As an aside, the value of an idea can be destroyed when more people know it - this is why we have trade secret law.
The value being destroyed is only the value as seen from the point of view of someone who had exclusive control over its use.
There is another value gained, but it is gained by all of society, by the widespread use of that idea.
In some cases, if all of society are free to use some piece of intellectual property as they see fit (including torrenting it), the incentive to create it is diminished, and in that case society loses something it could have gained if people abstained from exercising their natural freedoms.
It's a highly non-trivial problem to figure out when it's best to restrict people and when it's best to let them be free.
we'll see if this strategy will continue to work once everybody will be more fluent with computers.
I think "everybody" will be more fluent with computers only inasmuch as young people start to replace those who are old now.
Take your in-laws. Take your old English class. Take your band, or the stamp collectors' club. Of those, take all between 15 and 35. How many run Linux? They're the "everybody" who will inherit the earth.
I might redact this in 35 years when I have grandkids;-)
if the focus is on bringing stability, visual appeal, and new user interface innovations
Ubuntu is stable enough for me, and my debian box only goes down when I'm acting dumb. Both have compiz, so there's your visual appeal factor.
What "Linux" needs (speaking only for Ubuntu) is more work on basic usability stuff. No big gaping holes, but just stupid small cracks that are annoying time-wasters. We all know that programmer time is more expensive than CPU time; isn't the same true for end-user time?
I'v recently filed two bugs against really lame error messages.
One was when I installed a new kernel; dkms (I assume) complained when trying to make fglrx.ko, complaining that I didn't have kernel source installed. Ok fine, I install linux-source-2.6.27-11-generic. And dkms still complains. Ok, I hit google. Some talk of installing a package named something-fglrx-something. That's already installed.
It turns out I needed linux-headers-2.6.27-11-generic. Why the FUCK does dkms lead me astray but telling me I need the source? Lame. "That's great Trillian, 10 out of 10 for good intentions but minus several million for usefulness of error message."
Next is network-manager-vpnc. I ask it to import/etc/vpnc/foo.conf, and it says "Error: file not readable or (what amounts to 'syntax error', no line/col numbers, no nothing, just 'it dont look rite, i doun like it')".
Hello? If you can't read the file, how can you find a syntax error in it? Couldn't you just tell me which of the two MUTUALLY EXCLUSIVE error conditions have occured?
That shit needs to get better. We can masturbate our brains to multi-pointer X and computational telepathy when we have fixed the basic problems.
Linux Starter: LFS. You said you wanted Linux from the start, right?
Linux Home Basic: Ubuntu. Linux for human beings (but only when they're at home).
Linux Home Premium: Kubuntu. Linuks for human beinks, with tolerance for a higher kognitive load imposed by the konfiguration options, and the lack of leksikal korrektness
Linux Business: Linux Mint. Because that's what busineses try to make an exemplar of.
Linux Enterprise: CentOS. It's the community enterprise OS!
most of us chose for Linux because you get the choice, not because you want everything to be chosen for you. If you prefer that, go for a Mac or something.
Or just stick to GNOME;-)
[Sorry if I offended you. I use GNOME and I like it].
"Let's go to Iraq and export our democratic American values" "Hey, they're building a statue that says something that offends us" "Hey, that statute reminds us of violent attacks on rulers" "Let's tear down this statue, we don't like what it says"
Freedom is a multidimensional thing. It's an element of {0, 1}^n: the freedom to do thing 0,..., the freedom to do thing n.
Only using free software means you abstain from exercising freedoms k..k', in return for having freedoms 0..3 with regard to all software you use.
Whether you value freedoms 0..3 more or less than k..k' is an individual value judgement, and you're free (heh) to make that value judgment on your own. And there is no correct, true value judgment, only individual choices.
I can suggest reading rfc1178 (http://tools.ietf.org/html/rfc1178).
It contains some common-sense advice about host naming. Here's a sample:
Don't call them your own name
Don't call them fuckface
Don't spel teh namez w0rng
Don't name them after what they do
Don't give them a name that already has a meaning or refers to something.
Use names from some big set
I'm so far successfully naming my boxes after moons in the solar system. Pro: you can think of the boxes as A, B, C, etc., but let them have more interesting names than that.
Anime characters should be fine too. Usagi, Chiyo-Chan, Sakura,...:D
Or you could go for slashdot memes... natalie-portman, cowboyneal, in-soviet-russia, car-analogy, etc...;-)
[ruby is more OOP than python:] [].length vs len([]), [1, 2, 3].each vs for i in [1, 2, 3]
Given my understanding of OOP, those are not particularly good examples.
len([]) is equivalent to [].__len__, a method invocation on the object in question. Different classes can define different __len__ methods.
for i in some_object calls some_object.__iter__ behind the scenes; same thing again: a method with different definitions across different classes.
What defines OOP is not where you put the dot, but what the dot means, which roughly speaking is the method lookup in a vtable. Ruby and Python both do that, so they're both OO. Python just puts a little syntactic (and semantic) sugar on it.
