simple cases, and I occasionally use it.
Second, I've used Lambda. I don't like it. I don't find it a reasonable substitute for classes with inheritance.
Note: I am stating this as a personal preference. I know people who are enthusiastic over Scheme. I've occasionally considered Lisp. (It would have a lot going for it...fast, simple syntax, etc.) I don't use Lisp, because I really don't like the syntax. This is similar to the way I don't use C because I don't like pointers. OF COURSE everything is really done with pointers anyway, that's not the point. So if it's at all possible I will use D rather than C or C++, or I'll even use Eiffel. I'm not certain about Ada...that starts getting too verbose and finicky. (Multiple incompatible types of strings of exactly the same characters is a bad idea! At least when they only have syntactic meaning rather than semantic meaning.) Sometimes I consider Java. It's got a lot going for it. (Libraries, mostly.) But I prefer D, and if it's a reasonable choice I'll use it.
Actually, I really prefer Python, and even more Ruby...but both of those are slow, so I need a fall-back position. Somehow D ends up testing faster than C (which shouldn't be possible), so it's a fair choice for when things get time-critical. NONE of the languages that I seriously consider make heavy use of Lamba expressions...except internally. I'm sure that internally they make heavy use of assembler, too, but I don't want to.
FWIW, I once wrote a primitive Lisp and it did use lambda expressions. This did not cause me to love them. (It wasn't a great Lisp, and I'm rather glad I never made the code public, but I did learn about Lambda expressions. They aren't a reasonable substitute for objects and classes. Prototypes might be...but Io is slow compared to Python, and Prothon is dead. So is that Smalltalk derivative that was going to be based around prototypes. If I need them, I can do them in Python without too much grief.)
The phone company was the object of scorn for being high-handed and arbitrary. I never, however, heard anyone say that Bell Labs wasn't innovative. The phone company may have tried to stamp out modems, but Bell Labs wasn't blamed for that, and didn't deserve blame for that.
OTOH, can anyone say that MS has an equivalent organization? If so, they definitely hide it.
Unfortunately, Erlang is SLOW when doing anything EXCEPT concurrency. (At least when rated by http://shootout.alioth.debian.org/ ).
Another problem that I have with it is the lack of any object model, or any decent tutorial.
It does do concurrency well, but that, by itself, isn't sufficient. (I just checked it out for a couple of days recently. Not long enough to really know the language, but long enough to get a high-level feel of it.)
One nice thing about Erlang is it's good connection to a database. (Well, not just ANY database, but a particular one, Mnestra or something close. It's FOSS.) Unfortunately, this doesn't suffice. (The database probably does, but database + concurrency doesn't suffice to make Erlang a viable choice for me.)
My current idea for a hack is connection via TCP connections and pickled data. I'll need to nail down the details when I get closer to implementation, but for now it looks like Python is the high level language I'll choose, with either pyrex + C or pyd + D for low level modules. (D has LOTS of advantages over C, but C has the libraries...OTOH, so does Python.)
...fool me three times - that's enemy action.I think you got that part right. I stopped considering them reliable quite some while ago, though, and missed getting burned through trusting them. It was happenstance, and good fortune rather than good sense, but I wasn't ever that please by SuSE, and didn't like SUSE. (I bought a copy of 9.1 because I liked something that they had done, and wanted to support them, but it didn't stay installed very long.)
It's interesting, because I have 5 different partitions each with a different copy of Linux installed, but the only "commercial" ones are the two Ubuntu partitions, and I spend most of my time in Debian Etch. This isn't because I haven't tried lots of distributions...I could see 10 different boxed versions before I cleaned off my shelves earlier this year, but they aren't the ones I stayed with.
I have a suspicion that companies are basically untrustworthy. Red Hat will probably remain (relatively) trustworthy just a long as Bob Young remains in control. Ubuntu as long as Shuttleworth stays in charge. So the only answer is to have a strong license. GPL3 may suffice. Something that means that we WILL continue to be able to fork the code base, as it's only by frequent forks that FOSS can remain viable, and companies be prevented from being obnoxiously predatory.
Step 10. (You did say "Quick frozen", and that requires a blizzard of sub-Siberian temperature.)
OTOH, if you just mean the Thames freezing in autumn and thawing late in spring, that's step "The great conveyor shuts down".
