they choose not to use the particular software distribution method of your particular distro, and there are a dozen out there so you can hardly blame them.
My distro is Ubuntu, and most software at least contains instructions for how to install on Ubuntu.
Again, you're suggesting standardizing around something -- why not the debian package manager?
I think we could do better, but I think as it stands, apt and dpkg are already better than Windows. And Ubuntu is by far the most popular distro, so if you were going to target a distro, it seems like this would be the one to hit.
Then again, there are tools which will automatically build a deb, an rpm, and a tarball.
Bullshit. If the app in question does not use an install package, then it is completely self contained. At most it will use some files within its own directory. If you need to make registry hacks or install new DLLs to make it work, it means you are copying it directly from an install folder which was created via an installer at some earlier point.
Or it means the installer was poorly written. And yes, I have had this happen.
Have you ever tried to install a Windows app and found there's no exe, or msi?
Honestly? No.
That contradicts your mention of a zipfile.
No, the right way to deal with it is come up with a simpler way to install packages without requiring the approval of your favorite corporate repository owner.
Who said anything about "corporate"? And again,
Getting all defensive about how a flaw in the useability of Linux should not be fixed for some stupid reason or another is one of the reasons the Linux is still not mainstream.
You are missing the point.
I agree, it should be better. You've pointed out some situations which really should be improved.
However, your proposed solution makes things worse, not better. I would prefer a forever non-mainstream Linux to the bastardization that would be downloading random binaries from a Google search.
If you're interested in providing constructive criticism, tell me how we can keep the strengths of the repository system while also gaining the one-click you're suggesting.
I am a very technical computer user, but I'm also not into torturing myself or doing more work than I need to just to use my damn computer.
Neither am I, it's why I switched away from Windows in the first place.
If -I- have a problem with using Linux, you can bet your ass the vast majority of non-technical users will have a problem using it as well.
Sounds good. Except it's also arrogant and wrong.
The vast majority of non-technical users don't need to be downloading random pieces of software from the Internet, and are better off not doing so. They like the App Store for the iPhone, so I don't see how add/remove programs on Ubuntu is worse.
You have a harder time because you try to do more with your computer than the average non-technical user needs.
For example: Would you ever use the Chrome OS? I wouldn't. But I'd certainly recommend it to my grandmother. Less locally-installed software means less she can screw up.
It *is* preferable to have a separate ad-blocker, that should be a no-brainer within the unix philosophy.
Granted. But there are two large problems with it, in practice:
Firstly, while it's not entirely in line with the Unix philosophy, my browser already parsed the page. Why should my proxy have to, also?
And probably more importantly: Maybe this has changed, or maybe I just used the wrong proxy, but I remember using Polipo for a lot of things. I also remember finding out that it would cause certain pages to fail (even when I was just using it as a cache), and that, perversely, it'd often slow down my browsing.
Is it really any different clicking on an apt-url from some unfamiliar website?
Yes.
An apt-url is a URL to a package that is already in the repository, which means at least there's some minimal assurance that someone associated with Ubuntu or Canonical has looked at it, and some strong assurance that it hasn't been modified since then.
What about legitimate authors who didn't jump through the hoops to get their software into all the various repositories and packaging systems? Should they not be trusted simply because they didn't sign up with Canonical?
Pretty much. Ubuntu is community-maintained, and it's really not that difficult to at least get into the "universe" repository.
Of course, unlike the iPhone, it's always possible for the user to go around this, and I agree, it could be made easier than it is -- but it is pretty easy. I gave an example somewhere here involving Chrome -- it's a deb which sets up a repository as it's installed.
Now, you could make the case that downloading a random deb from the Internet is just as bad as downloading a random exe. But the point is, the capability is there if it's needed -- and most often, it makes much more sense to use something from an already-trusted repository.
All those -dev packages? Those are libraries that need to be on your Linux box in order to make the program. If, for some reason, the program you are trying to install cannot use a newer version of a library and you need that library for other software, you're quite thoroughly fucked.
