Kasparov is about 37 now. He's getting close to that age where one is no longer so sharp. Here's
a good article on aging from Feed.
I disagree. First, Kasparov reached his peak chess rating only recently, and actually had increased the gap between him and the others. IIRC, about 5 years ago both Kramnik and Kasparov, had the same rating: so it would look like the 32 years old improved more than the 17 years old.
Second, chess doesn't involve as much imagination as science, it is also technical: this is why Korchnoi was still the 18th player, 1 or 2 years ago, at the age of 68 ; actually Korchnoi reached his best level of play in his forties.
seeing what happened to Apple's stock this week... well, I don't feel bad for them at all
I don't think I've ever heard such a strange disconnected take on morality before. Apple abides by
US law which (like it or not) has been around for over a century, and you're glad their stock dropped?
And the stock dropped for reasons entirely unrelated to this?
Possibly, but I guess CmdrTaco was just expressing resentment at Apple attitude (you cut the "Very depressing")
Like Apple or not, that type of statement is just childish. I'm glad to see that some people on the board
realize that real companies don't have the freedom to give everyone the finger like the majority of
the sit-at-home Slashdot crowd. You only have that freedom because of your anonymity. Try
actually running a business and keeping that attitude. You're stock will end up as wallpaper.
Actually it is not that childish, given the way Apple advertised how they licensed "1-Click Buying". Amazon patent has become the perfect example of crap patent, not only because it is so simple, can be used so pervasively on all the Internet e-commerce sites, but also because it is actually used as a weapon and Amazon sues.
Now licensing doesn't look to bright (why not use 2-click shopping, instead ?), but maybe they needed it, you're right. On the other hand, the 1-click patent glorification and propaganda in the PR release is too much to swallow. Especially from one company advertising "being different".
At best they could have said simply "we are using 1-click ordering now". A more militant company could have say something in the lines "we have licensed 1-click ordering now, for your convenience, despite the patent is crap".
Look at the mess this attitiude has created in the current implementation of TCP/IP. The entire net is run on a
hideously patched together and crufty protocol because everyone just decided to support it because, hey, it was
there.
TCP/IP is quite good. It might reach its limits on terabit/second networks but it serves extremely well its purpose: a protocol that adapts to the available bandwidth, without creating congestion collapses. Most of its options address an important point (such as Nagle algorithm, or SACK). What do you want to replace TCP/IP with ?
Ok, ok. You're right. But I don't think that's right. The program should automatically detect which open I'm
talking about.
Hmm... Python doesn't. The reason is that most of the time, you don't use "from package import *", but simply "import package" and use "package.function(...)".
But it still doesn't allow writing from a variable. But I'm
not going to comment on that because it's probably some other dumb thing that I've done:|
What is "not writing from a variable" ? Or maybe try on comp.lang.python.
"testf" doesn't include the full path name, so of course it can't work unless direc="." (as in any language). Try 'FILE* f=fopen("dev/null", "r")' in C while being on your home directory, it won't work too well, will it ? The fix is:
You become an advocate of a language by recognizing its power?
Well, to compare with that, I use
LISP for 40 hours a week professionally. And in doing *that* I've come to the conclusion that it sucks.
At least, compared to a good OO language like C++ or Smalltalk.
I depends what you are doing. Writing a compiler in Lisp sounds sure much easier in Lisp than in C++ (I wrote one for a mini-language). Of course, I'm not using Lisp everyday so you have a point:-)
On the other hand, it *is* better than
Perl - it just made me laugh that you compare two shitty languages and say one's better than the other.
They're both shitty!;)
Hmm... Perl is uncessarily weird. Lack of syntax from Lisp is precisely what allows meta-programming. Are you using Lisp macros in their full power ?
he ultimate challenge is for someone to win with something written in APL. Then not only will they have
proven themselves truly gifted coders, but then we can have a contest to see if people can figure out
what they wrote!
This has already happened with an APL variant, "J" in 1998. They even won Judge's prize:-)
Shows how shitty LISP is that you consider its nearest rival to be Perl.
