Nostalgia will warp your judgement as to whether things have improved or not. In 1994, when Doom was the shiny new video game and you were 13 years younger and had never seen a 3D-ish FPS, Doom was pretty impressive. Doom 3 is only less impressive because the FPS was 10+ years old at that point.
Asking "how much better was Doom 3" is seriously missing the point. How much better was the movie "Apollo 13" then "2001 A Space Odyssey"? Although superficially similar, they're in different genres. The older one has special effects so painful that I don't really want to watch it again. When they came out, they were both legitimately very good.
If an operating system in 2007 can't run an app or two reasonably smoothly on a gig of RAM, that would be pretty sad. But, that doesn't mean that a system with a flashy interface like Vista wouldn't run snappier with more.
Unless you've actually compared the performance between 1 gig and 4 gigs of RAM, you aren't in a position to say that 4 gigs isn't the "sweet spot".
The important thing for the reader here is that we're doing a sort in descending order. Normally it would be ascending order, so descending order is special. If we're sorting strings, it's *obviously* alphabetical. All we care about "heap sort" is that it's a stable sort - stable / unstable should be specified in the documentation of the sort function and we can check if we care.
Paying "the industry" "compensation" for a personal copying provision is seriously wrong-headed. It's based on an assumption that they innately deserve to get money - they don't. If they can make money by offering some product in the free market, that's great. If they can't, that sucks - but I don't get free money either.
I can generate hundreds of setters and getters via Eclipse in a matter of seconds.
I consider this a failure of the programming language you're working in, rather than a beneficial feature of your IDE.
Languages like Java (the other offender being C#) *do* punish the programmer for working without a complex IDE. They require like a million lines of code before you can even start programming. They require method autocomplete because they have methods like "my_array.sortThisArrayInDescendingAlphabeticalOrd erUsingHeapSort()". They require that your editor supports having multiple files open because no non-trivial functionality can be implemented in a single file.
If you're working in C, or Perl, or Ruby, or Python, or Haskel, or even C++ then vi works really well. Sure, it's possible to argue that a programming language should be more verbose than Perl, but getting much more unnecessarily verbose and strict about form than Python is probably a bad idea.
There's a lot of human variability, and that includes mental capacity.
If you had the option to eliminate the genes that made men weigh over 200 pounds, would that be a good deal? It would definitely get rid of the male obesity problem. It would reduce deaths due heart disease. It would also make football and heavyweight boxing pretty lame, and it would reduce the physical labor capacity of the population immensely - sometimes a heavily muscled construction worker type is damn useful.
Sure, it's obvious that full blown autism is strictly bad - in extreme cases an autistic person my never develop language skills.
That's very different from "mild autistic spectrum disorders", where the symptoms seem to consist entirely of being intelligent and hating social interaction (at least that's how it's portrayed on the intertubes). If those are really the symptoms, that's not something we want to be fixing - people like that are extremely socially valuable.
Forcing a child to live their lives without a sense of hearing, or with autistism, or with any other treatable disability, is truely criminal.
You have to be careful - generalizing doesn't always work.
You'll get no argument from me that deafness, downs syndrome, or full-on autism are obviously detrimental compared to hearing / full mental function. The problem comes with the less obvious cases... attention deficit disorder for example, or color blindness.
I've got a friend who is red-green color blind, and we frequently play games (think "capture the flag") outside at night. I'm pretty good at operating in the dark, but this friend of mine can see clearly when I could swear it was pitch black out. That's not a disability, that's a functionality tradeoff - and I'm not sure the way my eyes work is more useful. If parents had the choice for their children of distinguishing specific shades of red vs. green or superior night vision, I wouldn't condemn them for either choice - but I would say that eliminating either from the gene pool entirely would be criminal.
Canotical provides Ubuntu support on clients and servers. If you want some ridiculous cluster thing, that requires a differently structured OS anyway. I'm not seeing what the problem would be with Ubuntu on Dell - the lack of a license fee would save a noticeable amount of money.
It's possible that you're deficient mentally at some specific task that the Blender interface assumes humans are good at - focusing on what you need out of a larger set of info might be it. That would suck for you, but unless it can be shown to be common it's not worth changing the interface over.
