Ah, I see that you are from the CS department at CMU. As it seems you are unaware let me relate to you a fundamental flaw in the general social environment of game development: game developers are not computing scientists. For the most part those constituting 'old hat' developers are self-taught and have developed a chauvinistic narrow-minded view of that which is pragmatic and applicable knowledge to game development. Eg, "You may have a Ph.D. in Physics but without being a C++ guru I feel I can belittle and admonish you for even presuming to enter my field." (Disclaimer: I am not a physicist).
How is that narrow-minded and chauvinistic? It sounds practical and eminently reasonable to me. If you want to do X, and you're not very good at X, others will make fun of you, no matter how many PhDs you have in other subjects.
A PhD is actually a very specialized qualification for a very, very particular field of research, like the high energy physics of certain particles. It's necessary to have if you want to be an important person in that field, but it won't help you in most other fields. Sad but true. It also doesn't mean you have common sense or the ability to work well on a team.
In my experience Game Developers are fair programmers but mathmeticians, physicists, and computing scientists they are not. The curiosity present in graduate researchers is absent from much of the lot.
Is it really absent, or is it just present about different topics? Perhaps game developers are interested mostly in things like gameplay. The topics you mentioned are more interesting to mathematicians and applied mathematicians. If you want to be a mathematician, you know where to go. (Back to college.)
I simply wish the elitism would dissipate and the field could learn to accept itself for what it is: a commercial, industrial endeavour to meet a target market demand in order to produce sustainable profit margins. It is not amenable to creative computing research, but merely a consumer of the output of said research.
I guess I haven't experienced any of the elitism you talk about. Then again, I've never worked in the game industry. I hear those people work long hours.
It's probably doable if you 1) Make the system ground based 2) Use lasers, which are (in practical terms) instantaneous, rather than trying to hit one moving object with another
But here's the thing... you can do all sorts of stuff that you essentially can't nicely without the ADS, because you can store file+metadata together *without changing the file format*.
In some sense, a file format is a contract between various developers and end users. It provides certain capabilities, and imposes certain limits, on what can be stored in the file.
If you can't accept those limits, you have to move to a different file format, which can accommodate your needs. Then your tools will need to be updated. For example, if you want to add a field called "cuteness" to your jpg files, all of your apps will have to be updated to know what "cuteness" means-- starting with image viewers and moving on down to basic things like "ls."
Streams can't change this. In order to work sensibly, your apps will still have to be rewritten each time you add a new stream to the file. Either that, or the apps will get out of date, and only be able to correctly understand part of the file. Neither of those fates seem really that desirable.
Streams are the wrong abstraction. The right abstraction is storing a separate file to house the metadata. If you are worried about the file getting separated from the metadata, create a new directory with just those two files in it. You could type "mkdir" all day and not run out of space on one of those shiny new 250 GB drives.
Putting the metadata in other files would be somewhat of a solution, but that has its own problems, because it's easy to get the files separated or out of sync, and they sit there cluttering the directory.
How is it any easier to get the separate file out of sync or out of date than it is to get the metadata stream out of sync? If anything, the separate file solution makes the situation clearer, because whenever you update the "main file" with some tool that doesn't support your metadata, it will be obvious that the modification times of the metadata file and the main file are out of sync.
I agree. File streams, or resource forks, or whatever you want to call them, just make things more confusing, without providing any real benefit.
The only benefit I can really think of is that the user doesn't have to see so many files cluttering up the place, because multiple files have been merged into one "stream." But Windows users are not supposed to be looking at those files anyway (Windows whines at you if you open "Program Files"), so who cares?
Don't be dense. Denying countries access to the sea is an old and time-honored tactic. It's called a blockade, and the US has used it many times. Our fleet of aircraft carriers effectively exerts control over the sea.
Blubbering on about how we would never, ever infringe on another countries right to access the sea or space is just ridiculous. That's what we have these fleets and satellites for-- to exert control in, and gain information about, distant lands. And no, of course we don't deny Britain or France access to anything. That's because they're our allies.