Well, no, it's just gotten harder. Generally, you can pass an argument to./configure to tell it where to find headers, and there can easily be multiple versions of a library installed -- that's why the.so files have version numbers on them.
I also can't remember the last time that happened to me.
MSI and DMG files have dependancies built in.
This has two major disadvantages:
First, it's just extra space, bandwidth, RAM, and cache. That last one is still not cheap.
And second, it makes it difficult to globally patch a vulnerability in such a library. Example: OpenSSL is likely to be used in a number of packages. The only way to patch it in all packages on, say, OS X, is to not bundle it and pray Apple includes it in the OS -- then it'll be handled by Software Update.
Updates are a single extension file (with MSI files anyway).
What does "extension file" mean, in this case?
And where is the option for me to tell my system to update all MSI-based apps I have installed?
I'm fairly sure this does not exist for OS X apps.
What would you suggest for me now? Got any more commands that will magically fix a retarded install method?
Contact the maintainer?
Your complaint here is that it doesn't work with newer libraries. But this is, again, similar to expecting a Win98 app to work in Windows 7. It often will, but sometimes it doesn't. Even when Microsoft pours blood, sweat, and tears into backwards compatibility, it's often working around a bug in the original app.
It's also worth mentioning (again) that the blame for this falls squarely on the shoulders of the app developer. Just as you would scream bloody murder if a Windows app required you to download Visual Studio Express and compile from source -- but you wouldn't be blaming Microsoft.
Of course it's not The One Thing(TM) holding Linux back, you'd have to be an idiot to think that. But you've got to have your blinders on pretty tight not to see that it would definitely improve the situation for Linux on the desktop.
I agree that the situation could be improved.
However, the methods most people suggest amount to "Copy Windows" or "Copy OS X", which would be a serious regression in many wa
That's my whole point, there is a shitload of software out there that it does not work for.
That is: There's a shitload of software out there that chooses not to work with it.
Worst case scenario for Windows is a zipped directory with some shortcuts to place on the desktop.
And some registry hacks. And some things you may have to manually compile. And a third-party tweak to make it work in Vista or Win7.
You're right that Windows is more unified. You're wrong that this is somehow the fault of Linux, or the distro, or some lack of functionality.
Again:
1. See there is no apt-url link - curse (shit is common here)
Have you ever tried to install a Windows app and found there's no exe, or msi?
That's the difference -- it's the app's own fault.
And the right way to deal with it is to go bother the app in question, not post pointless rants on Slashdot about how "Linux" is broken. It's a bit like saying Windows is broken because of identity theft.
Oh, and probably the strangest one: Suggesting a proxy-based adblocker.
Ok, I get that it's practical, for now. However, they seem to be saying this would be preferable to an adblocker built in to the browser, which makes no sense. Being able to right-click on an ad and figure out how to block it is not going to be replaced by editing some obscure config file in privoxy.
I like the idea, and I'd love to play with it a bit, but there are a few stupid design decisions:
Why don't you just use a reasonable config by default?
There really is no excuse for this. I mean, yes, I can understand where not everyone would want that "reasonable default", but that's why it's a default.
We don't want to store anything "automagically" in the users home. Some people prefer different file/directory layouts
Uhm... ~/.uzbl? How difficult is that? And if you don't like it, rm -rf ~/.uzbl!
Or just create an example script that sets up the default config, and put it in your FAQ.
We considered the option of having a global '/etc/uzbl' which user specific ones could override but that would overcomplicate things.
I'm sorry, but even mplayer is officially friendlier than uzbl. How the fuck is it "complicated" to read one config file, then another?
Uzbl itself doesn't use much gtk stuff (only the statusbar) so we could do without gtk. But Webkit needs a widget toolkit to create widgets (think javascript popups, html forms etc). Officially, it also supports QT and wxwigdets.
So, why doesn't uzbl also support these options? I'm using KDE, so Qt makes sense.
