When you are a halfway decent programmer and do it professionally, you will come across a wide range of projects. And what happens wich each new project? In a matter of days or 1-2 weeks you gotta acquire the knowledge of your new problem domain. And that is very often a problem domain others spend years to study and get a degree, yet as a software engineer, you gotta be able to understand the problem domain good enough to automate it. T means you must be able to understand all the nomenclature and the structure, you must be able to spot even the subtlest problems, you must know the pitfalls and special cases, you must be able to talk to the experts in their domain on somewhat equal footing. And if you wanna be any good, you must understand the new problem domain enough to generalize it and often to make it accessible to non-experts.
And that is only one half of the job. The other half is the broad field of computer science... which is the prerequisite for all this.
In my life as a programmer I have had to learn the termodynamics and streaming physics of internal combustion engines, the bioengineering for bio-processing tanks in pharmaceutical production. Dentist medicine. Law. Economics. Hard body physics and so much more. And about each year one or two neew things get added to the list.
Sure, I'm no expert in either of those fields. But I know (or knew) enough to find my way around by myself for 90% of the time, and to know when I should ask the expert. And the fact remains... it is your bread and butter to learn things in a matter of days others take and need years to learn.
If you make design and usability your main concern in the microprocessor business, you can't make x86 chips anymore.
Re:The GPL: Intellectual Theft
on
GPLv2 Vs. GPLv3
·
· Score: 1
Wrong. You can only use the hard work of other people if they use YOUR hard work first. Or if they license it in a GPL compatible license. In both cases I don't see how this could restrict any freedoms for others they haven't used first for themselves.
I'm aware of every single a fact and argument you brought up In fact, I have written assembly code for multiple POWER based Proessors, of which PowerPC is one implementation variant, no production code though, just for excecise, and I have written assembly code for each x86 iteration, including AMD-64 (which actually was production code). Still doesn't change anything. After NT Windows wasn'T shipped for anything but x86 (and a beta version for itanium). Fact is, you can't get rid of it, because of inertia (or momentum, as you describe it). There is no single technical reason for having it, in fact, I know for fact that the dominance of x86 architecture prevented a lot of innovation in software. Read some papers on the development of the JavaVM, and how many less than ideal decisions were made because of x86, or for the L4 kernel or the CoyotOS kernels. The ressources that were put in keeping x86 running are few orders of magnitude greater than the ressources put into all other architecures combined. That's not really cost efficiency I think. And the fact that Apple Laptop battery life is less than half of what it was before the intel switch, despite better battery technology, better chipset integration, better LCD efficency and all those advances should give you a reason to think twice.
There are tremendous costs bound to using x86. It's just that those aren't costs easily to understand by those that buy all those machines. You cannot imagine he amount of work and energy (unproductive work and energy basically) that has been put and still is put into keeping the MS/intel combination running. I have experienced some of it. And only scratched the surface.
The x86 dominance is basically a result of two crooked architecure holding each other up: if MS DOS wasn't so crappy that it depends on x86 then the processor could be changed. If x86 wasn't too crappy to properly emulate it, then MS DOS or it's successors could be changed. As it is, we are stuck with both, because noone wants to change both at the same time, and you cannot really change each independently. There is something I hope for: Vista Tanks mightly, OS X and it's successors become the dominant OS in 10 years. Those are instructions set agnostic, and have been proven to be able to run on multiple platforms with pretty little effort, and does so atm run on 3 different instruction sets: x86, POWER, and the iPhone on ARM. Linux runs everywhere too. And as soon as you have this, there is no reason not to drop the most expensive to develop for and least effcient architecture. But as long as people still use MS Operating systems, we will be stuck with x86 and have to pay the price... energy price.
Only way to set an industry standard is, to get so fast so big in a new market/technology that everybody has to follow. Problem is, when you get so big so fast, there are almost neccessarily major flaws in the designs. Problem is, you never get rid of them again.
Exactly... I hear that reiterated quite for quite some time already. To really get things going there, somehow it has to be figured out, which parts of the code are covered by 3rd party patents, and which 3rd parties own those patents.
Then we know where to go to and who to spam, or wether it is feasible to just drop those parts of the code, or wether to implement open alternatives. Maybe those parts are even some fraud patents, something like matrix multiplication or whatever.
If the governement system he deploys will ba anything close to the operating system he deploys, it will be a free for all for all the vile elements to spy on you. And it will make you used to substandard availability of basic services. And how a major crash every few days? And the famous tax you gotta pay wether you use it or not? A lot of this si implemented already of course. But I'm sure embrace and extend will work better on the middle east than bomb to shreds and see what happens.