Huh? Emacs Lisp might well be one of the worse Lisps. Lisp had the advantage on Perl that it has optional declarations, can be compiled, and has huge meta-level possibilities. Perl is an exemple of C with retrofitted 1/3rd of the possibilities of Lisp.
Now I don't use Lisp, except for very specific programs, but I sure can recognize its power.
Such event
members can be defined to take any delegate type. Delegates
are the C# version of function pointers. Whereas a C function
pointer consists of nothing but a callable address, a delegate is
an object reference as well as a method reference. Delegates
are callable, and when called, operate the stored method upon
the stored object reference. This design, which may seem less
object-oriented than the Java approach of defining a handler
interface and having subscribers subclass the interface and
instantiate a subscriber, is considerably more straightforward
and makes using events nearly as simple as invoking object
methods.
Events are one example of how C# takes a popular use of
pre-existing object-oriented mechanisms and makes it explicit
by giving it a name and logic of its own.
Except that it already existed decades ago in Lisp under the explicit name of "closures", exists also in Python under the name "bound methods", and exists in general in dynamically typed languages under the idea of "protocol". The difference is, in C#, you have type checking, so you have to declare the signature.
Delegate is rather an example of how "Those who don't use Lisp are doomed to reimplement it.":-)
If you are buying new machines, why not ask the vendor to install what you want?
If you are installing 2500 machines, you usually know better than your vendor. 1 hour maintenance per machine equals 1 man-year. You'd better have competent sysadmin(s) and automate as much as possible. You must be able to automate the system reinstallation, software upgrade, hardware upgrades,..., of the machines.
Speaking of which, I've used an S80 (the M80's big brother), there is NO WAY that any 4 CPU PC is 50% faster than an 8 CPU RS/6000 M80 on any comperable benchmark. This raises a large question-mark over the whole result.
Obviously, they deliberatly engineered the bottleneck
Redhat/Dell have done nobody any favours by publishing such a clearly unrepresentitive benchmark result. It's Mindcraft all over again, this time from the other side, redoubled.
I agree. If you look closer, Mindcraft tried to get a better image, by sponsoring one of the benchmarks in May 2000 (usually only hardware/software manufacturer/sellers do), but I guess some geeks at Dell wanted to give them a taste of their own medecine.
At least we have 1 Gbps NICs in this test (not like expensive 4 CPU server with 4 x 100 Mbps NICs), but I'd like to see a revamped SPEC test with no network splitting and better than RAID 0.
Why is SPEC biased in favor of Intel ? SPEC95 includes 2 programs which I frequently use, Perl and gcc, (not to mention compress and ijpeg, which are not an irrealistic type of programs). SPEC2000 even includes gcc, perl, crafty, gzip, bzip2 and mesa.
The idea is basically split information in several "pads", that truely look random. You get the information back by xor-ing several pads (or known pseudo-random generators).
The idea is not new. I even saw it in some magazine such as Scientific American or Dr Dobbs Journal: the idea there was to split information in several pieces, so that you need a number of them to get the information (at least a fixed number of them, but not necessarily all).
The problem is that it is pushed as a method for promoting free speech. IANAL, but if your free speech is legal, then there is no need to encrypt it ; if it is not, then you'll just prove that you were aware of its illegality, and tried to work around, which won't look too good in front of the judge. Also carrying crypted pieces of information can get you into trouble. You can't say you are unaware that it might be illegal: why was it crypted in the first place ?
The other problem is that once someone knows how to decipher the random pads, nothing prevents him to tell how to do to others. I expect NSA to be quickly informed. Of course: it is difficult to make a crypto scheme that the "good" people can decipher but not the "bad" people. You need a shared secret, and sharing a secret at Internet scale is not safe. Or you need to encrypt with the public keys of some safe people (which you know personally, or are 100% sure by other means) but then (mail|public repository)+PGP is a better solution.
In my defense I wasn't necessarily trying to say that Perl, Python, and Tcl were revolutionary. I was mostly trying to rebut the slide on page six that is subtitled "Hardware has changed dramatically; software is stagnant."
