This really isn't as big a deal as you make it out to be. Unix-ish systems have versioned libraries, which means that multiple versions of the same library can be installed in parallel. Further, it's possible to do what Windows developers do with their apps - just include a copy of the library with the app.
I don't run that many apps that aren't in the Ubuntu package repository. The only two I have installed are Unreal Tournament 2004 and Google Earth. Both apps just work, even though UT2004 is two years old (and therefore would have mad library problems if there really was a library compatibility issue).
Occasionally I decide I want to dig out Loki's Sid Meier's Alpha Centauri. It's using some ridiculous old version of libc, so I have to install an old-libc compatibility package. It's there in the repository, so I just have to fire up the package manager and grab one package to get my game working. The fact that installing a six year old binary application is that simple pretty much debunks the "hard target for ISVs" claim - my distro didn't even *exist* when the app was released.
The way the Ubuntu development cycle works is that they sync with Debian unstable at the beginning of their six month release cycle. Then they spend the next 6 months getting things stable (and ready for the last-minute Gnome drop).
Debian unstable is reasonably stable anyway (I ran it as a desktop for a year with no major issues), but Ubuntu is really a polished and stable distro fully ready for production use rather than just a copy of Debian's development branch.
The system as it works now has issues, but at least it works.
Making a law requiring that half of all university tuitions be spent on medical research would be a system that "has issues yet still works". My threshold for things being complaint worthy is way lower than that - down near "it works, and works well".
Have you actually test driven the hybrid? I haven't tried the hybrid Civic, but some of the other hybrids are pretty nice performance-wise. The improved torque at low speeds is *really* nice.
As a pedestrian, when trying to cross the road, I put some effort into looking like I'm completely oblivious to the cars in the road and just walking into the road like an idiot trying to get myself killed. This works way better at getting them to stop and let me cross than stopping and staring at them does. I do make sure that if the cars don't stop for me I'm not actually going to be in the way, but watching a car out of the corner of my eye while blatantly looking the other way and walking towards the road works wonders to getting them to stop.
You ask, basically, "Without patents, who will fund new drug research?"
That's a really interesting question, and I wish I had a good clean answer for you. First - and most importantly - I highly doubt that the pharmaceutical companies would outright stop their research efforts. A company needs new products, and the first-to-market advantage is non-trivial.
It would mean a significant tax increase if you think the government should pay for it.
That's a possible solution, and it might even be a good one. It's definitely not the only possible solution.
My first question is: Who's really paying for useful drug research today? Is it really the pharmaceutical companies encouraged by patents, or is it other groups for other reasons?
My next question is: How much money do we save if all current drugs can be produced as royalty-free generics?
I seriously don't think that there's any real risk of drug research stopping. People are too attached to "making the world a better place" - the work will get funded by private grants if necessary.
Now, I absolutely agree that Marijuana should be legal. Our taxes are higher due to an unwinnable fight against the United States' largest cash crop. If it were legal and taxed, my taxes would be lower twice: once because I wouldn't have to pay for the "war on drugs", and once because of the tax on pot.
As for smoking pot being a good idea to "expand your mind", that's utter bullshit unless your neurochemistry is completely different from anyone I know. Marijuana increases appetite, dulls pain, and helps people with ADD concentrate. It also makes people who don't have ADD worse at learning new things and abstract thought in general. So... Yes, it should be legal; No, people shouldn't be encouraged to use it more recreationally.
You are the worst economist / social architect ever.
The "pick a solution arbitrarily and mandate it" technique that you favor is not an especially effective way to solve problems. The problems you're thinking of are possible to solve, but your technique isn't going to do it well.
You've been confused by the flawed "intellectual property" terminology. The problem is that by treating medical techniques as property, people are killed. The solution isn't to continue to stretch the already-broken property metaphor (i.e. "eminent domain"), the solution is to stop trying to treat ideas as property.
Very simply, allowing pharmaceutical patents in poor countries is murder.
When I click on the [Watch FSF's year end video appeal.] link on http://www.fsf.org/ with my embedded media player, it plays instantly... like flash video, except it takes up the whole page and is way better quality that most flash video you see around. That's a theora video, and I'm not seeing the downside.
Normally the commonly used and simple stuff goes in the GUI, and the rarely used and complex stuff is only accessible through the config file. This is true for sudo on Ubuntu, this is true for Firefox on any platform, this is true for many Windows programs (with the registry).
The fact that there's frequently more powerful options available in config files on Unix-like platforms compared to Windows is a feature, not a bug.
