I disagree. It's not necessary to push out broken, untested software as the default in a stable release just to get the code tested and debugged--especially with open-source software. It will take longer, but it will work, and the extra time it takes is worth it because of all the man-hours that are saved from end users going in circles trying to fix regression bugs that shouldn't have been presented to them in the first place. Pushing beta testing off onto end users of stable releases is selfish.
Um...I don't think shipping Amarok is wrong. I think shipping Dolphin and KPackageKit and plasma-desktop that all segfault regularly is wrong, after 5 major releases no less.
I don't care whether a piece of software is free, or developed for free in spare time--pushing out broken software as "stable" default software in a "stable" distribution is a broken model. That's what beta-testing and optional packages are for. I don't care if it takes a while longer to discover and fix the bugs--what's more important is that users of this "stable" distribution get truly stable software they can rely on. Not one user should ever have to spend time fixing regressions in major, fundamental functionality because some impatient developers wanted to use entire user bases as their guinea pigs and some foolish distributors went along with it. An upgrade should never, ever turn out to be a downgrade. There is no excuse for it.
It's stupidity like this that will push me back to Debian, even if it does take a little more work to maintain or set up.
CrashPlan is excellent. $50/year for one computer and unlimited space, indefinitely-kept versioning and deleted files, and a daemon that runs in the background all the time, with a separate GUI frontend.
I wish there were a referral plan so I could get something from this plug, but as of now, there's not.:/ haha Anyway, check it out. For a long time I used Duplicity to a web hosting account, but CrashPlan is easier and more reliable.
Actually, they decided to waste the bandwidth, because there is zero guarantee of return-on-investment for advertising. There's no guarantee that I see ads, or think about them, or buy because of them. Period.
No money has changed hands, so I am not obligated to watch their ads--nor are they obligated to provide any services. It's like at-will employment: it's at-will, and can be stopped at any time by any party. If they don't like it, let them require registration, or go the Hulu route and make it hard or impossible to block ads--they still won't know if I see, hear, or think about the ads. If it ends up being a waste of their bandwidth, that's completely their decision.
If you choose not to watch ads, then YouTube is paying money (via bandwidth, maintenance, etc.) to deliver you content, and you are giving them nothing back in return. In effect, you are freeloading. This is fine - free market and whatnot - but at least understand that your behaviors are only possible because they're subsidized by other more-profitable viewers. You're the Internet's equivalent of a welfare family. I agree with you in some ways - there should be no right to make money. However, it goes both ways. They have no right to make money, and you have no right to use their service. Unless you're willing to participate in some form of quid pro quo arrangement with a service provider, don't be surprised if you find yourself excluded from their service in favor of profitable users. With free services, meeting them halfway is the name of the game.
There are too many technical problems with your argument. Just because my browser or their Flash applet downloads an ad doesn't even mean that I see it. Just because the ad shows up on my screen doesn't mean that I have paid attention to it. Just because I pay attention to an ad doesn't mean I will visit the advertiser's web site, nor does it mean that I will purchase from them.
To follow your logic, if advertisers support the service, then I should go buy from them, otherwise I'm freeloading by not giving them anything in return.
All advertising is subject to this fundamental problem, whether print ads, billboards, radio, TV, web, streaming web video, or even a guy standing on the street corner in a costume waving a sign: there is no guarantee to the advertiser that anyone will notice or pay attention to his ads; there is no guarantee of return on his investment.
And to that problem I say: tough cheese. I'm not obligated to watch or listen to or think about anyone's ads. I can look away from billboards, skip print ads, turn down radio ads, change channels (or walk away from) TV ads, and as it so happens, can use technical measures to avoid Internet ads. And the advertisers are free to try to use technical countermeasures--oh well, arms races can be fun. No money has changed hands between us, so neither of us is obligated to do anything.
Of course they would love to find ways to force us to see and think about ads (i.e. to brainwash us), but I'm under no obligation to let them. If they don't want people "freeloading" by skipping ads, then they should go Hulu's route, or require registration, or sell access--and if those measures drive away visitors such as myself, so be it. I can live without YouTube--the world got along fine without it for a few thousand years, and the Internet's full of alternatives.
No--didn't you notice where I mentioned "the sum of their uncoordinated expectations"?
There is no (or very little) hierarchy, but there ought to be more. Quality is not valued highly enough in KDE right now. My impression is that it's more of a playground for many of the devs--code that should still be considered beta, or even alpha, is pushed out in "stable" releases. Bugs and regressions and feature-disappearances are not taken seriously enough. As much as I prefer KDE over GNOME, it's no wonder distributors are putting GNOME up as the primary UI, because--sadly--it's more reliable.
Sure, they do it for free, so they're entitled to do what they want. But consider Debian: its developers and maintainers work for free as well, but they put quality at the top of their list. It's done when it's done, and it's not released with known, serious bugs. A little bit of self-sacrifice--giving new, fun code a lower priority than maintaining and fixing existing code--is called for, IMHO.
By the way, developer resources are attributed (did you mean distributed) arbitrarily, aren't they? Each dev does what he wants, when he wants--isn't that arbitrary?
Well certainly Comcast could eat the costs. They are almost certainly capable of doing this. Just like Wal Mart could eat the cost of certain levels of shrinkage, and Microsoft could probably give VS2010 away for free and still make money overall. That doesn't mean that they should.
Actually, I think that's exactly what they should do. How many gabillions of dollars do they need to make? Especially considering their practical monopolies, they ought to eat the costs, rather than passing them on to their already-abused customers. But, of course, that would cut into profits, and growth must continue indefinitely.