1. Controllers for other consoles are 40$ too 2. For other consoles the games cost 60$ 3. Without nunchuk, which isn't used for the multiplayer games I have seen so far, the Wii controller is 40$ too 4. You will buy 3 additional controllers at maximum for the Wii, most likely only one 5. You will definitely buy much more games for the Wii, which come at 40$ 6. If you don't buy more games for the Wii, you won't care about the additional controller too, or wouldn't buy a Wii in the first place
So, I can absolutel not understand how such a consumer friendly, and actually more affordable price decision by Nintendo can get flamed that much. But well, that's what most business decisions are made anyway: Look at the number you see immediately, compare greater than and lower than, and then take the lower number on buys and the higher number on sells. Any 5 year old can reason on that level. The scary thing is, that is how the world is run: with the reasoning of 5 year olds.
I don't know whwther you have read this review and benchmark comparison of the Old and the new PowerMac/MacPro.
See, on one side we have the 3 years old Dual Single-Core G5 with 2,5GHz. On the other side we have the newest Dual DualCore Xeon with 2,66GHz. That is 3 years advance in technology and manufactoring process (even a geberation generation difference, the G5 is 90nm and the Xeon is 65nm), double the cores cores and a neglectable 1% advantage in Clock. Yet the speed advanatge barely scratches 50% in it's best tests, most of the time the advantage is only between 20% and 30%. And for benovelence they haven't even mentioned Performance/Watt. For comparison, this is somewhat like comparing a dual Pentium to a single i486 and only barely beating it. So why is that? It's the architecture. Those are the things the average user can see when he looks carefully. The other thing is all the quircs and limitations (to name only a few: real mode, A20 gate, no execute flage etc, awkward paging mechanism, crufted, "baroque" instruction set) that an engineer has to work around to get it to work. That isn't immediately noticable to the end user. But what the end user notices is more glitches and unpredictable problems which are the results of not so clean designs. I'm only guessing, but I sort of have the impression that the well documented, but unheard of before problems with instability and and random shutdowns etc. are a result design not being so clean and managable as before.
I just hope this takeover won't hurt them in thier technology decisions. Because I want to see a Subnotebook with their MPC8641D their MRAM as a buffer and about 10-20GB Flash. Combined with an organic iode display or a high resolution monochrome titanium-oxide screen and a decent battery that would be the ultimate outdoor writing and coding machine. The only thing it won't be overly suited for is video and high end gaming, but still should be enough for most. A laptop with days of battery time...
Actually the Desktop is the only place Power architecture isn't used on a big scale anymore. Just as x86 isn't used anywhere in any significant amounts but on the desktop. This has reasons. The decision which processor to use is made by technicians who know what they are doing virtually everywhere but at the desktop. But the vast majority of desktop decisions are made by corporate managers, that are bought by Microsoft or don't know anything else, which means the have to use x86 for (backwards) compatibility, or by end users who don't know anything else. The only reason that x86 is still around is, that it is the only architecture Microsoft ever got their operating systems working on decently. Tells you a lot about code quality.
Well, I'm a debian user since 1998, and I also conributed half a dozen patches or so in that time. But I have to say, the only thing that ever caused me serious problems in using Debian (prevented me from doing what I wanted) was the Kernel. And that isn't Debian specific. Problems in the Kernel are prevalent in all distributions.
Hmmm... what is it when everybody in a project but one or two run nervously about for the stress they have? What what if those one or two are actually those that need to get the work done? I'd say that those one or two are pretty much resistant to stress.
I don't get stressed by jobs. I'm a contractor, so I don't play any role in internal career plans of others (or have any career plans in the company of myself that could affect my or others thinking). Basically I don't really care wether anyone makes money or not or looks good or not on those projects. All I care about is to get the job done properly. If anyone has unrealistic schedules, then it's not my fault. And in 90% of the cases, is some manager makes an unrealistic schedule, then he usually does some mistakes that are undeniably his own that will pull delay the project more and gives everybody else more time to do their jobs.
I can only remember one time when I was stressed out at work. But the source of that stress was private life, wich was falling apart in a lot of ways then. Obviously that affected my performance, and people noticed. But it didn't last too long.
Everybody stressing out over a job can just blame himself.
That being said, I usually have pretty good project managers.
Now, intellectual ability is very hard to measure. And the only reason chess is so good to measure is, that it is an almost mechanical undetaking (no wonder computers are better at it than humans).