I agree. Software has much changed ; scripting languages and Java, are a clean proof. They even brought some of older language research into commonly used systems. The extension of Pike's question would be now "is current language research relevant?", but I'm not a language researcher, so I suspect I'll have to wait 10 years to see novelties appearing (or not).
I also imagine that a paper describing how to scale Gnutella to the point where it would still be useful with two orders of magnitude more users would also be "research material."
Probably. Probably such a paper had already been written long ago:-) A recent candidate might be in Globe
Mr. Pike is just upset because his beloved innovative Plan 9 is being completely and totally eclipsed by Unix-like operating systems like Linux and FreeBSD.
You are telling us that his research was irrelevant. That's exactly his point: "system software research is irrelevant". It's the title. Not that I necessarily agree or disagree with his point.
However, unless this new system allows the end user to do the things that they traditionally use their computer for what's the point? Why would I want an innovative new system that can't send email or surf the web (or run Emacs for that matter)? I wouldn't, no matter how novel it might be.
Again you seem to violently agree with him. Look: "Who needs new operating systems, anyway?" you ask. Maybe no one, but then that supports my thesis (his words).
Napster, Gnutella, and FreeNet are all innovative ways of sharing files, yet all of them rely on existing standards.
This is more some kind of (limited) innovation than research. A paper describing Napster probably won't meet the standards for research publication, for starters. A bit too basic. Maybe if you throw in some measurements and statistics. He, again that's what he is saying.
Mr. Pike also totally belittles the success of scripting languages like Perl, Tcl, and Python.
This is a rather bad example ; Smalltalk in 1980 was already equal or superior to Perl, Tcl and even Python (I'm not too biased, since I'm a Python advocate). Lisp environments have been superior for long, and are still offering yet unequalled features.
I'd challenge you to say that "Perl, Tcl and Python have been major innovations" to some language researcher. He'd be more likely to yell at you "Where have you been since Lisp (60s), Smalltalk (80s), and others ? I've been telling you this for dozens of years". Or at worst, he would make some derogatory comments involving the smoking of illicit substances. If you want to talk about research in languages, you'd rather talk about FP (Haskell, [O]Caml,...), Self, meta-object protocols, intensional programming, distributed systems, program proofs, elaborate optimisation, sophisticated garbage collection, parallel compilation, etc... Not Perl, Tcl or Python! Do you really want the poor researchers to have a heart attack ?:-)
Python does NOT have all of the power of Perl <br> Where are closures? <br> If you wonder why I mention them, then you have never tried functional programming techniques. </em> <p> This is half-intentionnal. You aren't expected to use functional programming techniques in Python, but to use objects.<p> This is very debatable, but at least people won't argue forever whether to use closures or objects in a particular context, and you can do mostly what closures could do.<p>For instance, if you need a callback either pass an object "o" and later call a well-known function "o.callbackFunction(args)", or pass a bound method "m=o.callbackFunction", and later call "m(args)".
I've just downloaded and played a bit with the language. Basically, if I understood properly:
It is in fact quite close to Java.
It has garbage collection (or is it reference counting ? I created some circular dependencies and it didn't seem to collect).
It has high level types: string, array, mapping, multiset. It has simple syntax for defining constants with those objects, and basic operation on them (like lists, dictionnaries, hashes in Perl or Python).
It has OO, looks less hackish than Perl, but just doesn't look as clean as Java. Is class-shared data possible ?
The variables have to be declared. It is not statically typed though, and it has a "mixed" type which is a placeholder for any type (something like Delphi's "Variant").
It tries hard to be like C. Exactly the same declaration syntax, operators, control flow structures. Worrying point: it also has the preprocessor.
It is probably less dynamic than Perl or Python (I couldn't set the method of one object to be another, at runtime).
Basically it is closer to Java than to scripting languages because of the declarations. It might be simpler to use than Java if, for instance, you used the "mixed" type (can hold any object) everywhere, because you don't need casts ; it still looks heavier than Tcl, Perl or Python.
monopoly n. a business or inter-related group of businesses which controls so much of the production or sale of a product or kind of product as to control the market, including prices and distribution.