If so, you get the standard special-ed speech "Suck it up and cope, you're probably exaggerating your problem". Sorry. =P
Power users who assume they can just muddle their way through are the greatest enemy of any significant UI improvements in *any* software. Sometimes you can do that. Sometimes you have to suck it up and read a manual. Sometimes you even have to actively train yourself in something. That's OK... some things aren't trivial.
For a pure-software system like many DRM schemes, once one person has cracked the system they can release a program to do it. After that, cracking additional instances of that system is free.
With a well built theft-deterrent system based on tamper proof hardware, you've got to do a hardware crack every time. If you can make the hardware crack difficult enough, you can make it so that it requires a competent hardware guy hours for each and every unit that needs to get cracked. Think about modchipping x-boxes - it's not super-expensive, but it'll never be free.
In order for the OLPC security system to work, the cost of the crack only needs to stay more expensive than the resale value of the device. That's a much more feasible proposition than a software security system or an "impossible to crack" piece of tamper-resistant hardware.
In this case the security system is intended to protect *physical hardware* not data or tampering. They don't have to make it theoretically impossible to break, they just have to make it significantly more expensive to crack than the resale value of the device - that's damn easy.
You're underestimating how absurdly rich the first successful astroid mining company will get.
Interstellar travel *will* get cheaper, simply because the required resources will become more plentiful once someone commercializes interplanetary travel. And, aside from resources, remember: technology doesn't get worse, unless the religious fanatics take over and throw us into another dark age.
How would you accellerate something that big that fast.
We could do it today with nuclear pulse propulsion. It doesn't work especially well on an inhabited planet due to fallout, but it works fine for interplanetary / interstellar travel.
There are a *lot* of people in the USA and Europe with net worths in the $100,000+ range. Everyone who owns a house. If it costs $10,000,000,000 to fund a space colonization expedition, it only takes 10,000 to liquidate and move. 10,000 is getting in the order of magnitude of a functional breeding population, so that's about right...
Face it, ANYTHING can be cracked if you try hard enough.
That's a nice hope, but it's not true. Any file based DRM that is functional on general purpose computers can be cracked. This is a security solution rather than DRM, and it's implemented on custom hardware.
Friend of mine hates the way Ubuntu seperates things into a main package, a -dev package, a -docs package, etc.. There should be a --fat-packages option that he can use to combine all them together.. and yeah, this should be in the gui package installers too.
No, no, and oh god no. There should not be an utterly insignificant feature added that clutters up the UI unnecessarily and complicates the package system simply because your friend has an irrational dislike of having to check 3 boxes in the package manager. It's *possible* that there should be a "consider suggested packages as dependencies" checkbox in Synaptic next to "consider recommended packages as dependencies", but there shouldn't be a modification in the handling of packages nor a major UI complication as the result of some guy's arbitrary preference.
Why is there this obsession with the awful Windows package system? Have you legitimately used a repository-based system with a GUI?
Setup.exe + an Uninstall menu item is strictly worse than, say, the way packages work in Ubuntu. If you want to just distribute a package file and have the user double click to install, that works great. But... there's also a giant fully-supported package repository.
I guess it basically comes down to one thing: As they would say on Fark.com... "No you can't have Linux be an exact copy of your favorite version of Windows. Not yours. (Picture of pony)"
I guess if a "dose" of artificial sweetener is actually a reasonable number, it's fine. I put down quite a bit of aspartame. If a "dose" is what's in an 8-ounce serving of soda, I easily put down 16 'doses' some days.
But until the science of genetic manipulation is (close to)perfected, all they are doing is 'fooling' with it.
Science doesn't get perfected until quite a bit of "fooling" gets done... even after that, "perfected" is rarely the right word.
Sure... if I were world dictator I'd probably shoot for more lab time and slower market penetration for some of these GM products, but given the choice between nessisary research getting funded by agribusiness and public phobia of science holding it back, I'll chose the research over the Frankenstein myth every day.
If 1000 times the normal dosage of an artificial sweetener produces a significant increase in cancer risk, that's *damn well* worthy of note. Three orders of magnitude isn't quite good enough for me there. Eight orders of magnitude? Sure - but 3 isn't a good enough margin of error.
Slashdot Mirror
Nostalgia will warp your judgement as to whether things have improved or not. In 1994, when Doom was the shiny new video game and you were 13 years younger and had never seen a 3D-ish FPS, Doom was pretty impressive. Doom 3 is only less impressive because the FPS was 10+ years old at that point.