Uzbl.run( ) command is any uzbl command as defined above return value: a string, either empty or containing the output of the command. Very few commands return their output currently, including js, script, and print.
They obviously realize that JS runs in a single thread. So the obvious implementation here would be to use a callback, not a return value, so you don't block the entire page while you run that script.
I mean, I want to like it, but that's a number of facepalms right off the bat, so I think I'll stick with Chrome until I have time to fix them.
Why would anybody develop a repository when all you have to do is Google it, clicky click the pretty blue link, and lo and behold it's installed (ignoring the plethora of security checks of course
Because that's really fucking insecure, compared to a repository?
Especially with.msi installs, way beyond anything I could find in Linux.
What does an MSI do that Linux package managers don't?
Because I can think of things Linux package managers do that MSIs don't.
There are a shitload of applications out there that do not exist in either, and if you want them they suck monkey testicles to install.
That's the fault of the application, not of the distro.
For example, try installing Chrome on Ubuntu -- you click either the 32-bit link or the 64-bit link, and it Just Works. It also sets up a Google repository so that it will continue to auto-update, through the package manager.
Your complaint would be somewhat like complaining because some Windows app demanded that you download and install Visual Studio Express first, then downloaded and compiled the source for it and several libraries it needs. It's certainly possible to make installation easier, but it's up to the app to do so.
Why hasn't anybody come up with the Linux equivalent of the.dmg file?
Linux had disk images for long before OS X did, and it's a terrible way of distributing software.
Seriously, I think you've been brainwashed a bit by the Apple flashiness. Think about what's going on here:
Click download link
Open DMG file
Drag app from DMG to Applications
Drag DMG drive to trash
Drag DMG file to trash
Empty trash
Every non-technical Mac user I know never gets to step 3. They get to step 2, say "Oh, there's my app!" and double-click it. Which means they run Firefox out of a disk image for years at a time, and never upgrade, because the DMG is read-only and Firefox won't auto-update itself.
Compare to:
Click the apt-url link
Follow the instructions
There is no step 3, nor any files to clean up.
In other words, the reason there isn't the "Linux equivalent to the DMG file" is because what we have actually is better, and easier to use, when it works.
For heaven's sake, it's 2009, why the hell do I have to friggin compile every damn piece of software that isn't in a repository?
Mostly because the people who provided that software didn't know what they were doing. I just untar things not in a repository.
For heaven's sake, it's 2009, why am I downloading and executing random, untrusted binaries from the Internet as the standard way to install software on Windows?
As it is now you have to dump a tarball into directory, run a few scripts while crossing your fingers that all your libraries match up, then make the binary.
If by "run a few scripts" you mean "exactly the same scripts for every single package", fine. I mean, it's going to be./configure && make && make install.
Where I think you're confused is the library dependencies. Generally, when I download the source package, they'll also give me a list of '-dev' packages I need to install on Ubuntu. And it's a hell of a lot easier to do than trying to compile something on Windows.
As wonderful as the repository idea is - frankly I love that everything is right at your fingertips - it is completely unnecessary with Windows, because Google works just fine as a repository. Click the link and you're installing the program, no mess no hassle.
No security.
And if you wanted to set up a repository, it would not be hard, it would be little more than a database of.msi files, which install automatically.
And dependencies. And reverse-dependencies. And automatic updates. And third-party repositories.
No, if you wanted to do it right, you'd probably start with Windows Update -- except Microsoft has that locked down against third parties.
Frankly, some kind of unified one-step scripted install structure, preferably all in a single container, that actually worked as intended would catapult linux on the desktop by leaps and bounds.
Bullshit.
See, there's really no way such a system would see any wider adoption than any existing package manager. In other words, we already have this, and it's called dpkg. And you can, in fact, click a dpkg in a web browser, and expect it to work -- or even click an apturl, and it'll pull it through the repository, rather than the browser.
The absolute worst any modern desktop user has to do is run some commands -- that is, copy and paste something from a website into a commandline.