So, Microsoft is a monopoly because IBM, a major member of the distribution is forced to distribute only Microsoft's operating system?
No, because most of the IBM clients want Windows (for desktop OS), so IBM distributes it. And IBM has to abide by some Microsoft conditions.
If most clients didn't want Windows on their desktop OS, then Microsoft won't be a monopoly.
Does anybody realize just how dated this 'Python 3000' stuff is going to sound in a thousand years' time? </em> <p> Well it was called "Python 2000" until this year:-)
But if I need something clean or structured why should I use Python instead of, say, C++?
The main factor is that it is easier to program in Python than in C++. Of course depending on other factors, the balance can be in favor of one or the other language , and often complementary.
Re:Prodigies in math, music -- programming?
on
Too Old To Code?
·
· Score: 1
I remember reading once that "experts" (that nebulous phrase) only allow for the existence of prodigies in two fields: music and math. That is, young people (espec. children) can only be as good as adults in those two fields.
That's right ; in math, in particular, young people can do better than older people. As Hardy ranted, "no mathematician has done a major discovery after 30" (Erdos came later, Wiles too?). The worst case is in sports.
However, intellectual skills don't change that much. A skilled mathematician at 50 would still do better than an average one at 25. This is also true in board games ; many top champions stay on the top long, even if they are past their prime, because the youngsters have less talent they had themselves in their prime. Kasparov (who should be around 40), has never dominated all the other like he does now ; Kortchnoi (around 69) is still in the top 30 ; Marion Timsley was a famous example for checkers (he stopped for many years, came back, beat every one, and started matchs with computers because only them could be a challenge to him).
And these are competitive games (which mathematics is also, to a point), where the difference in ability is magnified.
The point is, a good programmer at 20, should be a good one at 50.
The smallest installation of Linux with X that I've seen takes up 40 megs.
Of course there is lots of unnecessary stuff in those 40 MB. For comparison, back in 1994, the first Debian released took about 40 MB with every package installed.
Go to Compaq's web site, download the plans, and build one yourself..:-) </em> <p> I investigated, but I was scared at the price of individual components, and this resulted into a question to <a href="http://slashdot.org/article.pl?sid=00/05/27/ 1846219">askslashdot</a>. <p> <em> ike the PDA market is really heating up and getting seriously competitve at this point, and you can't make a lot of money just on the few gear-heads who buy your product because its' guts are cool </em> <p> Hmmm... no money.... Well, remember in the eighties the flury of micro-computers (Pet, VIC, Apple II, Commodore64, ZX81 Spectrum, QL, Oric, Aquarium, MSX,...) ? Granted, maybe they were also used for games (like many PC nowadays).
I think what the critics of open source security are missing is the deterrent power of open source. If they are really right in their claim that more crackers than good guys will be finding flaws in my program, then that's a strong deterrent for me to just code away as I wish. I have a sort of moral responsibility for the code I write (the warranty disclaimers notwithstanding) and I would be peeved if a cracker penetrated a system because of gaping security holes in my work.
The incentive for writing better code is that much lesser if I know that "hell, who's going to be spending time disassembling this code, I've got a deadline to meet".
I think this can't be analyzed in such simple terms, and only a statistical analysis can show who is winning and why.
For starters, in Open Source, there can also be similar problems "hell, it works for most people, I've better things to do", or even "I or someone else, will add necessary features later, including security", "whoever need it, just has to code it".
Second, I think that at least some people in some companies will review or audit their code, especially if it is security-related. Do you think that repeated and deliberatesloppy coding won't be of some consequence for its author ?
I agree that probably the Apache group, or OpenBSD teams are probably falling in the "deterrent power of open source". However for many other Open Source projects it is not a factor. Hell, some projects are grown, and people are happy that the program works at all.
More time spent on security means less time spent on interesting features. So it depends.
Slashdot Mirror
I disagree. First, Kasparov reached his peak chess rating only recently, and actually had increased the gap between him and the others. IIRC, about 5 years ago both Kramnik and Kasparov, had the same rating: so it would look like the 32 years old improved more than the 17 years old.