There must have been an issue with SATA for like 10 minutes in some weird distro and I missed it or something. I've been buying SATA harddrives and using them with Linux systems since they became widely available, and I've *never* had an issue with it. In fact, I've had more trouble installing Windows on SATA drives - simply because Windows XP was released before the SATA standard.
As for applications, and Photoshop vs. The GIMP - that's the most overhyped argument ever. Sure... some people are addicted to Photoshop's specific interface to such a large extent that even GimpShop can't help them. Sucks to be them. Most people don't even use photo editing software. Of the people who do, The GIMP is more than good enough for most of them. For those few people who require native CMYK color or can't handle an option being under a different menu... well, I hear Photoshop runs pretty well under Windows.
How often have you seen this as a generic response to a problem the user does have to care about, particularly when the user is the administrator?
Recently, I hear the complaint applied incorrectly significantly more frequently than I've seen that actual problem. Realistically, tweaking config files in a modern user-oriented Linux distro needs to come up just about as frequently as tweaking registry entries under Windows comes up - i.e. never for normal users.
Basically: Don't use Gentoo or NetBSD as a normal user and don't try to configure your cross-platform network to use IPv6 tunneled over Appletalk, and you'll be fine.
Some distributions target desktop users, others do not.
Ubuntu, CentOS, OpenSuSE, SLED, RHEL D, maybe Mandriva - that's more than enough desktop distros. Stuff like Gentoo is no more relevant to a non-expert computer user than NetBSD or Plan 9.
Installation issues that severe just don't happen without broken hardware. Having to swap out hard drives is exactly the sort of nightmare scenario that Linux installs didn't cause even in 1997.
My last two installs Nvidia drivers worked flawlessly. Installing a dual boot setup on my cousin's computer, the NTFS partition was automatically detected and showed up as an icon on the desktop.
Beryl? Wine? I have an idea... rather than using this nice stable and functional desktop that you've just installed, how about you check the forums for things that people are having trouble with and try installing them so your computer breaks? Beryl is one of those fun things to do when you're willing to break everything for pretty pictures... i.e., it's not ready for the inexperienced yet. I'm not sure why anyone would install Wine at all. If you want to run Windows programs, why are you in Linux?
Linux is more than ready for users who want to accomplish things with their computer. It's never going to be "ready" for users who think they can install the newest unsupported alpha-quality hack like Beryl and not run into problems. I suggest you go install Project Looking Glass and Litestep on Windows at the same time and see if you run into problems.
Debian *always* is having internal infighting. Always. It's one of the things that slows down their releases a bit, but it's nothing that you shouldn't be expecting after nearly 10 years of it. It doesn't change the fact that what they do release is one of the most stable and easy to administer distros out there.
RedHat is all that's left.
There are two other major distros that you forgot: Ubuntu and Mandriva. I haven't really been watching Mandriva (ever), but they still seem to be going strong. Ubuntu, on the other hand, remains my top choice for a full featured Linux distro. It's got almost the stability of Debian, with all the package management power, better hardware support, and worldwide enterprise-grade commercial support from multiple vendors.
Re:It's too early to discount Oracle/MS/Novell
on
Red Hat Sales Surge
·
· Score: 2, Informative
Another one of the unique abilities of FOSS is the fact that it allows you to rely on the work of others. Rather than screwing around with Linux From Scratch (which is an amusing thing to do once as a hobbyist, not a serious business solution) it's perfectly possible to chose another Linux vendor with a better product pricing model... say Canonical with Ubuntu. If they're more attached to the Red Hat model than they are to decent external support, something like CentOS might be appropriate. There's no need for people to be hand-rolling their own distros.
I'm well aware that referrers aren't a reliable source of information to make security decisions. That's not what I was suggesting. I was suggesting that the referrer header usually *does* contain valid information about what page is embedding content - and can be used to server alternate offensive content to specific unwanted embedders. If the offensive content isn't enough to discourage them from embedding your content, you can then resort to other techniques to block it technologically - but the "Goatse Guy / THIS WEBSITE STEALS BANDWIDTH" image replacement technique almost always works.
If you use the actual tactic that I suggested, rather than assuming that I'm dumb and think that the referrer header is something it isn't, this isn't a problem.
Again: If you detect that the HTTP request is coming as the result of being embedded in a specific, known, site that you don't want embedding your content and ship something wildly offensive and vomit inducing instead, the problem will go away.
You see... that's where you're wrong. If he wants to communicate the true and useful information that everything but page 7 is ads, you damn well don't have a right to legally stop him.
This still isn't a legal issue. It's not hard to set up a referrer check that ships along some shit porn or something to viewers from the unwanted website.