Somehow, I don't think that this is The One Thing(TM) that is holding Linux back. I'm going to say it's instead the lack of applications, the less-than-perfect Wine, the fact that it's unfamiliar, and the fact that most people aren't getting it preloaded.
I can think of a few ways to make the repository system better. But they don't involve mirroring the absolutely retarded way that Windows and OS X handle software installation.
A couple of minor good uses doesn't balance out myriad terrible abuses.
I don't know... It sounds terrible, but if you think about it, would you rather have Hiroshima and space travel, or no Hiroshima and be forever earthbound?
There is such a thing as a bad feature. For example, consider two computers, completely identical, except that one randomly kicks you in the nuts, and the other does not.
Except that randomly kicking you in the nuts has no good applications. I can't think of a good application of <blink> offhand. I can, however, think of good applications of animated gifs -- pretty much anywhere I need a few frames of animation, for example, it's going to be less CPU-intensive than flipping through a set of PNGs with JavaScript.
Having a feature which can be disabled is much better than not having a feature, I would think.
For example, if my browser supports audio and video (at least through Flash), I can watch YouTube.
If it starts annoying me elsewhere, I can always use something like adblock, flashblock, etc, to disable it. But I'd much rather have the feature, if it has any good application at all, than not have the feature.
Depending on the luck of the draw, any particular device may be more or less compatible with Linux depending what was included.
And do you really think Dell would risk their reputation on "the luck of the draw"?
I'll give you a small (but annoying) example: My laptop could come with Ubuntu or with Vista. To get it with Ubuntu, I have to order it from an entirely different section of the website, with entirely different customization options -- obviously, I can't order it with Office preloaded.
The Vista version clearly comes with an 802.11n card. There's even a little "n-series" sticker on it.
But it came with an 802.11g card instead.
Want to guess why? I'm guessing it's because the G card had better Linux support.
IQ is meant to measure intelligence, not knowledge. Knowing more about history and politics doesn't make him "smarter", just better educated -- the IQ test suggests that you could have just as thorough an understanding of these things, if you cared to.
I could be wrong, but it seems to me that how good we are at something that requires intelligence is largely a function of how much attention we put in it, and how interested we are in it, and how generally intelligent we are.
Would it be possible to try this out with people in varying stages of a relationship?
That is, do we eventually recover by adjusting to the person in question, or are married men who are truly still in love necessarily dumber around their wife?
You have to do this specifically because AAC is patented.
That is: Third-party AAC libraries do exist. But they infringe on AAC patents. If software patents didn't exist, or if patents had a shorter term, reverse engineering and reimplementing these formats would be legal, no matter what the licensing restrictions.
Now go back and read my comment, and the thread in general -- this is about what would happen if there were no software patents. On the one hand, this would mean vorbis and theora might never see the light of day, and certainly wouldn't have seen what success they have. On the other hand, we'd be able to use AAC and h.264 without those licensing restrictions.
The current period (I think it's seventeen years, but I Could Be Wrong.) is the product of centuries of experience with patents.
And during those centuries, we've had more than one example -- going back as far as Watt and Hornblower, of course -- of that term, or of the existence of patents at all, being a significant detriment.
There is actually a good case that the existence of patents at all hinders innovation more than it helps.
One thing I do know, however: simplistic solutions to complex problems don't work.
"For every problem, there is a solution which is simple, neat, and wrong."
Of course, that applies to language, too. Simplistic solutions can solve complex problems, and simple problems can require complex solutions.
Slashdot Mirror
they choose not to use the particular software distribution method of your particular distro, and there are a dozen out there so you can hardly blame them.
My distro is Ubuntu, and most software at least contains instructions for how to install on Ubuntu.
Again, you're suggesting standardizing around something -- why not the debian package manager?
I think we could do better, but I think as it stands, apt and dpkg are already better than Windows. And Ubuntu is by far the most popular distro, so if you were going to target a distro, it seems like this would be the one to hit.