Asking "how much better was Doom 3" is seriously missing the point. How much better was the movie "Apollo 13" then "2001 A Space Odyssey"? Although superficially similar, they're in different genres. The older one has special effects so painful that I don't really want to watch it again. When they came out, they were both legitimately very good.
If an operating system in 2007 can't run an app or two reasonably smoothly on a gig of RAM, that would be pretty sad. But, that doesn't mean that a system with a flashy interface like Vista wouldn't run snappier with more.
Unless you've actually compared the performance between 1 gig and 4 gigs of RAM, you aren't in a position to say that 4 gigs isn't the "sweet spot".
I agree. The important thing is not to enforce a general policy in either direction for questionable edge cases.
You're off by a couple months. The high end Athlon FX processors have been dual core for a while now.
http://www.amdcompare.com/us-en/desktop/details.as px?opn=ADAFX70GAA6DI
s px?opn=ADAFX62IAA6CS
http://www.amdcompare.com/us-en/desktop/details.a
Note how it says "1MB x2" for L2 cache.
I'm going to go with the python-esque:
The important thing for the reader here is that we're doing a sort in descending order. Normally it would be ascending order, so descending order is special. If we're sorting strings, it's *obviously* alphabetical. All we care about "heap sort" is that it's a stable sort - stable / unstable should be specified in the documentation of the sort function and we can check if we care.
Paying "the industry" "compensation" for a personal copying provision is seriously wrong-headed. It's based on an assumption that they innately deserve to get money - they don't. If they can make money by offering some product in the free market, that's great. If they can't, that sucks - but I don't get free money either.
I consider this a failure of the programming language you're working in, rather than a beneficial feature of your IDE.
Languages like Java (the other offender being C#) *do* punish the programmer for working without a complex IDE. They require like a million lines of code before you can even start programming. They require method autocomplete because they have methods like "my_array.sortThisArrayInDescendingAlphabeticalOrd erUsingHeapSort()". They require that your editor supports having multiple files open because no non-trivial functionality can be implemented in a single file.
If you're working in C, or Perl, or Ruby, or Python, or Haskel, or even C++ then vi works really well. Sure, it's possible to argue that a programming language should be more verbose than Perl, but getting much more unnecessarily verbose and strict about form than Python is probably a bad idea.
There's a lot of human variability, and that includes mental capacity.
If you had the option to eliminate the genes that made men weigh over 200 pounds, would that be a good deal? It would definitely get rid of the male obesity problem. It would reduce deaths due heart disease. It would also make football and heavyweight boxing pretty lame, and it would reduce the physical labor capacity of the population immensely - sometimes a heavily muscled construction worker type is damn useful.
Sure, it's obvious that full blown autism is strictly bad - in extreme cases an autistic person my never develop language skills.
That's very different from "mild autistic spectrum disorders", where the symptoms seem to consist entirely of being intelligent and hating social interaction (at least that's how it's portrayed on the intertubes). If those are really the symptoms, that's not something we want to be fixing - people like that are extremely socially valuable.
You have to be careful - generalizing doesn't always work.
You'll get no argument from me that deafness, downs syndrome, or full-on autism are obviously detrimental compared to hearing / full mental function. The problem comes with the less obvious cases... attention deficit disorder for example, or color blindness.
I've got a friend who is red-green color blind, and we frequently play games (think "capture the flag") outside at night. I'm pretty good at operating in the dark, but this friend of mine can see clearly when I could swear it was pitch black out. That's not a disability, that's a functionality tradeoff - and I'm not sure the way my eyes work is more useful. If parents had the choice for their children of distinguishing specific shades of red vs. green or superior night vision, I wouldn't condemn them for either choice - but I would say that eliminating either from the gene pool entirely would be criminal.
You legitimately support a tax to pay a for-profit corporation for their commercial product?
What next? A tax to pay Nike for sneakers, with everyone getting 12 tax-funded pairs every year?
Canotical provides Ubuntu support on clients and servers. If you want some ridiculous cluster thing, that requires a differently structured OS anyway. I'm not seeing what the problem would be with Ubuntu on Dell - the lack of a license fee would save a noticeable amount of money.
Other people seem to be able to handle it.