Second, chess doesn't involve as much imagination as science, it is also technical: this is why Korchnoi was still the 18th player, 1 or 2 years ago, at the age of 68 ; actually Korchnoi reached his best level of play in his forties.
I don't think I've ever heard such a strange disconnected take on morality before. Apple abides by US law which (like it or not) has been around for over a century, and you're glad their stock dropped? And the stock dropped for reasons entirely unrelated to this?
Possibly, but I guess CmdrTaco was just expressing resentment at Apple attitude (you cut the "Very depressing")
Like Apple or not, that type of statement is just childish. I'm glad to see that some people on the board realize that real companies don't have the freedom to give everyone the finger like the majority of the sit-at-home Slashdot crowd. You only have that freedom because of your anonymity. Try actually running a business and keeping that attitude. You're stock will end up as wallpaper.
Actually it is not that childish, given the way Apple advertised how they licensed "1-Click Buying". Amazon patent has become the perfect example of crap patent, not only because it is so simple, can be used so pervasively on all the Internet e-commerce sites, but also because it is actually used as a weapon and Amazon sues.
Now licensing doesn't look to bright (why not use 2-click shopping, instead ?), but maybe they needed it, you're right. On the other hand, the 1-click patent glorification and propaganda in the PR release is too much to swallow. Especially from one company advertising "being different". At best they could have said simply "we are using 1-click ordering now". A more militant company could have say something in the lines "we have licensed 1-click ordering now, for your convenience, despite the patent is crap".
TCP/IP is quite good. It might reach its limits on terabit/second networks but it serves extremely well its purpose: a protocol that adapts to the available bandwidth, without creating congestion collapses. Most of its options address an important point (such as Nagle algorithm, or SACK). What do you want to replace TCP/IP with ?
Hmm... Python doesn't. The reason is that most of the time, you don't use "from package import *", but simply "import package" and use "package.function(...)".
But it still doesn't allow writing from a variable. But I'm not going to comment on that because it's probably some other dumb thing that I've done :|
What is "not writing from a variable" ? Or maybe try on comp.lang.python.
Nice way to get your program corrected :-)
"testf" doesn't include the full path name, so of course it can't work unless direc="." (as in any language). Try 'FILE* f=fopen("dev/null", "r")' in C while being on your home directory, it won't work too well, will it ? The fix is:
f=open(direc+"/"+testf,"r")
with the "/" being unnessecary if direc=="/".
I depends what you are doing. Writing a compiler in Lisp sounds sure much easier in Lisp than in C++ (I wrote one for a mini-language). Of course, I'm not using Lisp everyday so you have a point :-)
On the other hand, it *is* better than Perl - it just made me laugh that you compare two shitty languages and say one's better than the other. They're both shitty! ;)
Hmm... Perl is uncessarily weird. Lack of syntax from Lisp is precisely what allows meta-programming. Are you using Lisp macros in their full power ?
This has already happened with an APL variant, "J" in 1998. They even won Judge's prize :-)
Huh? Emacs Lisp might well be one of the worse Lisps. Lisp had the advantage on Perl that it has optional declarations, can be compiled, and has huge meta-level possibilities. Perl is an exemple of C with retrofitted 1/3rd of the possibilities of Lisp.
Now I don't use Lisp, except for very specific programs, but I sure can recognize its power.
Events are one example of how C# takes a popular use of pre-existing object-oriented mechanisms and makes it explicit by giving it a name and logic of its own.
Except that it already existed decades ago in Lisp under the explicit name of "closures", exists also in Python under the name "bound methods", and exists in general in dynamically typed languages under the idea of "protocol". The difference is, in C#, you have type checking, so you have to declare the signature.
Delegate is rather an example of how "Those who don't use Lisp are doomed to reimplement it." :-)
If you are installing 2500 machines, you usually know better than your vendor. 1 hour maintenance per machine equals 1 man-year. You'd better have competent sysadmin(s) and automate as much as possible. You must be able to automate the system reinstallation, software upgrade, hardware upgrades, ..., of the machines.