Then again, there are tools which will automatically build a deb, an rpm, and a tarball.
Bullshit. If the app in question does not use an install package, then it is completely self contained. At most it will use some files within its own directory. If you need to make registry hacks or install new DLLs to make it work, it means you are copying it directly from an install folder which was created via an installer at some earlier point.
Or it means the installer was poorly written. And yes, I have had this happen.
Have you ever tried to install a Windows app and found there's no exe, or msi?
Honestly? No.
That contradicts your mention of a zipfile.
No, the right way to deal with it is come up with a simpler way to install packages without requiring the approval of your favorite corporate repository owner.
Who said anything about "corporate"? And again,
Getting all defensive about how a flaw in the useability of Linux should not be fixed for some stupid reason or another is one of the reasons the Linux is still not mainstream.
You are missing the point.
I agree, it should be better. You've pointed out some situations which really should be improved.
However, your proposed solution makes things worse, not better. I would prefer a forever non-mainstream Linux to the bastardization that would be downloading random binaries from a Google search.
If you're interested in providing constructive criticism, tell me how we can keep the strengths of the repository system while also gaining the one-click you're suggesting.
I am a very technical computer user, but I'm also not into torturing myself or doing more work than I need to just to use my damn computer.
Neither am I, it's why I switched away from Windows in the first place.
If -I- have a problem with using Linux, you can bet your ass the vast majority of non-technical users will have a problem using it as well.
Sounds good. Except it's also arrogant and wrong.
The vast majority of non-technical users don't need to be downloading random pieces of software from the Internet, and are better off not doing so. They like the App Store for the iPhone, so I don't see how add/remove programs on Ubuntu is worse.
You have a harder time because you try to do more with your computer than the average non-technical user needs.
For example: Would you ever use the Chrome OS? I wouldn't. But I'd certainly recommend it to my grandmother. Less locally-installed software means less she can screw up.
It *is* preferable to have a separate ad-blocker, that should be a no-brainer within the unix philosophy.
Granted. But there are two large problems with it, in practice:
Firstly, while it's not entirely in line with the Unix philosophy, my browser already parsed the page. Why should my proxy have to, also?
And probably more importantly: Maybe this has changed, or maybe I just used the wrong proxy, but I remember using Polipo for a lot of things. I also remember finding out that it would cause certain pages to fail (even when I was just using it as a cache), and that, perversely, it'd often slow down my browsing.
you just need to copy the example config from /usr/share/uzbl/examples/
Yeah... Why do I need to do that?
Is it really any different clicking on an apt-url from some unfamiliar website?
Yes.
An apt-url is a URL to a package that is already in the repository, which means at least there's some minimal assurance that someone associated with Ubuntu or Canonical has looked at it, and some strong assurance that it hasn't been modified since then.
What about legitimate authors who didn't jump through the hoops to get their software into all the various repositories and packaging systems? Should they not be trusted simply because they didn't sign up with Canonical?
Pretty much. Ubuntu is community-maintained, and it's really not that difficult to at least get into the "universe" repository.
Of course, unlike the iPhone, it's always possible for the user to go around this, and I agree, it could be made easier than it is -- but it is pretty easy. I gave an example somewhere here involving Chrome -- it's a deb which sets up a repository as it's installed.
Now, you could make the case that downloading a random deb from the Internet is just as bad as downloading a random exe. But the point is, the capability is there if it's needed -- and most often, it makes much more sense to use something from an already-trusted repository.
All those -dev packages? Those are libraries that need to be on your Linux box in order to make the program. If, for some reason, the program you are trying to install cannot use a newer version of a library and you need that library for other software, you're quite thoroughly fucked.
Well, no, it's just gotten harder. Generally, you can pass an argument to ./configure to tell it where to find headers, and there can easily be multiple versions of a library installed -- that's why the .so files have version numbers on them.
I also can't remember the last time that happened to me.
MSI and DMG files have dependancies built in.