China is not an ally. It's rapidly becoming something the U.S. has not had for a while: a peer. This may not be entirely a good thing for the world, but it's coming. India, too, will soon be very important. It's only natural that these countries would come up with weapons to counter the U.S. push towards domination of space.
This laser is a very Chinese weapon. It's purely defensive, to shoot down enemy spy satellites or the hypothetical "star wars" satellites Reagan was so fond of pushing. Another way of trying to protect China from the outside world, like the Great Wall of times past.
Personally, I think we should be investing more into weapons of this sort. Perhaps a laser like this could be used against ICBMs. Instead, though, we will probably sink more billions into 1984-esque surveillance technologies and killer robots.
It would be hilarious if the next Indiana Jones movie WAS about Dr. Jones being a crotchety, washed-up academic. They could spend an hour and a half describing his latest ethnographic study.
"Indiana Jones and the thought-provoking statistical anomaly in pre-mayan econometrics"
Ok. I know that this issue has been discussed to death, and it's hard for people to think rationally about it. But bear with me for a moment.
Sadam killed everyone in Iraq...everyone that stood against him. Terrorists, Muslims, Christians, football (soccer) players, left handed people, right handed people, . So what exactly is your point?
Saddam Hussein was a secular leader who kept a lid on the power of the religious radicals in the country. He fought a war against the Islamic fundamentalist government of Iran. During this time, he received funding and military backing from the U.S. government.
Whatever other terrible things he did, Saddam was NOT a terrorist or Islamic fundamentalist. That's the point.
Bush never said Iraq was behind 9-11, he said that Iraq had weapons of mass distruction like 50 million times and said places like Iraq were part of the Axis of Evil (Iraq, Iran, and North Korea).
Bush lumped together many unrelated countries in his "axis of evil." The Big Lie of the Bush administration is the idea that the war in Iraq somehow advances the cause of national security. Do you remember Bush repeating this lie over and over throughout the last 7 years?
It's become obvious even to the average voter just how silly it is to keep pissing away resources in Iraq. I think the people who opposed Iraq from the beginning deserve at least an "I told you so." Too bad we can't get that trillion dollars back, though.
If you want to know why Bush really brought us into Iraq, read up on "The Project for a New American Century." This is the stuff that makes some people hysterial-- the "neoconservative" ideology. Generally speaking, neocons favor U.S. military hegemony, free trade, and the establishment of western-style republics throughout the world.
Ironically, the US is the country with the most to lose in any war, because we have the most money and the most political power. By overextending ourselves, we have actually weakened our position across the board. Try as they might, China, Afghanistan, or Iraq could never have done the United States a more grievous hurt than G.W. Bush has in the last decade.
What I'm saying (and have been saying) is that dealing in bandwidth with bytes has a lot more ambiguity than dealing with bandwidth in bits...
Um... divide by 8? Don't let those EE courses go to your head... not all math is complex.
The irony is, bytes are actually more standardized than prefixes like "m," "k,", or billion. Remember that there is, or used to be at least, an "english million" and "english billion" that were NOT 10^6 and 10^9. And there's sometimes confusion between mega- and milli-, although this is usually resolved by the context.
The real reason electrical engineers don't measure things in bytes is because, in the past at least, they didn't care about the software side of things. So there was no need to group things into bytes. Now that everyone and his dog has a computer, and most EEs have computers on their desks, it would probably be best if line speeds were in bytes per second. But in an industry where L2TPV2 and BFD are viewed as legitimate acronyms, I wouldn't expect much!
Here's a summary of this decision for dumbass politicos:
1) AT&T can kick the cable companies in the nuts 2) Cable companies can't kick back
In most cases, both telephone service and cable are monopolies, so all the free-market spew that's going to come out in the next few comments is irrelevant.
mySql cheats on a lot of database guarantees, and didn't handle views correctly for a long time. postgresql is much more solid.
That being said, I think the choice of language is much more important than the choice of database. Whatever you do, don't use PHP-- just "training wheels without the bike," as the Perl people like to say.
Um, maybe the slight "element of unfairness" was the fact that Microsoft was giving away their product for free, hence destroying their business model? Common sense, have you heard of it?