Obviously, they deliberatly engineered the bottleneck
Redhat/Dell have done nobody any favours by publishing such a clearly unrepresentitive benchmark result. It's Mindcraft all over again, this time from the other side, redoubled.
I agree. If you look closer, Mindcraft tried to get a better image, by sponsoring one of the benchmarks in May 2000 (usually only hardware/software manufacturer/sellers do), but I guess some geeks at Dell wanted to give them a taste of their own medecine.
At least we have 1 Gbps NICs in this test (not like expensive 4 CPU server with 4 x 100 Mbps NICs), but I'd like to see a revamped SPEC test with no network splitting and better than RAID 0.
Why is SPEC biased in favor of Intel ? SPEC95 includes 2 programs which I frequently use, Perl and gcc, (not to mention compress and ijpeg, which are not an irrealistic type of programs). SPEC2000 even includes gcc, perl, crafty, gzip, bzip2 and mesa.
The idea is not new. I even saw it in some magazine such as Scientific American or Dr Dobbs Journal: the idea there was to split information in several pieces, so that you need a number of them to get the information (at least a fixed number of them, but not necessarily all).
The problem is that it is pushed as a method for promoting free speech. IANAL, but if your free speech is legal, then there is no need to encrypt it ; if it is not, then you'll just prove that you were aware of its illegality, and tried to work around, which won't look too good in front of the judge.
Also carrying crypted pieces of information can get you into trouble. You can't say you are unaware that it might be illegal: why was it crypted in the first place ?
The other problem is that once someone knows how to decipher the random pads, nothing prevents him to tell how to do to others. I expect NSA to be quickly informed. Of course: it is difficult to make a crypto scheme that the "good" people can decipher but not the "bad" people. You need a shared secret, and sharing a secret at Internet scale is not safe. Or you need to encrypt with the public keys of some safe people (which you know personally, or are 100% sure by other means) but then (mail|public repository)+PGP is a better solution.
I agree. Software has much changed ; scripting languages and Java, are a clean proof. They even brought some of older language research into commonly used systems.
The extension of Pike's question would be now "is current language research relevant?", but I'm not a language researcher, so I suspect I'll have to wait 10 years to see novelties appearing (or not).
I also imagine that a paper describing how to scale Gnutella to the point where it would still be useful with two orders of magnitude more users would also be "research material."
Probably. Probably such a paper had already been written long ago :-) A recent candidate might be in Globe
You are telling us that his research was irrelevant. That's exactly his point: "system software research is irrelevant". It's the title. Not that I necessarily agree or disagree with his point.
However, unless this new system allows the end user to do the things that they traditionally use their computer for what's the point? Why would I want an innovative new system that can't send email or surf the web (or run Emacs for that matter)? I wouldn't, no matter how novel it might be.
Again you seem to violently agree with him. Look:
"Who needs new operating systems, anyway?" you ask. Maybe no one, but then that supports my thesis (his words).
Napster, Gnutella, and FreeNet are all innovative ways of sharing files, yet all of them rely on existing standards.
This is more some kind of (limited) innovation than research. A paper describing Napster probably won't meet the standards for research publication, for starters. A bit too basic. Maybe if you throw in some measurements and statistics. He, again that's what he is saying.
Mr. Pike also totally belittles the success of scripting languages like Perl, Tcl, and Python.
This is a rather bad example ; Smalltalk in 1980 was already equal or superior to Perl, Tcl and even Python (I'm not too biased, since I'm a Python advocate). Lisp environments have been superior for long, and are still offering yet unequalled features.
I'd challenge you to say that "Perl, Tcl and Python have been major innovations" to some language researcher. He'd be more likely to yell at you "Where have you been since Lisp (60s), Smalltalk (80s), and others ? I've been telling you this for dozens of years". Or at worst, he would make some derogatory comments involving the smoking of illicit substances. :-)
If you want to talk about research in languages, you'd rather talk about FP (Haskell, [O]Caml,...), Self, meta-object protocols, intensional programming, distributed systems, program proofs, elaborate optimisation, sophisticated garbage collection, parallel compilation, etc... Not Perl, Tcl or Python!