This has two major disadvantages:
First, it's just extra space, bandwidth, RAM, and cache. That last one is still not cheap.
And second, it makes it difficult to globally patch a vulnerability in such a library. Example: OpenSSL is likely to be used in a number of packages. The only way to patch it in all packages on, say, OS X, is to not bundle it and pray Apple includes it in the OS -- then it'll be handled by Software Update.
Updates are a single extension file (with MSI files anyway).
What does "extension file" mean, in this case?
And where is the option for me to tell my system to update all MSI-based apps I have installed?
I'm fairly sure this does not exist for OS X apps.
What would you suggest for me now? Got any more commands that will magically fix a retarded install method?
Contact the maintainer?
Your complaint here is that it doesn't work with newer libraries. But this is, again, similar to expecting a Win98 app to work in Windows 7. It often will, but sometimes it doesn't. Even when Microsoft pours blood, sweat, and tears into backwards compatibility, it's often working around a bug in the original app.
It's also worth mentioning (again) that the blame for this falls squarely on the shoulders of the app developer. Just as you would scream bloody murder if a Windows app required you to download Visual Studio Express and compile from source -- but you wouldn't be blaming Microsoft.
Of course it's not The One Thing(TM) holding Linux back, you'd have to be an idiot to think that. But you've got to have your blinders on pretty tight not to see that it would definitely improve the situation for Linux on the desktop.
I agree that the situation could be improved.
However, the methods most people suggest amount to "Copy Windows" or "Copy OS X", which would be a serious regression in many wa
at one point
Really? Let me Google that for you.
Yes, they are still selling Ubuntu boxes.
It's easy enough to disable, and won't re-enable on its own. And they indicate how to do so right there on the homepage.
This is pretty much par for the course, for any end-user-oriented app. Firefox auto-updates itself, on Windows and OS X.
That's my whole point, there is a shitload of software out there that it does not work for.
That is: There's a shitload of software out there that chooses not to work with it.
Worst case scenario for Windows is a zipped directory with some shortcuts to place on the desktop.
And some registry hacks. And some things you may have to manually compile. And a third-party tweak to make it work in Vista or Win7.
You're right that Windows is more unified. You're wrong that this is somehow the fault of Linux, or the distro, or some lack of functionality.
Again:
1. See there is no apt-url link - curse (shit is common here)
Have you ever tried to install a Windows app and found there's no exe, or msi?
That's the difference -- it's the app's own fault.
And the right way to deal with it is to go bother the app in question, not post pointless rants on Slashdot about how "Linux" is broken. It's a bit like saying Windows is broken because of identity theft.
Oh, and probably the strangest one: Suggesting a proxy-based adblocker.
Ok, I get that it's practical, for now. However, they seem to be saying this would be preferable to an adblocker built in to the browser, which makes no sense. Being able to right-click on an ad and figure out how to block it is not going to be replaced by editing some obscure config file in privoxy.
I like the idea, and I'd love to play with it a bit, but there are a few stupid design decisions:
Why don't you just use a reasonable config by default?
There really is no excuse for this. I mean, yes, I can understand where not everyone would want that "reasonable default", but that's why it's a default.
We don't want to store anything "automagically" in the users home. Some people prefer different file/directory layouts
Uhm... ~/.uzbl? How difficult is that? And if you don't like it, rm -rf ~/.uzbl!
Or just create an example script that sets up the default config, and put it in your FAQ.
We considered the option of having a global '/etc/uzbl' which user specific ones could override but that would overcomplicate things.
I'm sorry, but even mplayer is officially friendlier than uzbl. How the fuck is it "complicated" to read one config file, then another?
Uzbl itself doesn't use much gtk stuff (only the statusbar) so we could do without gtk. But Webkit needs a widget toolkit to create widgets (think javascript popups, html forms etc). Officially, it also supports QT and wxwigdets.
So, why doesn't uzbl also support these options? I'm using KDE, so Qt makes sense.