Even the article itself acknowledges that there ar egood learners and bad learners (and what else is that difference than talent). There is a myriad of menatal disadvantages, like legastenics, dawn syndrome etc. Making experts of them becomes pretty impossible in many cases. And it's safe to say that there are mental advantages just as well. The decades or centuries of experience of learning in school are a good example that talent matters. If talent didn't matter, all school classes would be on the same level. And don't tell me that homework matters. Sure it helps. But there are those that are top of the class without doing homework. I know, I was one of them. I was even top of the class without going to school half of the time (due to illness).
Well, my fingers are emacs wired. So let the flame wars begin;)
But I have to say, that in all my years programming, I have never been more productive than when using emacs on a widescreen with 2 vertically seperated windows open at all times. And if neccessary, a few additional console windows. The hands never leave the keyboard in that setting. You have svn/shell/dired/ssh/editing/documentation readily available, all only being a tenth of a second away at your fingertips. Even for file management that that double window setup beats even the * Commander tools by far. Teh only beef I have is, that etags (for better source browsing... sort of a half-intellisense) has problems with generating tags with MACRO defined entities in C and C++. But well.
Well, but it could be used as a cache for other permanent storage like flash or a hd. This could have strong improvements in database performance especially in transaction management.
Yep, it's refreshing to see that there actually are managers that like to think sometimes and come up and learn something new once in a while. Everywhere else the horror of having to adapt or change a business model or having to learn something new about the market is so obvious and palpable. Music industry having to change their thinking for the first time in decades scream world ending. Software companies seeing that cheap cash from licenses on every device, even processor, might not be acceptable on the customer side entrenching themselves more and more into "intellectual property".
As a developer who has to learn something new, and figure out new ways of doing something almost every day this is almost ridiculous if it wasn't so sad.
Right. Learning, ant at least fiddling around with a lot of different languages is the most best way to learn programming. With that you do several things:
- you get to know a lot of different programming paradims, ways of solving problems, that would be completely lost to you, if you stick to just one language. The least it does is to give you an understanding, when someone tries to explicitely implement one of those paradigms in a language that isn't meant for it (because the problem requires it). Tremendously helps understandaning code and actually desinging solutions. - you get a feel for what language implements wich concept elegantly and which doesn't. Fact is, often people just don't use powerful features of a language, because it is implemented awkwardly. If you get to use this feature in another language, where it is solved elegantly, you might find yourself learnign it better, and then using it more often, when it is appropriate. For example, people who have done some Lisp completely lose their trouble with recursion. I too often find people who only basically learned C(++) and java and all this often don't even think of it... despite the language allowing it. And it is so often the most elegant and shortest solution. Same for Reflection. People who just use Java and.Net often shy away from it (since it is awkward), and instead prefer to maintain long tables and constant definitions. But if you have used reflection a little (and it is easier in Lisp or Smalltalk), the you know from experience how much typing and maintenance it can save you, and you still use it, despite the implementation being a little awkward. OR people from the widows world often shay away from regular expressions. But people with some unix and perl experience often use them for what they are meant to, despite first having to import some external library in C++ and having to explicitely create objects and all this. - you learn to derive knowledge from language similarities. Just as with natural languages, Learning you first foraign language is the hardest, the more you actually know, the easier it becomes to learn new one, due to all those similarities. And with each new language you learn new ways of expressions and concepts, that tremendously help you expressing yourself in the languages you already know.
Well, once x86 come to cellphones 30min battery time will become standard.
Have you noticed how Apple dropped battery runtime from their marketing, once they switched to intel? My G4 iBook runs almost 8 hours, when all I do is reading/editing documents, surfing, and the occasional compile. With the MacBook you are lucky to have 3 hours.
So what is more productive? When you can work with your code/documents and are available 8 hours, and the occasional compile takes a minute? Or when you can with your code/documents and are available 3 hours, and the occasional compile takes 30 seconds?
I won't even get to servers. The only real selling point intel has there, is that it runs windows. And anyone seriously willing to run server applications with windows, is insane anyway. Far less performance, far higher licensing and support costs, much lower uptime, a bitch to administrate. but well, the suits that decide what to buy know the name windows, and get their conferences andpaid by MS and certificates. So those decisions are easy for them.
Slashdot Mirror
They already said, the bigger version is supposed to run under Windows CE, and nothing else planned so far.
Been waiting for something like this for years too. But considering nVidia's history, I don't think we will get proper linux drivers for this.