PS. I agree with you that the gaming industry isn't the best place to work at the moment. Not sure why that is, but many in the industry end up working 80-hour weeks or ridiculous stunts like that.
Kind of ironic considering that Electronic Arts was founded by a group of programmers unhappy with the profit-sharing setup at Atari. Oh well-- all flows acording to the whims of the Great Magnet, as Hunter S. would say.
Computing can be a difficult field. In most industries, change comes gradually, but in the computer field, every year brings some new surprises.
I understand your point about sysadmins, although I'm too young to have personal experience with it. When I was in college not too long ago, IT was a different, more practical degree than computer science. The specialization sucks in a way, because it means that good programmers might never have a chance to know what it's like to be a sysadmin. Their range of experience will be that much smaller. But it is inevitable as the industry matures.
New specialties create opportunities as well as problems. Who could have even understood what a microprocessor designer or compiler expert was in 1950? These disciplines have been created out of thin air. Some other disciplines may vanish into thin air eventually, just like professional woodworking or professional fly fishing vanished. You just have to keep an open mind. Progress is the hallmark of life-- it IS creativity itself. You just have to ride the wave while it's peaking, and have fun with it.
I don't even know where to start with this post... it is so wrong-headed. First of all, what is "pure engineering"? No form of engineering is really pure, only science can be that. There was a point when software engineering split from electrical engineering... maybe that is what you're thinking of?
Secondly, your three phases of job evolution is complete bullshit. No reputable economist believes that. Every industry is different, and trying to split things into arbitrary phases is boneheaded. And what's with comparing the computer industry to the garment industry? Next thing you know, you'll be telling me that the internet is like a series of tubes.
Finally... the hardware of a decade or two from now will be very different from today's. I don't doubt that there will be a market for new code that pushes those hardware limits. Just because you can't imagine any levels of abstraction beyond they ones we have now, doesn't mean they aren't waiting to be created. And there will be jobs.
We are very close to the point where no group or team will be able to understand the entire code for a major program... We are, right now, at the point where the code for major programs can't be understood by any single individual (just like microprocessor design, in fact).
"X cannot be understood by a single individual" is such a meaningless statement. There are people who understand the high-level organization of a microprocessor. They draw charts with boxes and arrows, run simulations and optimization programs, and get paid a whole lot of money. Generally they have PhDs in computer architecture and work for Intel or AMD, or similar company. There are people who understand the high-level organization of the Linux kernel, or other major software packages. Division of labor-- what is the big deal?
The result is that we'll create compilers that create code to be fed into compilers, which create lower-level code, which will be fed into compilers which create machine code... They already have that. It's called Java or C#, in which code is compiled into bytecode, and then the JVM compiles it into machine code. Or shell scripts that generate other scripts, like autoconf. And someone needs to write those JVMs, the operating systems that they run on, the compilers that they use, and the code that they execute. The sky is not falling-- there will always be interesting work for computer people.
Eventually, the "programmer's code" will be no more than a concise description of what the program should do. (This is a good thing, by the way.) At that point, programming as a field of progressive science will dwindle, and become a technical trade, like riveting or welding or house framing. The programmer's code was always "no more than a concise description of what the program should do." Some people were more concise than others, but then again, some riveters were better at riveting than others. Programming was never a science-- that's why it's called software engineering. In academia, computer science is mostly a melange of pure mathematics (theorical CS), psychology (HCI), and statistics (artificial intelligence).
The field is definitely becoming more specialized. "Information technology" has emerged as a separate field for network administrators and help desk people. As long as people have help desk questions, and as long as software can be misconfigured or hardware can break down, they will have jobs, although maybe not glamorous ones. Web programming is emerging as a field. It's hot now, but the really interesting thing for me is what's behind the shiny web interfaces. Game programming has had its own industrial revolution, several of them in fact, and may not be a good place to work for most people (several other posters have commented on this.) Fewer people have to deal with low-level programming, but it's still around if you want it. Projects are getting larger. Hardware design is becoming more like software. Some things are being outsourced. But that's not news, unless you live in the 1970s. Which one of these trends signals "the end of computing" for you?