Something I'm more likely to agree with: Python and Ruby are both more OO than C++ with respect to the collection classes.
Why is that? Because (for example) vector.size and set.size aren't overridden versions of (a hypothetical) collection.size.
In Haskell parlance (note: "class" means two different things in C++ and Haskell), the code template <T> int foo(const T& t) { return t.size(); }
uses class-bound polymorphism with inferred classes. Meaning that the set of types T for which the code is valid is the class of types that have a size method. The inferred-ness (roughly speaking) means that you don't write template <T extends has_a_size_method>...
If you still believe Ruby is more OOP than Python, could you point out either (1) another example, or (2) why you disagree with my thinking?
I hope one or more of us can learn something from this discussion:)
Re:I am afraid, there is lack of direction for Rub
on
Ruby 1.9.1 Released
·
· Score: 1
Lisp is good for people who think very mathematically.
Which aspects of mathematical thinking is it that is well aligned with Lisp?
What I think characterizes mathematical thinking (as opposed to programming thinking) is the declarative and/or pure nature of math: variables don't change, and there's no notion of time.
I think a pure functional language, such as Haskell (or at least pure code) would fit better with mathematical thinking, because it has the same unchanging nature that math has.
I'm guessing that a pure logic programming language would also make a good fit, but I haven't (really) done much logic programming.
But then again, map and filter are axioms in ZFC and they appear in roughly speaking all functional languages as well, so maybe there's something to it:)
So now it's only "really slow" as opposed to "really really slow"?
It's really really simple: a speed improvement of a factor x means you multiply the number of "really"s by 1/x.
Had it been three times as fast, it would have been merely "real slow".
If you're ever reduced (hah!) to fractions of a single letter, just scale the ASCII value of that letter. So if it was 24 (=2*12) times as fast, it would be half an r, or "8 slow".
Slashdot Mirror
What about apologies to Canadians? >:[
Not unless you can make an oval smiley :p
most Canadians see an ad for "Joe's plumbing [...]"
Him again?? The campaigns are over! :(
As an aside, the value of an idea can be destroyed when more people know it - this is why we have trade secret law.
The value being destroyed is only the value as seen from the point of view of someone who had exclusive control over its use.
There is another value gained, but it is gained by all of society, by the widespread use of that idea.
In some cases, if all of society are free to use some piece of intellectual property as they see fit (including torrenting it), the incentive to create it is diminished, and in that case society loses something it could have gained if people abstained from exercising their natural freedoms.
It's a highly non-trivial problem to figure out when it's best to restrict people and when it's best to let them be free.
we'll see if this strategy will continue to work once everybody will be more fluent with computers.
I think "everybody" will be more fluent with computers only inasmuch as young people start to replace those who are old now.
Take your in-laws. Take your old English class. Take your band, or the stamp collectors' club. Of those, take all between 15 and 35. How many run Linux? They're the "everybody" who will inherit the earth.
I might redact this in 35 years when I have grandkids ;-)
sentences apart into two halves that can't be understood on their own, and the subject isn't descriptive of the body.
Sorry to pee on your sugarcane, but it's really annoying... And sorry for not being above it myself.
And there is no pink lizard in your room.
Dammit. Did someone step on my pink-lizard-shaped mind control probe? :(
if the focus is on bringing stability, visual appeal, and new user interface innovations
Ubuntu is stable enough for me, and my debian box only goes down when I'm acting dumb. Both have compiz, so there's your visual appeal factor.
What "Linux" needs (speaking only for Ubuntu) is more work on basic usability stuff. No big gaping holes, but just stupid small cracks that are annoying time-wasters. We all know that programmer time is more expensive than CPU time; isn't the same true for end-user time?
I'v recently filed two bugs against really lame error messages.
One was when I installed a new kernel; dkms (I assume) complained when trying to make fglrx.ko, complaining that I didn't have kernel source installed. Ok fine, I install linux-source-2.6.27-11-generic. And dkms still complains. Ok, I hit google. Some talk of installing a package named something-fglrx-something. That's already installed.
It turns out I needed linux-headers-2.6.27-11-generic. Why the FUCK does dkms lead me astray but telling me I need the source? Lame. "That's great Trillian, 10 out of 10 for good intentions but minus several million for usefulness of error message."
Next is network-manager-vpnc. I ask it to import /etc/vpnc/foo.conf, and it says "Error: file not readable or (what amounts to 'syntax error', no line/col numbers, no nothing, just 'it dont look rite, i doun like it')".
Hello? If you can't read the file, how can you find a syntax error in it? Couldn't you just tell me which of the two MUTUALLY EXCLUSIVE error conditions have occured?
That shit needs to get better. We can masturbate our brains to multi-pointer X and computational telepathy when we have fixed the basic problems.
Sorry, I just had to get it out. Ahh.... :-)
most of us chose for Linux because you get the choice, not because you want everything to be chosen for you. If you prefer that, go for a Mac or something.