When you are a halfway decent programmer and do it professionally, you will come across a wide range of projects. And what happens wich each new project? In a matter of days or 1-2 weeks you gotta acquire the knowledge of your new problem domain. And that is very often a problem domain others spend years to study and get a degree, yet as a software engineer, you gotta be able to understand the problem domain good enough to automate it. T means you must be able to understand all the nomenclature and the structure, you must be able to spot even the subtlest problems, you must know the pitfalls and special cases, you must be able to talk to the experts in their domain on somewhat equal footing. And if you wanna be any good, you must understand the new problem domain enough to generalize it and often to make it accessible to non-experts. ... which is the prerequisite for all this. ... it is your bread and butter to learn things in a matter of days others take and need years to learn.
And that is only one half of the job. The other half is the broad field of computer science
In my life as a programmer I have had to learn the termodynamics and streaming physics of internal combustion engines, the bioengineering for bio-processing tanks in pharmaceutical production. Dentist medicine. Law. Economics. Hard body physics and so much more. And about each year one or two neew things get added to the list.
Sure, I'm no expert in either of those fields. But I know (or knew) enough to find my way around by myself for 90% of the time, and to know when I should ask the expert. And the fact remains
If you make design and usability your main concern in the microprocessor business, you can't make x86 chips anymore.
Wrong. You can only use the hard work of other people if they use YOUR hard work first. Or if they license it in a GPL compatible license.
In both cases I don't see how this could restrict any freedoms for others they haven't used first for themselves.
I'm aware of every single a fact and argument you brought up In fact, I have written assembly code for multiple POWER based Proessors, of which PowerPC is one implementation variant, no production code though, just for excecise, and I have written assembly code for each x86 iteration, including AMD-64 (which actually was production code).
Still doesn't change anything. After NT Windows wasn'T shipped for anything but x86 (and a beta version for itanium).
Fact is, you can't get rid of it, because of inertia (or momentum, as you describe it). There is no single technical reason for having it, in fact, I know for fact that the dominance of x86 architecture prevented a lot of innovation in software. Read some papers on the development of the JavaVM, and how many less than ideal decisions were made because of x86, or for the L4 kernel or the CoyotOS kernels. The ressources that were put in keeping x86 running are few orders of magnitude greater than the ressources put into all other architecures combined. That's not really cost efficiency I think.
And the fact that Apple Laptop battery life is less than half of what it was before the intel switch, despite better battery technology, better chipset integration, better LCD efficency and all those advances should give you a reason to think twice.
There are tremendous costs bound to using x86. It's just that those aren't costs easily to understand by those that buy all those machines. You cannot imagine he amount of work and energy (unproductive work and energy basically) that has been put and still is put into keeping the MS/intel combination running. I have experienced some of it. And only scratched the surface.
The x86 dominance is basically a result of two crooked architecure holding each other up: if MS DOS wasn't so crappy that it depends on x86 then the processor could be changed. If x86 wasn't too crappy to properly emulate it, then MS DOS or it's successors could be changed. As it is, we are stuck with both, because noone wants to change both at the same time, and you cannot really change each independently. ... energy price.
There is something I hope for:
Vista Tanks mightly, OS X and it's successors become the dominant OS in 10 years. Those are instructions set agnostic, and have been proven to be able to run on multiple platforms with pretty little effort, and does so atm run on 3 different instruction sets: x86, POWER, and the iPhone on ARM. Linux runs everywhere too. And as soon as you have this, there is no reason not to drop the most expensive to develop for and least effcient architecture.
But as long as people still use MS Operating systems, we will be stuck with x86 and have to pay the price
Only way to set an industry standard is, to get so fast so big in a new market/technology that everybody has to follow.
Problem is, when you get so big so fast, there are almost neccessarily major flaws in the designs.
Problem is, you never get rid of them again.
Exactly ... I hear that reiterated quite for quite some time already. To really get things going there, somehow it has to be figured out, which parts of the code are covered by 3rd party patents, and which 3rd parties own those patents.
Then we know where to go to and who to spam, or wether it is feasible to just drop those parts of the code, or wether to implement open alternatives. Maybe those parts are even some fraud patents, something like matrix multiplication or whatever.
You are aware that it's the PHBs that are gonna do the firing.
Can you imagine a thing more scary?
If the governement system he deploys will ba anything close to the operating system he deploys, it will be a free for all for all the vile elements to spy on you. And it will make you used to substandard availability of basic services. And how a major crash every few days? And the famous tax you gotta pay wether you use it or not?