I'm going to guess that you're a young, idealistic kid at university. You have very little experience with the industry or its history, and you don't realize that all of your "big ideas" and predictions have been thought of and predicted already. A lot of bright people have thought about these topics, from Alan Turing and Herbert Simon, to more practical folks like Sergey Brin and Bill Gates. Before you sound off, how about reading what they have to say?
Computers generating code? Get the president of IBM on the line! We'll call the code-generating programs "Compilers"
Self-administering systems? I know! We can create a program called "cron" that runs jobs at periodic intervals! Get the president of Bell labs on the line!
Or maybe, just maybe, we can't even look five years ahead, let alone twenty. And given the pace of technology, the idea of any field of engineering going away in our lifetime is boneheaded. If anything, there's going to be a lot more computer- and science-related fields in a decade.
... understand that a Master's degree is for masters, not for slaves (employees)...
I LOLed.
I do not want people who like being led, I want to get other self-starters and leaders collaborating with me (with profit sharing of course). I would prefer a hacker...
Ok, these are all good points, but have you ever actually hired someone?
If you had, I think you would realize that it takes all kinds of people to make a business successful. Some people are really ambitious, some are not.
All engineering is a tradeoff. In the case of UTF-8 versus UTF-16, UTF-32, Shift/JIS, etc., the tradeoffs mostly involve compatibility, space, and speed.
As you've noticed, for non-English languages, UTF-32 often takes up less space in memory than UTF-8. However, converting applications to use UTF-32 usually requires an expensive rewrite. Also, UTF-16 and UTF-32 have endianness issues. Basically, in order to achieve real interoperability, you must either mandate that one endianness or the other be used (which slows down the CPU if it has the "wrong" endianness), or tell programmers to plop in conversions in wherever they feel appropriate (compatibility hell, BOMs.)
UTF-8 sidesteps all of those endianness issues by being a byte-oriented standard. It also allows existing C code to work well with foreign text. You may look down your nose at C code, and admittedly some of it is bad, but it is not going away any time soon. Both the Linux kernel and the Windows kernel are written in C, and there are no plans to change! Keep in mind that the computer industry is dominated by compatibility concerns. There are still people hacking FORTAN or COBOL code from the 70s. Telling people they need to throw away their own code is like telling people they need to stop using gasoline engines and start using your new, slightly better petroleum fuel. People will nod, smile and keep on doing what they were doing. Sorry, Charlie, it's called installed base!
UTF-8's only disadvantage, the space issue, is overhyped anyway. It's very likely that Asian or Eastern European text encoded in UTF-8 will compress very well with gzip or another free compression algorithm. Also, English text, the common case for most people, does better with UTF-8 than under any other encoding! No matter how you feel about England or the United States, you cannot deny that English has become the new lingua franca of commerce, and most organizations support it in one way or another.
Also, text is a very small part of the total size of most applications. If every string constant in Linux started taking double the size, would anyone even notice? 18 months later, hard disk sizes and memory densities would double, and it would be a big yawn for everyone involved. On the other hand, rewriting every piece of software from the ground up, which is basically what you're proposing with UTF-32, is an enormous and ongoing cost.
As for your comments about UNIX... those who don't understand UNIX are doomed to reimplement it, poorly.
More people died in just one day of World War I (aka the Great War) than on all of 9/11. By the way, world war I went on for 2 years. And then another two decades later, World War II was even worse.
Yup, sure is a mystery why the United States took the technological and economic lead after WWII. They must be a special people superior to all others.
Slashdot Mirror
Ah, I see that you are from the CS department at CMU. As it seems you are unaware let me relate to you a fundamental flaw in the general social environment of game development: game developers are not computing scientists. For the most part those constituting 'old hat' developers are self-taught and have developed a chauvinistic narrow-minded view of that which is pragmatic and applicable knowledge to game development. Eg, "You may have a Ph.D. in Physics but without being a C++ guru I feel I can belittle and admonish you for even presuming to enter my field." (Disclaimer: I am not a physicist).
How is that narrow-minded and chauvinistic? It sounds practical and eminently reasonable to me.