Uzbl.run( )
command is any uzbl command as defined above
return value: a string, either empty or containing the output of the command. Very few commands return their output currently, including js, script, and print.
They obviously realize that JS runs in a single thread. So the obvious implementation here would be to use a callback, not a return value, so you don't block the entire page while you run that script.
I mean, I want to like it, but that's a number of facepalms right off the bat, so I think I'll stick with Chrome until I have time to fix them.
Why would anybody develop a repository when all you have to do is Google it, clicky click the pretty blue link, and lo and behold it's installed (ignoring the plethora of security checks of course
Because that's really fucking insecure, compared to a repository?
Especially with .msi installs, way beyond anything I could find in Linux.
What does an MSI do that Linux package managers don't?
Because I can think of things Linux package managers do that MSIs don't.
There are a shitload of applications out there that do not exist in either, and if you want them they suck monkey testicles to install.
That's the fault of the application, not of the distro.
For example, try installing Chrome on Ubuntu -- you click either the 32-bit link or the 64-bit link, and it Just Works. It also sets up a Google repository so that it will continue to auto-update, through the package manager.
Your complaint would be somewhat like complaining because some Windows app demanded that you download and install Visual Studio Express first, then downloaded and compiled the source for it and several libraries it needs. It's certainly possible to make installation easier, but it's up to the app to do so.
Why hasn't anybody come up with the Linux equivalent of the .dmg file?
Linux had disk images for long before OS X did, and it's a terrible way of distributing software.
Seriously, I think you've been brainwashed a bit by the Apple flashiness. Think about what's going on here:
Every non-technical Mac user I know never gets to step 3. They get to step 2, say "Oh, there's my app!" and double-click it. Which means they run Firefox out of a disk image for years at a time, and never upgrade, because the DMG is read-only and Firefox won't auto-update itself.
Compare to:
In other words, the reason there isn't the "Linux equivalent to the DMG file" is because what we have actually is better, and easier to use, when it works.
For heaven's sake, it's 2009, why the hell do I have to friggin compile every damn piece of software that isn't in a repository?
Mostly because the people who provided that software didn't know what they were doing. I just untar things not in a repository.
For heaven's sake, it's 2009, why am I downloading and executing random, untrusted binaries from the Internet as the standard way to install software on Windows?
As it is now you have to dump a tarball into directory, run a few scripts while crossing your fingers that all your libraries match up, then make the binary.
If by "run a few scripts" you mean "exactly the same scripts for every single package", fine. I mean, it's going to be ./configure && make && make install.
Where I think you're confused is the library dependencies. Generally, when I download the source package, they'll also give me a list of '-dev' packages I need to install on Ubuntu. And it's a hell of a lot easier to do than trying to compile something on Windows.
As wonderful as the repository idea is - frankly I love that everything is right at your fingertips - it is completely unnecessary with Windows, because Google works just fine as a repository. Click the link and you're installing the program, no mess no hassle.
No security.
And if you wanted to set up a repository, it would not be hard, it would be little more than a database of .msi files, which install automatically.
And dependencies. And reverse-dependencies. And automatic updates. And third-party repositories.
No, if you wanted to do it right, you'd probably start with Windows Update -- except Microsoft has that locked down against third parties.
Frankly, some kind of unified one-step scripted install structure, preferably all in a single container, that actually worked as intended would catapult linux on the desktop by leaps and bounds.
Bullshit.
See, there's really no way such a system would see any wider adoption than any existing package manager. In other words, we already have this, and it's called dpkg. And you can, in fact, click a dpkg in a web browser, and expect it to work -- or even click an apturl, and it'll pull it through the repository, rather than the browser.
The absolute worst any modern desktop user has to do is run some commands -- that is, copy and paste something from a website into a commandline.
Somehow, I don't think that this is The One Thing(TM) that is holding Linux back. I'm going to say it's instead the lack of applications, the less-than-perfect Wine, the fact that it's unfamiliar, and the fact that most people aren't getting it preloaded.