Or just stick to GNOME ;-)
[Sorry if I offended you. I use GNOME and I like it].
We're not out to become rich billionaires
I am. But only because I like the offspring of Debian and the color brown.
"Let's go to Iraq and export our democratic American values"
"Hey, they're building a statue that says something that offends us"
"Hey, that statute reminds us of violent attacks on rulers"
"Let's tear down this statue, we don't like what it says"
Export democratic values my $DONKEY
true freedom.
No.
Freedom is a multidimensional thing. It's an element of {0, 1}^n: the freedom to do thing 0, ..., the freedom to do thing n.
Only using free software means you abstain from exercising freedoms k..k', in return for having freedoms 0..3 with regard to all software you use.
Whether you value freedoms 0..3 more or less than k..k' is an individual value judgement, and you're free (heh) to make that value judgment on your own. And there is no correct, true value judgment, only individual choices.
Happed to one of my colleagues... He was reading reddit when he blueit.
I can suggest reading rfc1178 (http://tools.ietf.org/html/rfc1178).
It contains some common-sense advice about host naming. Here's a sample:
I'm so far successfully naming my boxes after moons in the solar system. Pro: you can think of the boxes as A, B, C, etc., but let them have more interesting names than that.
Anime characters should be fine too. Usagi, Chiyo-Chan, Sakura, ... :D
Or you could go for slashdot memes... natalie-portman, cowboyneal, in-soviet-russia, car-analogy, etc... ;-)
There's really no reason to run Linux on the PS3.
Neither is there any reason to run NetBSD on a toaster, but it's still fun (to some people).
And that's reason enough.
it was never meant to be 'consumer protection'
Of course it was. It always ways. The EULA is there to protect the corporation from its consumers.
a temporary decrease in fertility is a feature
Only if the girl's birth control pill fails to work, and in that case the after-oops pill will work just as fine.
I don't buy it as an argument for frying my balls. But feel free to do what you want to your own testicles :)
MS doing this is them trying to ensure that Firefox will work with their web apps
Call me jaded and cynical, but if MS was aiming for interoperability, where are the OS X and Linux versions?
Sigh... Was it not clear I was going for funny? Or was my joke that offensive? :(
Accompanying the article with the FreeBSD logo is slightly tasteless, no?
Well, assume Beastie breathes fire.
That kinda' makes him a toaster...
Python is skiing and ruby is snow boarding.
I guess that makes C drunk driving. Much faster and more crash-prone.
--*car_analogy_quota("jonaskoelker");
[ruby is more OOP than python:] [].length vs len([]), [1, 2, 3].each vs for i in [1, 2, 3]
Given my understanding of OOP, those are not particularly good examples.
len([]) is equivalent to [].__len__, a method invocation on the object in question. Different classes can define different __len__ methods.
for i in some_object calls some_object.__iter__ behind the scenes; same thing again: a method with different definitions across different classes.
What defines OOP is not where you put the dot, but what the dot means, which roughly speaking is the method lookup in a vtable. Ruby and Python both do that, so they're both OO. Python just puts a little syntactic (and semantic) sugar on it.
Something I'm more likely to agree with: Python and Ruby are both more OO than C++ with respect to the collection classes.
Why is that? Because (for example) vector.size and set.size aren't overridden versions of (a hypothetical) collection.size.
In Haskell parlance (note: "class" means two different things in C++ and Haskell), the code
...
template <T> int foo(const T& t) { return t.size(); }
uses class-bound polymorphism with inferred classes. Meaning that the set of types T for which the code is valid is the class of types that have a size method. The inferred-ness (roughly speaking) means that you don't write template <T extends has_a_size_method>
If you still believe Ruby is more OOP than Python, could you point out either (1) another example, or (2) why you disagree with my thinking?
I hope one or more of us can learn something from this discussion :)
Lisp is good for people who think very mathematically.
Which aspects of mathematical thinking is it that is well aligned with Lisp?
What I think characterizes mathematical thinking (as opposed to programming thinking) is the declarative and/or pure nature of math: variables don't change, and there's no notion of time.
I think a pure functional language, such as Haskell (or at least pure code) would fit better with mathematical thinking, because it has the same unchanging nature that math has.
I'm guessing that a pure logic programming language would also make a good fit, but I haven't (really) done much logic programming.
But then again, map and filter are axioms in ZFC and they appear in roughly speaking all functional languages as well, so maybe there's something to it :)
So now it's only "really slow" as opposed to "really really slow"?
It's really really simple: a speed improvement of a factor x means you multiply the number of "really"s by 1/x.
Had it been three times as fast, it would have been merely "real slow".
If you're ever reduced (hah!) to fractions of a single letter, just scale the ASCII value of that letter. So if it was 24 (=2*12) times as fast, it would be half an r, or "8 slow".
In fact, I test it on all 4 platforms on a regular basis.
Why test? You only have to write once, and then it runs everywhere...
What do you mean "cut down on the kool aid"?? ;)