If you want to do X, and you're not very good at X, others will make fun of you, no matter how many PhDs you have in other subjects.
A PhD is actually a very specialized qualification for a very, very particular field of research, like the high energy physics of certain particles. It's necessary to have if you want to be an important person in that field, but it won't help you in most other fields. Sad but true. It also doesn't mean you have common sense or the ability to work well on a team.
In my experience Game Developers are fair programmers but mathmeticians, physicists, and computing scientists they are not. The curiosity present in graduate researchers is absent from much of the lot.
Is it really absent, or is it just present about different topics? Perhaps game developers are interested mostly in things like gameplay.
The topics you mentioned are more interesting to mathematicians and applied mathematicians. If you want to be a mathematician, you know where to go. (Back to college.)
I simply wish the elitism would dissipate and the field could learn to accept itself for what it is: a commercial, industrial endeavour to meet a target market demand in order to produce sustainable profit margins. It is not amenable to creative computing research, but merely a consumer of the output of said research.
I guess I haven't experienced any of the elitism you talk about. Then again, I've never worked in the game industry. I hear those people work long hours.
It's probably doable if you
1) Make the system ground based
2) Use lasers, which are (in practical terms) instantaneous, rather than trying to hit one moving object with another
Then again, I'm not a mad scientist.
Is there ANY non-experimental filesystem that runs on a standard desktop OS that has this feature?
Waybackfs on
Linux.
But here's the thing... you can do all sorts of stuff that you essentially can't nicely without the ADS, because you can store file+metadata together *without changing the file format*.
In some sense, a file format is a contract between various developers and end users. It provides certain capabilities, and imposes certain limits, on what can be stored in the file.
If you can't accept those limits, you have to move to a different file format, which can accommodate your needs. Then your tools will need to be updated. For example, if you want to add a field called "cuteness" to your jpg files, all of your apps will have to be updated to know what "cuteness" means-- starting with image viewers and moving on down to basic things like "ls."
Streams can't change this. In order to work sensibly, your apps will still have to be rewritten each time you add a new stream to the file. Either that, or the apps will get out of date, and only be able to correctly understand part of the file. Neither of those fates seem really that desirable.
Streams are the wrong abstraction. The right abstraction is storing a separate file to house the metadata. If you are worried about the file getting separated from the metadata, create a new directory with just those two files in it. You could type "mkdir" all day and not run out of space on one of those shiny new 250 GB drives.
Putting the metadata in other files would be somewhat of a solution, but that has its own problems, because it's easy to get the files separated or out of sync, and they sit there cluttering the directory.
How is it any easier to get the separate file out of sync or out of date than it is to get the metadata stream out of sync? If anything, the separate file solution makes the situation clearer, because whenever you update the "main file" with some tool that doesn't support your metadata, it will be obvious that the modification times of the metadata file and the main file are out of sync.
I agree. File streams, or resource forks, or whatever you want to call them, just make things more confusing, without providing any real benefit.
The only benefit I can really think of is that the user doesn't have to see so many files cluttering up the place, because multiple files have been merged into one "stream." But Windows users are not supposed to be looking at those files anyway (Windows whines at you if you open "Program Files"), so who cares?
Don't be dense. Denying countries access to the sea is an old and time-honored tactic. It's called a blockade, and the US has used it many times.
Our fleet of aircraft carriers effectively exerts control over the sea.
Blubbering on about how we would never, ever infringe on another countries right to access the sea or space is just ridiculous. That's what we have these fleets and satellites for-- to exert control in, and gain information about, distant lands. And no, of course we don't deny Britain or France access to anything. That's because they're our allies.
China is not an ally. It's rapidly becoming something the U.S. has not had for a while: a peer. This may not be entirely a good thing for the world, but it's coming. India, too, will soon be very important. It's only natural that these countries would come up with weapons to counter the U.S. push towards domination of space.
This laser is a very Chinese weapon. It's purely defensive, to shoot down enemy spy satellites or the hypothetical "star wars" satellites Reagan was so fond of pushing. Another way of trying to protect China from the outside world, like the Great Wall of times past.