Slashdot Mirror
This really isn't as big a deal as you make it out to be. Unix-ish systems have versioned libraries, which means that multiple versions of the same library can be installed in parallel. Further, it's possible to do what Windows developers do with their apps - just include a copy of the library with the app.
I don't run that many apps that aren't in the Ubuntu package repository. The only two I have installed are Unreal Tournament 2004 and Google Earth. Both apps just work, even though UT2004 is two years old (and therefore would have mad library problems if there really was a library compatibility issue).
Occasionally I decide I want to dig out Loki's Sid Meier's Alpha Centauri. It's using some ridiculous old version of libc, so I have to install an old-libc compatibility package. It's there in the repository, so I just have to fire up the package manager and grab one package to get my game working. The fact that installing a six year old binary application is that simple pretty much debunks the "hard target for ISVs" claim - my distro didn't even *exist* when the app was released.
The way the Ubuntu development cycle works is that they sync with Debian unstable at the beginning of their six month release cycle. Then they spend the next 6 months getting things stable (and ready for the last-minute Gnome drop).
Debian unstable is reasonably stable anyway (I ran it as a desktop for a year with no major issues), but Ubuntu is really a polished and stable distro fully ready for production use rather than just a copy of Debian's development branch.
Making a law requiring that half of all university tuitions be spent on medical research would be a system that "has issues yet still works". My threshold for things being complaint worthy is way lower than that - down near "it works, and works well".
Have you actually test driven the hybrid? I haven't tried the hybrid Civic, but some of the other hybrids are pretty nice performance-wise. The improved torque at low speeds is *really* nice.
As a pedestrian, when trying to cross the road, I put some effort into looking like I'm completely oblivious to the cars in the road and just walking into the road like an idiot trying to get myself killed. This works way better at getting them to stop and let me cross than stopping and staring at them does. I do make sure that if the cars don't stop for me I'm not actually going to be in the way, but watching a car out of the corner of my eye while blatantly looking the other way and walking towards the road works wonders to getting them to stop.
You ask, basically, "Without patents, who will fund new drug research?"
That's a really interesting question, and I wish I had a good clean answer for you. First - and most importantly - I highly doubt that the pharmaceutical companies would outright stop their research efforts. A company needs new products, and the first-to-market advantage is non-trivial.
That's a possible solution, and it might even be a good one. It's definitely not the only possible solution.
My first question is: Who's really paying for useful drug research today? Is it really the pharmaceutical companies encouraged by patents, or is it other groups for other reasons?
My next question is: How much money do we save if all current drugs can be produced as royalty-free generics?
I seriously don't think that there's any real risk of drug research stopping. People are too attached to "making the world a better place" - the work will get funded by private grants if necessary.
Now, I absolutely agree that Marijuana should be legal. Our taxes are higher due to an unwinnable fight against the United States' largest cash crop. If it were legal and taxed, my taxes would be lower twice: once because I wouldn't have to pay for the "war on drugs", and once because of the tax on pot.
As for smoking pot being a good idea to "expand your mind", that's utter bullshit unless your neurochemistry is completely different from anyone I know. Marijuana increases appetite, dulls pain, and helps people with ADD concentrate. It also makes people who don't have ADD worse at learning new things and abstract thought in general. So... Yes, it should be legal; No, people shouldn't be encouraged to use it more recreationally.
You are the worst economist / social architect ever.
The "pick a solution arbitrarily and mandate it" technique that you favor is not an especially effective way to solve problems. The problems you're thinking of are possible to solve, but your technique isn't going to do it well.
Patents are not property, and should not be considered as such.
The correct solution here is to change the patent law to make it no longer cover drugs. That will solve the problem very simply.
You've been confused by the flawed "intellectual property" terminology. The problem is that by treating medical techniques as property, people are killed. The solution isn't to continue to stretch the already-broken property metaphor (i.e. "eminent domain"), the solution is to stop trying to treat ideas as property.
Very simply, allowing pharmaceutical patents in poor countries is murder.
When I click on the [Watch FSF's year end video appeal.] link on http://www.fsf.org/ with my embedded media player, it plays instantly... like flash video, except it takes up the whole page and is way better quality that most flash video you see around. That's a theora video, and I'm not seeing the downside.
Normally the commonly used and simple stuff goes in the GUI, and the rarely used and complex stuff is only accessible through the config file. This is true for sudo on Ubuntu, this is true for Firefox on any platform, this is true for many Windows programs (with the registry).
The fact that there's frequently more powerful options available in config files on Unix-like platforms compared to Windows is a feature, not a bug.
There must have been an issue with SATA for like 10 minutes in some weird distro and I missed it or something. I've been buying SATA harddrives and using them with Linux systems since they became widely available, and I've *never* had an issue with it. In fact, I've had more trouble installing Windows on SATA drives - simply because Windows XP was released before the SATA standard.