Do you really want the poor researchers to have a heart attack ?
This might be a problem for commercial software (like Blender), or for free software not GPL-compatible.
Python does NOT have all of the power of Perl
<br>
Where are closures?
<br>
If you wonder why I mention them, then you have never tried functional programming techniques.
</em>
<p>
This is half-intentionnal. You aren't expected to use functional programming techniques in Python, but to use objects.<p>
This is very debatable, but at least people won't argue forever whether to use closures or objects in a particular context, and you can do mostly what closures could do.<p>For instance, if you need a callback either pass an object "o" and later call a well-known function "o.callbackFunction(args)", or pass a bound method "m=o.callbackFunction", and later call "m(args)".
Basically it is closer to Java than to scripting languages because of the declarations. It might be simpler to use than Java if, for instance, you used the "mixed" type (can hold any object) everywhere, because you don't need casts ; it still looks heavier than Tcl, Perl or Python.
So, Microsoft is a monopoly because IBM, a major member of the distribution is forced to distribute only Microsoft's operating system?
No, because most of the IBM clients want Windows (for desktop OS), so IBM distributes it. And IBM has to abide by some Microsoft conditions.
If most clients didn't want Windows on their desktop OS, then Microsoft won't be a monopoly.
Does anybody realize just how dated this 'Python 3000' stuff is going to sound in a thousand years' time?
</em>
<p>
Well it was called "Python 2000" until this year
The main factor is that it is easier to program in Python than in C++. Of course depending on other factors, the balance can be in favor of one or the other language , and often complementary.
That's right ; in math, in particular, young people can do better than older people. As Hardy ranted, "no mathematician has done a major discovery after 30" (Erdos came later, Wiles too?). The worst case is in sports.
However, intellectual skills don't change that much. A skilled mathematician at 50 would still do better than an average one at 25. This is also true in board games ; many top champions stay on the top long, even if they are past their prime, because the youngsters have less talent they had themselves in their prime. Kasparov (who should be around 40), has never dominated all the other like he does now ; Kortchnoi (around 69) is still in the top 30 ; Marion Timsley was a famous example for checkers (he stopped for many years, came back, beat every one, and started matchs with computers because only them could be a challenge to him).
And these are competitive games (which mathematics is also, to a point), where the difference in ability is magnified.
The point is, a good programmer at 20, should be a good one at 50.
Of course there is lots of unnecessary stuff in those 40 MB. For comparison, back in 1994, the first Debian released took about 40 MB with every package installed.
Go to Compaq's web site, download the plans, and build one yourself..
</em>
<p>
I investigated, but I was scared at the price of individual components, and this resulted into a question to
<a href="http://slashdot.org/article.pl?sid=00/05/27
<p>
<em>
ike the PDA market is really heating up and getting seriously competitve at this point, and you can't make a lot of
money just on the few gear-heads who buy your product because its' guts are cool
</em>
<p>
Hmmm... no money.... Well, remember in the eighties the flury of micro-computers (Pet, VIC, Apple II, Commodore64, ZX81 Spectrum, QL, Oric, Aquarium, MSX,
The incentive for writing better code is that much lesser if I know that "hell, who's going to be spending time disassembling this code, I've got a deadline to meet".
I think this can't be analyzed in such simple terms, and only a statistical analysis can show who is winning and why.
For starters, in Open Source, there can also be similar problems "hell, it works for most people, I've better things to do", or even "I or someone else, will add necessary features later, including security", "whoever need it, just has to code it".
Second, I think that at least some people in some companies will review or audit their code, especially if it is security-related. Do you think that repeated and deliberatesloppy coding won't be of some consequence for its author ?
I agree that probably the Apache group, or OpenBSD teams are probably falling in the "deterrent power of open source". However for many other Open Source projects it is not a factor. Hell, some projects are grown, and people are happy that the program works at all.
More time spent on security means less time spent on interesting features. So it depends.