Personally, I think we should be investing more into weapons of this sort. Perhaps a laser like this could be used against ICBMs. Instead, though, we will probably sink more billions into 1984-esque surveillance technologies and killer robots.
Torvalds is a brilliant programmer and administrator. He is more than "slightly connected" to the issue of what license to use for the kernel.
I think you have been misled by RMS's fine-sounding speeches and writings. Think for yourself.
Here's a hint: the GPLv3 can't stop DRM or software patents.
Here's another hint: the GPLv3 CAN cripple projects that adopt it.
It would be hilarious if the next Indiana Jones movie WAS about Dr. Jones being a crotchety, washed-up academic.
They could spend an hour and a half describing his latest ethnographic study.
"Indiana Jones and the thought-provoking statistical anomaly in pre-mayan econometrics"
Wow. Take a deep breath, comrade.
1% of 53 million dollars = 530,000
66% of 530,000 = 331,980 after tax
That's enough to buy a house, but hardly more than that.
Yes, Virginia, extreme income inequality is a bad thing. But so are frothy rants by people who should know better by now.
Ok. I know that this issue has been discussed to death, and it's hard for people to think rationally about it. But bear with me for a moment.
Sadam killed everyone in Iraq...everyone that stood against him. Terrorists, Muslims, Christians, football (soccer) players, left handed people, right handed people, . So what exactly is your point?
Saddam Hussein was a secular leader who kept a lid on the power of the religious radicals in the country. He fought a war against the Islamic fundamentalist government of Iran. During this time, he received funding and military backing from the U.S. government.
Whatever other terrible things he did, Saddam was NOT a terrorist or Islamic fundamentalist. That's the point.
Bush never said Iraq was behind 9-11, he said that Iraq had weapons of mass distruction like 50 million times and said places like Iraq were part of the Axis of Evil (Iraq, Iran, and North Korea).
According to president Bush himself, 9/11 was the justification for attacking Iraq. (Thanks to theskipper for finding this.)
Bush lumped together many unrelated countries in his "axis of evil." The Big Lie of the Bush administration is the idea that the war in Iraq somehow advances the cause of national security. Do you remember Bush repeating this lie over and over throughout the last 7 years?
It's become obvious even to the average voter just how silly it is to keep pissing away resources in Iraq. I think the people who opposed Iraq from the beginning deserve at least an "I told you so." Too bad we can't get that trillion dollars back, though.
If you want to know why Bush really brought us into Iraq, read up on "The Project for a New American Century." This is the stuff that makes some people hysterial-- the "neoconservative" ideology. Generally speaking, neocons favor U.S. military hegemony, free trade, and the establishment of western-style republics throughout the world.
Ironically, the US is the country with the most to lose in any war, because we have the most money and the most political power. By overextending ourselves, we have actually weakened our position across the board. Try as they might, China, Afghanistan, or Iraq could never have done the United States a more grievous hurt than G.W. Bush has in the last decade.
What I'm saying (and have been saying) is that dealing in bandwidth with bytes has a lot more ambiguity than dealing with bandwidth in bits...
Um... divide by 8?
Don't let those EE courses go to your head... not all math is complex.
The irony is, bytes are actually more standardized than prefixes like "m," "k,", or billion.
Remember that there is, or used to be at least, an "english million" and "english billion" that were NOT 10^6 and 10^9. And there's sometimes confusion between mega- and milli-, although this is usually resolved by the context.
The real reason electrical engineers don't measure things in bytes is because, in the past at least, they didn't care about the software side of things. So there was no need to group things into bytes. Now that everyone and his dog has a computer, and most EEs have computers on their desks, it would probably be best if line speeds were in bytes per second. But in an industry where L2TPV2 and BFD are viewed as legitimate acronyms, I wouldn't expect much!
Here's a summary of this decision for dumbass politicos:
1) AT&T can kick the cable companies in the nuts
2) Cable companies can't kick back
In most cases, both telephone service and cable are monopolies, so all the free-market spew that's going to come out in the next few comments is irrelevant.
mySql cheats on a lot of database guarantees, and didn't handle views correctly for a long time.
postgresql is much more solid.
That being said, I think the choice of language is much more important than the choice of database.