As for applications, and Photoshop vs. The GIMP - that's the most overhyped argument ever. Sure... some people are addicted to Photoshop's specific interface to such a large extent that even GimpShop can't help them. Sucks to be them. Most people don't even use photo editing software. Of the people who do, The GIMP is more than good enough for most of them. For those few people who require native CMYK color or can't handle an option being under a different menu... well, I hear Photoshop runs pretty well under Windows.
Recently, I hear the complaint applied incorrectly significantly more frequently than I've seen that actual problem. Realistically, tweaking config files in a modern user-oriented Linux distro needs to come up just about as frequently as tweaking registry entries under Windows comes up - i.e. never for normal users.
Basically: Don't use Gentoo or NetBSD as a normal user and don't try to configure your cross-platform network to use IPv6 tunneled over Appletalk, and you'll be fine.
Some distributions target desktop users, others do not.
Ubuntu, CentOS, OpenSuSE, SLED, RHEL D, maybe Mandriva - that's more than enough desktop distros. Stuff like Gentoo is no more relevant to a non-expert computer user than NetBSD or Plan 9.
This is an excellent example of anti-Linux FUD.
Installation issues that severe just don't happen without broken hardware. Having to swap out hard drives is exactly the sort of nightmare scenario that Linux installs didn't cause even in 1997.
My last two installs Nvidia drivers worked flawlessly. Installing a dual boot setup on my cousin's computer, the NTFS partition was automatically detected and showed up as an icon on the desktop.
Beryl? Wine? I have an idea... rather than using this nice stable and functional desktop that you've just installed, how about you check the forums for things that people are having trouble with and try installing them so your computer breaks? Beryl is one of those fun things to do when you're willing to break everything for pretty pictures... i.e., it's not ready for the inexperienced yet. I'm not sure why anyone would install Wine at all. If you want to run Windows programs, why are you in Linux?
Linux is more than ready for users who want to accomplish things with their computer. It's never going to be "ready" for users who think they can install the newest unsupported alpha-quality hack like Beryl and not run into problems. I suggest you go install Project Looking Glass and Litestep on Windows at the same time and see if you run into problems.
We're talking about the specifics of configuring sudo here... not something a user needs to care about.
Very simply... Users don't need to do fine-grained privilege delegation.
If you're an admin, why do you care about a GUI? I know I damn well wouldn't hire an admin who couldn't read a manpage and edit a config file.
Unnecessary replication of effort is a waste. If you need a general purpose Linux distro, at least start with something like Ubuntu or CentOS.
Debian *always* is having internal infighting. Always. It's one of the things that slows down their releases a bit, but it's nothing that you shouldn't be expecting after nearly 10 years of it. It doesn't change the fact that what they do release is one of the most stable and easy to administer distros out there.
There are two other major distros that you forgot: Ubuntu and Mandriva. I haven't really been watching Mandriva (ever), but they still seem to be going strong. Ubuntu, on the other hand, remains my top choice for a full featured Linux distro. It's got almost the stability of Debian, with all the package management power, better hardware support, and worldwide enterprise-grade commercial support from multiple vendors.
Another one of the unique abilities of FOSS is the fact that it allows you to rely on the work of others. Rather than screwing around with Linux From Scratch (which is an amusing thing to do once as a hobbyist, not a serious business solution) it's perfectly possible to chose another Linux vendor with a better product pricing model... say Canonical with Ubuntu. If they're more attached to the Red Hat model than they are to decent external support, something like CentOS might be appropriate. There's no need for people to be hand-rolling their own distros.
I'm well aware that referrers aren't a reliable source of information to make security decisions. That's not what I was suggesting. I was suggesting that the referrer header usually *does* contain valid information about what page is embedding content - and can be used to server alternate offensive content to specific unwanted embedders. If the offensive content isn't enough to discourage them from embedding your content, you can then resort to other techniques to block it technologically - but the "Goatse Guy / THIS WEBSITE STEALS BANDWIDTH" image replacement technique almost always works.
If you use the actual tactic that I suggested, rather than assuming that I'm dumb and think that the referrer header is something it isn't, this isn't a problem.
Again: If you detect that the HTTP request is coming as the result of being embedded in a specific, known, site that you don't want embedding your content and ship something wildly offensive and vomit inducing instead, the problem will go away.
You see... that's where you're wrong. If he wants to communicate the true and useful information that everything but page 7 is ads, you damn well don't have a right to legally stop him.
This still isn't a legal issue. It's not hard to set up a referrer check that ships along some shit porn or something to viewers from the unwanted website.