It's possible that you're deficient mentally at some specific task that the Blender interface assumes humans are good at - focusing on what you need out of a larger set of info might be it. That would suck for you, but unless it can be shown to be common it's not worth changing the interface over.
If so, you get the standard special-ed speech "Suck it up and cope, you're probably exaggerating your problem". Sorry. =P
Power users who assume they can just muddle their way through are the greatest enemy of any significant UI improvements in *any* software. Sometimes you can do that. Sometimes you have to suck it up and read a manual. Sometimes you even have to actively train yourself in something. That's OK... some things aren't trivial.
Wait a second... what's wrong with programming in vi? It's way easier to use than some random IDE with an overly-complex GUI.
I mean... a Unix CLI *is* an IDE - and it's not like a programmer should be complaining about needing idiot-proof tools.
For a pure-software system like many DRM schemes, once one person has cracked the system they can release a program to do it. After that, cracking additional instances of that system is free.
With a well built theft-deterrent system based on tamper proof hardware, you've got to do a hardware crack every time. If you can make the hardware crack difficult enough, you can make it so that it requires a competent hardware guy hours for each and every unit that needs to get cracked. Think about modchipping x-boxes - it's not super-expensive, but it'll never be free.
In order for the OLPC security system to work, the cost of the crack only needs to stay more expensive than the resale value of the device. That's a much more feasible proposition than a software security system or an "impossible to crack" piece of tamper-resistant hardware.
That's actually pretty funny.
In this case the security system is intended to protect *physical hardware* not data or tampering. They don't have to make it theoretically impossible to break, they just have to make it significantly more expensive to crack than the resale value of the device - that's damn easy.
You're underestimating how absurdly rich the first successful astroid mining company will get.
Interstellar travel *will* get cheaper, simply because the required resources will become more plentiful once someone commercializes interplanetary travel. And, aside from resources, remember: technology doesn't get worse, unless the religious fanatics take over and throw us into another dark age.
We could do it today with nuclear pulse propulsion. It doesn't work especially well on an inhabited planet due to fallout, but it works fine for interplanetary / interstellar travel.
How cheap does it really have to get?
There are a *lot* of people in the USA and Europe with net worths in the $100,000+ range. Everyone who owns a house. If it costs $10,000,000,000 to fund a space colonization expedition, it only takes 10,000 to liquidate and move. 10,000 is getting in the order of magnitude of a functional breeding population, so that's about right...
That's a nice hope, but it's not true. Any file based DRM that is functional on general purpose computers can be cracked. This is a security solution rather than DRM, and it's implemented on custom hardware.
No, no, and oh god no. There should not be an utterly insignificant feature added that clutters up the UI unnecessarily and complicates the package system simply because your friend has an irrational dislike of having to check 3 boxes in the package manager. It's *possible* that there should be a "consider suggested packages as dependencies" checkbox in Synaptic next to "consider recommended packages as dependencies", but there shouldn't be a modification in the handling of packages nor a major UI complication as the result of some guy's arbitrary preference.
Why is there this obsession with the awful Windows package system? Have you legitimately used a repository-based system with a GUI?
Setup.exe + an Uninstall menu item is strictly worse than, say, the way packages work in Ubuntu. If you want to just distribute a package file and have the user double click to install, that works great. But... there's also a giant fully-supported package repository.
I guess it basically comes down to one thing: As they would say on Fark.com... "No you can't have Linux be an exact copy of your favorite version of Windows. Not yours. (Picture of pony)"
I guess if a "dose" of artificial sweetener is actually a reasonable number, it's fine. I put down quite a bit of aspartame. If a "dose" is what's in an 8-ounce serving of soda, I easily put down 16 'doses' some days.
Science doesn't get perfected until quite a bit of "fooling" gets done... even after that, "perfected" is rarely the right word.
Sure... if I were world dictator I'd probably shoot for more lab time and slower market penetration for some of these GM products, but given the choice between nessisary research getting funded by agribusiness and public phobia of science holding it back, I'll chose the research over the Frankenstein myth every day.
If 1000 times the normal dosage of an artificial sweetener produces a significant increase in cancer risk, that's *damn well* worthy of note. Three orders of magnitude isn't quite good enough for me there. Eight orders of magnitude? Sure - but 3 isn't a good enough margin of error.
Like $8, because the patent would have expired and someone else could make the drug.