Whatever you do, don't use PHP-- just "training wheels without the bike," as the Perl people like to say.
Not only that, but just because someone told you that the range was clear doesn't mean it's clear 10 seconds later, when you finally take the shot.
This law is stupid. Mindless political correctness at its best.
Um, maybe the slight "element of unfairness" was the fact that Microsoft was giving away their product for free, hence destroying their business model?
Common sense, have you heard of it?
PS. I agree with you that the gaming industry isn't the best place to work at the moment. Not sure why that is, but many in the industry end up working 80-hour weeks or ridiculous stunts like that.
Kind of ironic considering that Electronic Arts was founded by a group of programmers unhappy with the profit-sharing setup at Atari.
Oh well-- all flows acording to the whims of the Great Magnet, as Hunter S. would say.
Computing can be a difficult field. In most industries, change comes gradually, but in the computer field, every year brings some new surprises.
I understand your point about sysadmins, although I'm too young to have personal experience with it. When I was in college not too long ago, IT was a different, more practical degree than computer science. The specialization sucks in a way, because it means that good programmers might never have a chance to know what it's like to be a sysadmin. Their range of experience will be that much smaller. But it is inevitable as the industry matures.
New specialties create opportunities as well as problems. Who could have even understood what a microprocessor designer or compiler expert was in 1950? These disciplines have been created out of thin air. Some other disciplines may vanish into thin air eventually, just like professional woodworking or professional fly fishing vanished. You just have to keep an open mind. Progress is the hallmark of life-- it IS creativity itself. You just have to ride the wave while it's peaking, and have fun with it.
I don't even know where to start with this post... it is so wrong-headed. First of all, what is "pure engineering"? No form of engineering is really pure, only science can be that. There was a point when software engineering split from electrical engineering... maybe that is what you're thinking of?
Secondly, your three phases of job evolution is complete bullshit. No reputable economist believes that. Every industry is different, and trying to split things into arbitrary phases is boneheaded. And what's with comparing the computer industry to the garment industry? Next thing you know, you'll be telling me that the internet is like a series of tubes.
Finally... the hardware of a decade or two from now will be very different from today's. I don't doubt that there will be a market for new code that pushes those hardware limits. Just because you can't imagine any levels of abstraction beyond they ones we have now, doesn't mean they aren't waiting to be created. And there will be jobs.
We are very close to the point where no group or team will be able to understand the entire code for a major program...
We are, right now, at the point where the code for major programs can't be understood by any single individual (just like microprocessor design, in fact).
"X cannot be understood by a single individual" is such a meaningless statement. There are people who understand the high-level organization of a microprocessor. They draw charts with boxes and arrows, run simulations and optimization programs, and get paid a whole lot of money. Generally they have PhDs in computer architecture and work for Intel or AMD, or similar company. There are people who understand the high-level organization of the Linux kernel, or other major software packages. Division of labor-- what is the big deal?
The result is that we'll create compilers that create code to be fed into compilers, which create lower-level code, which will be fed into compilers which create machine code...
They already have that. It's called Java or C#, in which code is compiled into bytecode, and then the JVM compiles it into machine code. Or shell scripts that generate other scripts, like autoconf. And someone needs to write those JVMs, the operating systems that they run on, the compilers that they use, and the code that they execute. The sky is not falling-- there will always be interesting work for computer people.
Eventually, the "programmer's code" will be no more than a concise description of what the program should do. (This is a good thing, by the way.) At that point, programming as a field of progressive science will dwindle, and become a technical trade, like riveting or welding or house framing.
The programmer's code was always "no more than a concise description of what the program should do." Some people were more concise than others, but then again, some riveters were better at riveting than others. Programming was never a science-- that's why it's called software engineering. In academia, computer science is mostly a melange of pure mathematics (theorical CS), psychology (HCI), and statistics (artificial intelligence).