I can think of a few ways to make the repository system better. But they don't involve mirroring the absolutely retarded way that Windows and OS X handle software installation.
I can think of a lot better ways to spend $5k, if the goal was getting dates.
A couple of minor good uses doesn't balance out myriad terrible abuses.
I don't know... It sounds terrible, but if you think about it, would you rather have Hiroshima and space travel, or no Hiroshima and be forever earthbound?
There is such a thing as a bad feature. For example, consider two computers, completely identical, except that one randomly kicks you in the nuts, and the other does not.
Except that randomly kicking you in the nuts has no good applications. I can't think of a good application of <blink> offhand. I can, however, think of good applications of animated gifs -- pretty much anywhere I need a few frames of animation, for example, it's going to be less CPU-intensive than flipping through a set of PNGs with JavaScript.
Having a feature which can be disabled is much better than not having a feature, I would think.
For example, if my browser supports audio and video (at least through Flash), I can watch YouTube.
If it starts annoying me elsewhere, I can always use something like adblock, flashblock, etc, to disable it. But I'd much rather have the feature, if it has any good application at all, than not have the feature.
If you'd bothered to Google it, you'd know that Gif can do true color, too.
Depending on the luck of the draw, any particular device may be more or less compatible
with Linux depending what was included.
And do you really think Dell would risk their reputation on "the luck of the draw"?
I'll give you a small (but annoying) example: My laptop could come with Ubuntu or with Vista. To get it with Ubuntu, I have to order it from an entirely different section of the website, with entirely different customization options -- obviously, I can't order it with Office preloaded.
The Vista version clearly comes with an 802.11n card. There's even a little "n-series" sticker on it.
But it came with an 802.11g card instead.
Want to guess why? I'm guessing it's because the G card had better Linux support.
I think the key element here is interest.
If she had the interest and the desire, she could get the experience and training. I suppose right now, being a mom is more important.
IQ is meant to measure intelligence, not knowledge. Knowing more about history and politics doesn't make him "smarter", just better educated -- the IQ test suggests that you could have just as thorough an understanding of these things, if you cared to.
I could be wrong, but it seems to me that how good we are at something that requires intelligence is largely a function of how much attention we put in it, and how interested we are in it, and how generally intelligent we are.
Would it be possible to try this out with people in varying stages of a relationship?
That is, do we eventually recover by adjusting to the person in question, or are married men who are truly still in love necessarily dumber around their wife?
A netbook is no more "designed" to run Ubuntu than the Kindle is.
Depends which netbook you get.
You have to do this specifically because AAC is patented.
That is: Third-party AAC libraries do exist. But they infringe on AAC patents. If software patents didn't exist, or if patents had a shorter term, reverse engineering and reimplementing these formats would be legal, no matter what the licensing restrictions.
Now go back and read my comment, and the thread in general -- this is about what would happen if there were no software patents. On the one hand, this would mean vorbis and theora might never see the light of day, and certainly wouldn't have seen what success they have. On the other hand, we'd be able to use AAC and h.264 without those licensing restrictions.
Yes, GIF really was "that bad", and PNG was a major improvement on it.
Which does raise the question, would people really have never added these to GIF, or developed PNG, if it weren't for the GIF patent?
And GIF's only "advantage" is mostly used for distracting annoyances,
Yet is still used for, to take an obvious example, AJAX spinners.
it would be quite reasonable to consider it a disadvantage.
Having a feature vs not having a feature?
The current period (I think it's seventeen years, but I Could Be Wrong.) is the product of centuries of experience with patents.
And during those centuries, we've had more than one example -- going back as far as Watt and Hornblower, of course -- of that term, or of the existence of patents at all, being a significant detriment.
There is actually a good case that the existence of patents at all hinders innovation more than it helps.
One thing I do know, however: simplistic solutions to complex problems don't work.
"For every problem, there is a solution which is simple, neat, and wrong."
Of course, that applies to language, too. Simplistic solutions can solve complex problems, and simple problems can require complex solutions.