Still, do remember that I said I didn't know how long any step took. I can't rule out several of them happening within the same year. (OTOH, the ocean temperatures change rather slower than air and land temperatures. Allow at least a few decades for most of these. The great conveyor shutting down would likely be due to Greenland shedding lots of fresh water, which would decrease the salinity of the surface, resulting in the surface water not being denser than the deeper water even when it was colder. [So floating buoys measuring temperature would be much use. Instead they need to measure BOTH temperature and salinity.])
N.B.: Current results indicate that ice sheets can "melt" a lot faster than previously supposed, due to surface water tunneling down to the base rock and lubricating the ice flow. I haven't heard of any good current estimates of the maximum expectable rate of "ice shedding" that include this effect in the models. I also don't know just how much ice needs to be shed the shut down the great conveyor. It's happened before, however. (But WE weren't around then, so we didn't care.) The last time it's reported to have happened was towards the end of the last ice age...but then the ocean was already cool, and so water wasn't being evaporated.
P.S.: If I'm remembering things correctly, the great conveyor shutting down would, indeed, make Europe VERY cold, but since the ocean wouldn't be circulating warm water up along the coast there wouldn't be much precipitation. Thus you wouldn't get an ice age, because the winter would be too dry. Perhaps summer would be also...in which case Europe would, instead, turn into a cold desert...think steppes. (Why do the models focus on Europe instead of the US? My guess: Europe is simpler to model.) If this should be correct, then could I point you to the recent development of a wave powered desalinization device? It's not yet ready for engineering trials, but perhaps someone should invest in it? (Sorry, I forget the source. Check New Scientist for the last couple of months.)
Do you know that it's easier to predict the seasons than the weather?
......
Why should you be able to predict the seasons when you can't tell whether it will rain tomorrow?
I'm sorry, but your argument "Why wasn't 2004 worse than 2003 which was worse than 2002 and so on? " appears silly. Your general argument would be worth addressing, except that your tone implies you won't understand the replies.
However, I consider planetary albedo (generally ice coverage) more significant than hurricanes. Frequent and intense hurricanes are a predicted end result from certain weather models. Ice is directly measurable, and directly causative.
Also, global changes are more significant than local changes. Rising sea levels are more significant that melting glaciers. One may contribute in a minor way to the other.
P.S.: "The Day After Tomorrow" was based on an absurd story line, and it compressed (hopefully) centuries worth of change into a year or less, but the basic concept is reasonable. (Except that Mexico would welcome folk from the US in such an event.) It bears as much relation to science as Shakespeare Histories bear to history. I.e., there are several points of match-up, but it's been adapted to suit supporters more interested in political propaganda than facts.
P.P.S.: Also there's currently a question of whether the shutdown of the great conveyor would actually cause an ice age in Europe. AFAIK this hasn't yet been decided. But at the time that the movie was made it was the orthodox belief, where now it's disputed. (I haven't heard any report on the effect on the US. One can presume that this is also in dispute.)
N.B.: IF the shutdown of the great conveyor leads to an ice age, then a "big melt" can be expected to be the normal prequel to a new ice age. The mechanism would work sort of like this:
1) Temperatures rise
2) Ice melts
3) The earth starts absorbing more sunlight, causing temperatures to rise more quickly.
4) Enough ice melts to shut down the great conveyor. (Say, much of Greenland.)
5) The great conveyor shuts down.
6) The poles freeze and the equator heats up.
7) The new ice on the poles reflects lots of sunlight, so the earth cools down, but the oceans are still warm, so there's lots of moisture evaporating.
8) Lots of precipitation, cloud coverage increases. In the north snow starts piling up.
9) More clouds & snow means more sunlight reflected. Things cool more.
10) The glaciers begin walking south.
11) Finally the oceans cool off. There's now less precipitation.
12) Cloud coverage decreases, but there's all that new snow...so things keep getting cooler for awhile.
I don't know what's supposed to happen here. Something eventually causes the glaciers to retreat.
Note that an important part of this cycle is the thermal mass of the oceans. When they are warm, they evaporate a lot of water. When they are cool, the amount of evaporation is a lot less.
N.B.: I don't say anything here about how long any particular step lasts. I don't know. Some indications seem to show that the glaciers may begin a rapid walk south in as little as one year. (Quick frozen Mammoths with tender green vegetation still in their stomachs, e.g. Perhaps there are other explanations.)