The field is definitely becoming more specialized. "Information technology" has emerged as a separate field for network administrators and help desk people. As long as people have help desk questions, and as long as software can be misconfigured or hardware can break down, they will have jobs, although maybe not glamorous ones. Web programming is emerging as a field. It's hot now, but the really interesting thing for me is what's behind the shiny web interfaces. Game programming has had its own industrial revolution, several of them in fact, and may not be a good place to work for most people (several other posters have commented on this.) Fewer people have to deal with low-level programming, but it's still around if you want it. Projects are getting larger. Hardware design is becoming more like software. Some things are being outsourced. But that's not news, unless you live in the 1970s. Which one of these trends signals "the end of computing" for you?
I'm going to guess that you're a young, idealistic kid at university. You have very little experience with the industry or its history, and you don't realize that all of your "big ideas" and predictions have been thought of and predicted already. A lot of bright people have thought about these topics, from Alan Turing and Herbert Simon, to more practical folks like Sergey Brin and Bill Gates. Before you sound off, how about reading what they have to say?
Ooh, a "futurist."
Computers generating code? Get the president of IBM on the line! We'll call the code-generating programs "Compilers"
Self-administering systems? I know! We can create a program called "cron" that runs jobs at periodic intervals! Get the president of Bell labs on the line!
Or maybe, just maybe, we can't even look five years ahead, let alone twenty. And given the pace of technology, the idea of any field of engineering going away in our lifetime is boneheaded. If anything, there's going to be a lot more computer- and science-related fields in a decade.
... understand that a Master's degree is for masters, not for slaves (employees)...
I LOLed.
I do not want people who like being led, I want to get other self-starters and leaders collaborating with me (with profit sharing of course). I would prefer a hacker...
Ok, these are all good points, but have you ever actually hired someone?
If you had, I think you would realize that it takes all kinds of people to make a business successful. Some people are really ambitious, some are not.
All engineering is a tradeoff. In the case of UTF-8 versus UTF-16, UTF-32, Shift/JIS, etc., the tradeoffs mostly involve compatibility, space, and speed.
As you've noticed, for non-English languages, UTF-32 often takes up less space in memory than UTF-8. However, converting applications to use UTF-32 usually requires an expensive rewrite. Also, UTF-16 and UTF-32 have endianness issues. Basically, in order to achieve real interoperability, you must either mandate that one endianness or the other be used (which slows down the CPU if it has the "wrong" endianness), or tell programmers to plop in conversions in wherever they feel appropriate (compatibility hell, BOMs.)
UTF-8 sidesteps all of those endianness issues by being a byte-oriented standard. It also allows existing C code to work well with foreign text. You may look down your nose at C code, and admittedly some of it is bad, but it is not going away any time soon. Both the Linux kernel and the Windows kernel are written in C, and there are no plans to change! Keep in mind that the computer industry is dominated by compatibility concerns. There are still people hacking FORTAN or COBOL code from the 70s. Telling people they need to throw away their own code is like telling people they need to stop using gasoline engines and start using your new, slightly better petroleum fuel. People will nod, smile and keep on doing what they were doing. Sorry, Charlie, it's called installed base!
UTF-8's only disadvantage, the space issue, is overhyped anyway. It's very likely that Asian or Eastern European text encoded in UTF-8 will compress very well with gzip or another free compression algorithm. Also, English text, the common case for most people, does better with UTF-8 than under any other encoding! No matter how you feel about England or the United States, you cannot deny that English has become the new lingua franca of commerce, and most organizations support it in one way or another.
Also, text is a very small part of the total size of most applications. If every string constant in Linux started taking double the size, would anyone even notice? 18 months later, hard disk sizes and memory densities would double, and it would be a big yawn for everyone involved. On the other hand, rewriting every piece of software from the ground up, which is basically what you're proposing with UTF-32, is an enormous and ongoing cost.
As for your comments about UNIX... those who don't understand UNIX are doomed to reimplement it, poorly.
Death Race 2000!
I thought I was the only person to have seen that bizarre movie!
personal gain != collective gain
Not that I'm agreeing with colonialism.
More people died in just one day of World War I (aka the Great War) than on all of 9/11.
By the way, world war I went on for 2 years. And then another two decades later, World War II was even worse.
Yup, sure is a mystery why the United States took the technological and economic lead after WWII. They must be a special people superior to all others.