A lot of this si implemented already of course. But I'm sure embrace and extend will work better on the middle east than bomb to shreds and see what happens.
1. Controllers for other consoles are 40$ too
2. For other consoles the games cost 60$
3. Without nunchuk, which isn't used for the multiplayer games I have seen so far, the Wii controller is 40$ too
4. You will buy 3 additional controllers at maximum for the Wii, most likely only one
5. You will definitely buy much more games for the Wii, which come at 40$
6. If you don't buy more games for the Wii, you won't care about the additional controller too, or wouldn't buy a Wii in the first place
So, I can absolutel not understand how such a consumer friendly, and actually more affordable price decision by Nintendo can get flamed that much. But well, that's what most business decisions are made anyway: Look at the number you see immediately, compare greater than and lower than, and then take the lower number on buys and the higher number on sells. Any 5 year old can reason on that level. The scary thing is, that is how the world is run: with the reasoning of 5 year olds.
I don't know whwther you have read this review and benchmark comparison of the Old and the new PowerMac/MacPro.
See, on one side we have the 3 years old Dual Single-Core G5 with 2,5GHz. On the other side we have the newest Dual DualCore Xeon with 2,66GHz. That is 3 years advance in technology and manufactoring process (even a geberation generation difference, the G5 is 90nm and the Xeon is 65nm), double the cores cores and a neglectable 1% advantage in Clock. Yet the speed advanatge barely scratches 50% in it's best tests, most of the time the advantage is only between 20% and 30%. And for benovelence they haven't even mentioned Performance/Watt. For comparison, this is somewhat like comparing a dual Pentium to a single i486 and only barely beating it. So why is that? It's the architecture. Those are the things the average user can see when he looks carefully.
The other thing is all the quircs and limitations (to name only a few: real mode, A20 gate, no execute flage etc, awkward paging mechanism, crufted, "baroque" instruction set) that an engineer has to work around to get it to work. That isn't immediately noticable to the end user. But what the end user notices is more glitches and unpredictable problems which are the results of not so clean designs. I'm only guessing, but I sort of have the impression that the well documented, but unheard of before problems with instability and and random shutdowns etc. are a result design not being so clean and managable as before.
I just hope this takeover won't hurt them in thier technology decisions. Because I want to see a Subnotebook with their MPC8641D their MRAM as a buffer and about 10-20GB Flash. Combined with an organic iode display or a high resolution monochrome titanium-oxide screen and a decent battery that would be the ultimate outdoor writing and coding machine. The only thing it won't be overly suited for is video and high end gaming, but still should be enough for most. A laptop with days of battery time ...
Actually the Desktop is the only place Power architecture isn't used on a big scale anymore. Just as x86 isn't used anywhere in any significant amounts but on the desktop.
This has reasons. The decision which processor to use is made by technicians who know what they are doing virtually everywhere but at the desktop. But the vast majority of desktop decisions are made by corporate managers, that are bought by Microsoft or don't know anything else, which means the have to use x86 for (backwards) compatibility, or by end users who don't know anything else.
The only reason that x86 is still around is, that it is the only architecture Microsoft ever got their operating systems working on decently. Tells you a lot about code quality.
Well, I'm a debian user since 1998, and I also conributed half a dozen patches or so in that time.
But I have to say, the only thing that ever caused me serious problems in using Debian (prevented me from doing what I wanted) was the Kernel. And that isn't Debian specific. Problems in the Kernel are prevalent in all distributions.
Hmmm ... what is it when everybody in a project but one or two run nervously about for the stress they have? What what if those one or two are actually those that need to get the work done? I'd say that those one or two are pretty much resistant to stress.
I don't get stressed by jobs. I'm a contractor, so I don't play any role in internal career plans of others (or have any career plans in the company of myself that could affect my or others thinking). Basically I don't really care wether anyone makes money or not or looks good or not on those projects. All I care about is to get the job done properly. If anyone has unrealistic schedules, then it's not my fault. And in 90% of the cases, is some manager makes an unrealistic schedule, then he usually does some mistakes that are undeniably his own that will pull delay the project more and gives everybody else more time to do their jobs.
I can only remember one time when I was stressed out at work. But the source of that stress was private life, wich was falling apart in a lot of ways then. Obviously that affected my performance, and people noticed. But it didn't last too long.
Everybody stressing out over a job can just blame himself.
That being said, I usually have pretty good project managers.
Now, intellectual ability is very hard to measure. And the only reason chess is so good to measure is, that it is an almost mechanical undetaking (no wonder computers are better at it than humans).