This think reads like it was scripted by the PR department.
Also, I notice that they had things rigged so that they could censor any questions they didn't like. (Reasonable, an open forum would have been a mad house, but not exactly a process that builds trust.)
They also didn't say anything about which of their customers could redistribute what. The short answer appears to be "We aren't interested in developers."
Could you please specify?
I'll accept that GNU/Linux is "non-sense". (It isn't really, but it's a bad idea because short words tend to be more acceptable for high usage.) So I mean something besides that.
Nothing that we know of obligate Novell to publish MS IP. But they've just bought insurance that allows them to do so without worrying about getting sued. Also there are many parts of the deal that have not been made public.
There is no *proof* that we have any more to worry about than before. But given patent law, if you wait for proof your business will be toast. The question is "Can you trust them", and they've just bought insurance that allows them to grossly violate your trust without worrying about getting sued. It's clearly not proof...but it removes most grounds for trusting them. And if you don't trust them, then public statements that aren't legally binding are worthless.
They do have valid reasons for declaring the contract void. Both sides have publicly declared that there has been "no meeting of minds". But if they were to void the contract, then they would need to return the money, so they don't want to.
And they it is proper that they be despised and turned into pariahs. They should be seen as the carriers of a loathsome infectious disease, and folk who value their own well-being should refuse to approach them, or to accept merchandise from them.
The thing that bothers me about Gnome is their lack of a decent menu editor. This appears trivial, but I find it so annoying that I've never looked much further.
This isn't quite true, I used Gnome by preference back around KDE1...but not since they removed the menu editor.
I believe that BSD uses the gcc. I could be mistaken. as I'm not really that well acquainted with it.
But note that Darwin uses the gcc under the GPL, and doesn't try to impose non-GPL acceptable conditions on the use of GPL code. Novell doesn't appear to have that option.
I don't think that's going to be optional. Rejecting code from Novell is going to be a necessary defensive act. I've already stopped accepting updates from them to old software (so I'm looking for alternatives when the applications are necessary...fortunately I'd only just started investigating Mono, as I'd been worried about IP of SOME [undetermined] nature all along).
If Novell has been making significant contributions, then this will be a sad loss. But necessary.
If you think this is purely an emotional argument, then you have read the publicly released portions of the agreement. Or you haven't analyzed what actions could be taken on the basis of those agreements.
"IBM thinks its a good agreement"? This bothers me, and if the arguments presented made sense, then I'd re-think my position. When I read them, though, they don't make sense unless you're a large company that is already maintaining a stable of lawyers. I can clearly see why the agreement wouldn't significantly bother or threaten IBM, but IBM's situation is not at all similar to that of the average FOSS developer.
You must not be an experienced programmer. ALL code develops incompatibilities and bugs appear as the environment changes. Only "short-time" projects can even appear stable. I've had code last for five years. It wasn't doing much that was environment facing.
Well, I suppose that if it's simple enough.... "hello, world" still works, and that's been decades, but once things start getting at all complex things change.
Didn't MS get a patent on adding two numbers together in basic? I think Gambas violates that.
And to cap the joke, NTP doesn't appear to have had a valid patent to complain about. That must have left all their lawyers rolling on the floor.
You've got the point, but you don't understand it.
My wife teaches privately, and can select here students. She has noticed that when the mother is working outside the home, the kids are less developed. Her belief (based on evidence, but I don't know it all) is that when the mother's come home they are too tired to do a decent job of being a mother, also. There's a lot of variation here, depending partially on what job the mother has, and what shape she's in, and, of course, on both her skill level, and the willingness of the child to be trained. But usually she can work with a child awhile and tell whether or not his mother is employed outside the home. (This is not a fair test, because she usually already knows, but she believes that this information tends to be expressed in the child's behavior.)
So... what's a fair answer, that doesn't have a cost to society that isn't excessively high?
P.S.: Wages have fallen since the 1980's, relative to prices, so that now it's usually mandatory for both parents to be employed. The two salary family now has no more money than the one salary family did in the 1980's, when measured by value of residual discretionary funds.
OK, I'm no expert, and my information is 2-3 decades old...however...
I have this strange suspicion that the stuff they enter into cell-phones is the rough equivalent to "texting" (or whatever that set of peculiar abbreviations is called).