Even the article itself acknowledges that there ar egood learners and bad learners (and what else is that difference than talent).
There is a myriad of menatal disadvantages, like legastenics, dawn syndrome etc. Making experts of them becomes pretty impossible in many cases. And it's safe to say that there are mental advantages just as well.
The decades or centuries of experience of learning in school are a good example that talent matters. If talent didn't matter, all school classes would be on the same level. And don't tell me that homework matters. Sure it helps. But there are those that are top of the class without doing homework. I know, I was one of them. I was even top of the class without going to school half of the time (due to illness).
Without those kinds of spending no kid would ever have heard of a computer, since their wouldn't be any, since noone got the funds to develop some.
Just bite into a tomato. Those little yellowish-wite bits in there are the seeds. Never eaten a tomato?
Well, my fingers are emacs wired. So let the flame wars begin ;)
... sort of a half-intellisense) has problems with generating tags with MACRO defined entities in C and C++. But well.
But I have to say, that in all my years programming, I have never been more productive than when using emacs on a widescreen with 2 vertically seperated windows open at all times. And if neccessary, a few additional console windows. The hands never leave the keyboard in that setting. You have svn/shell/dired/ssh/editing/documentation readily available, all only being a tenth of a second away at your fingertips. Even for file management that that double window setup beats even the * Commander tools by far. Teh only beef I have is, that etags (for better source browsing
Well, but it could be used as a cache for other permanent storage like flash or a hd. This could have strong improvements in database performance especially in transaction management.
Yep, it's refreshing to see that there actually are managers that like to think sometimes and come up and learn something new once in a while. Everywhere else the horror of having to adapt or change a business model or having to learn something new about the market is so obvious and palpable. Music industry having to change their thinking for the first time in decades scream world ending. Software companies seeing that cheap cash from licenses on every device, even processor, might not be acceptable on the customer side entrenching themselves more and more into "intellectual property".
As a developer who has to learn something new, and figure out new ways of doing something almost every day this is almost ridiculous if it wasn't so sad.
Right. Learning, ant at least fiddling around with a lot of different languages is the most best way to learn programming. With that you do several things:
... despite the language allowing it. And it is so often the most elegant and shortest solution. Same for Reflection. People who just use Java and .Net often shy away from it (since it is awkward), and instead prefer to maintain long tables and constant definitions. But if you have used reflection a little (and it is easier in Lisp or Smalltalk), the you know from experience how much typing and maintenance it can save you, and you still use it, despite the implementation being a little awkward. OR people from the widows world often shay away from regular expressions. But people with some unix and perl experience often use them for what they are meant to, despite first having to import some external library in C++ and having to explicitely create objects and all this.
- you get to know a lot of different programming paradims, ways of solving problems, that would be completely lost to you, if you stick to just one language. The least it does is to give you an understanding, when someone tries to explicitely implement one of those paradigms in a language that isn't meant for it (because the problem requires it). Tremendously helps understandaning code and actually desinging solutions.
- you get a feel for what language implements wich concept elegantly and which doesn't. Fact is, often people just don't use powerful features of a language, because it is implemented awkwardly. If you get to use this feature in another language, where it is solved elegantly, you might find yourself learnign it better, and then using it more often, when it is appropriate. For example, people who have done some Lisp completely lose their trouble with recursion. I too often find people who only basically learned C(++) and java and all this often don't even think of it
- you learn to derive knowledge from language similarities. Just as with natural languages, Learning you first foraign language is the hardest, the more you actually know, the easier it becomes to learn new one, due to all those similarities. And with each new language you learn new ways of expressions and concepts, that tremendously help you expressing yourself in the languages you already know.
Well, once x86 come to cellphones 30min battery time will become standard.
Have you noticed how Apple dropped battery runtime from their marketing, once they switched to intel? My G4 iBook runs almost 8 hours, when all I do is reading/editing documents, surfing, and the occasional compile. With the MacBook you are lucky to have 3 hours.
So what is more productive? When you can work with your code/documents and are available 8 hours, and the occasional compile takes a minute?
Or when you can with your code/documents and are available 3 hours, and the occasional compile takes 30 seconds?
I won't even get to servers. The only real selling point intel has there, is that it runs windows. And anyone seriously willing to run server applications with windows, is insane anyway. Far less performance, far higher licensing and support costs, much lower uptime, a bitch to administrate. but well, the suits that decide what to buy know the name windows, and get their conferences andpaid by MS and certificates. So those decisions are easy for them.