Telecoms and ISPs ought to be non-profit corporations or coops. Then they'd actually use fees to serve customers well, rather than lining pockets of CEOs and entertainment corporations. But people ought not to kill each other, too. And so the world turns....:(
I think the opposite could be true: if the right is exclusive to the author, it means that it is exclusive to the author. As the GP said, the Constitution says what may be done, not what may not be done. Nowhere does it mention transference, or descendants, or corporations.
Remember to consider the Spirit, not merely the Letter.
DSL is mostly another type of monopoly, since it's run by the phone company. WiMAX and 3GPP LTE are a whole 'nother ballgame because of technical limitations, and may not be able to fully compete with wired services.
Obviously, a geographically-enabled monopoly does not qualify as a free market. If Comcast had competition, and its customers had more choice of ISP, it would be at least somewhat of a free market, and it would change things considerably. Since Comcast has a monopoly in many areas, and since it's providing competing content--a conflict of interest--it needs more regulation.
the equipment necessary to stream the kind of bandwidth Netflix needs to a significant portion of their customers simply can not be purchased and maintained for the current price of a residential broadband connection.
Do we know that for a fact? I am skeptical. Bandwidth usage globally is increasing, and the rate of increase is increasing, and it's only going to get worse. Every ISP in the world has to deal with this every day, every year, and so on. Comcast is a huge company. If carrying Netflix is putting them in the red, why doesn't it do the same to small, local cable ISPs, who only have a few thousand customers? Why aren't the local ISPs' upstream providers doing the same thing? What about ISPs in Europe and Japan, where they provide comparatively enormous amounts of bandwidth to users? Why aren't they going bankrupt when they're sending 10x the bandwidth Comcast provides to each customer?
I may be wrong, but I suspect it's not a matter of losing money carrying Netflix content, but simply a matter of corporate greed.
Nope. Even with KDE 4.5.1, plasma-desktop (the primary UI process!), Dolphin (the primary file manager!), and especially KPackageKit (the primary package manager!) segfault with some regularity.
It's not a system problem--it's unfixed bugs in KDE, bugs which I don't encounter when using 3.5.10.
No, he's not ignorant. His observations are spot-on from an end-user's point-of-view. This illustrates how the KDE devs are scratching their own itches. That's expected in open-source development, but KDE is a huge project, and it's released to end-users with the expectation that they will use it day-in, day-out. It's delusional to expect end-users to put up with segfaults and utter failures in software after five major iterations. But that seems to be the expectation of many KDE devs--or, at least, the sum of all their uncoordinated expectations.
What KDE needs is an overriding commitment to quality: it should be job #1. Bugs first, features (and ripping-out-and-replacing huge chunks) second.
Even KDE 4.5.1 is not rock solid. I still have segfaults in Dolphin, plasma-desktop, and KPackageKit. KDE 3.5.10 is more stable and reliable, and that's why I still use it on my laptop. It's sad, really.
You are absolutely right. Akonadi should be tested thoroughly for at least a year with no changes other than bugfixes before being put into a stable release.
Sad but true. Even in 4.5.1 I still get segfaults from Dolphin and plasma-desktop from time to time. Shouldn't those have been ironed out by, oh, I don't know, 4.2, at least?
Government wasn't out to punish every infraction. They were out to teach if they thought you were capable of learning.
Sorry, but that's the dumbest response I've seen in a while. Where you drew that logical fallacy from is beyond me. It's some of the worst logic I've ever seen.
KDE should be approaching around version 4.3 now, not 4.6. Why? It still has alpha-level software in full releases. Example: yesterday I filed a bug on Nepomuk because it fails to follow moved files (files moved in Dolphin, no less), and it loses the assigned tags and ratings. It's completely undependable, and therefore completely useless. I might as well put the tags and ratings in the filename. Nepomuk is missing basic functionality--it should be considered alpha-level software--yet it's presented to the user in a full "KDE SC release" as if it's feature-complete, reliable, and ready to be trusted and used to its full extent.
KDE needs to go the way of Debian and release "when it's ready." If I were a betting man I'd put money on people losing email to this new Akonadi backend because I know many bugs are going to be discovered and fixed only after 4.6 is released and put into the hands of unsuspecting users. Akonadi and the KDE PIM software should be tested to smithereens before it's even released as a first beta! There's no way I'd trust this new backend with my personal email archive! I lost email to old versions of KMail in 3.5, and that was a codebase that'd been worked on for years! There's absolutely no way Akonadi should be considered ready for primetime and released until it's been heavily tested and verified to function correctly and reliably, preferably with extensive unit tests. I don't care that KDE is developed by people for free in their spare time; they're expecting users to trust this new software with their important data, so they should make absolutely sure it's as tested and vetted as possible before releasing it to average users.
I still use KDE 3.5 on my laptop because even in KDE 4.5 I can't get the Plasma taskbar to look as clean, readable, and usable as Kicker in 3.5. I'd have to spend a week hacking a Plasma theme with custom PNGs just to get a clean taskbar with a plain translucent background that doesn't waste screen space on big, ugly borders and shiny, barely-readable buttons.
Not to mention that KPackageKit constantly crashes in Kubuntu 10.10, popping up a segfault error on top of whatever you're doing whenever it feels like it. (Don't even ask how many dupes of this bug are reported on bugs.kde.org.)
The bottom line, IMHO, is this: KDE shouldn't release anything (going back to 4.x) until it has feature-parity with previous versions, and until it is as tested and reliable as possible--in other words, "when it's ready." Anything else merely frustrates users and hurts its reputation--and sometimes risks data loss, as well.
But, of course, "voices of reason" like this (IMHO) have been sounding off since 4.0 and have always fallen on deaf ears. *sigh*
Slashdot Mirror
I disagree. It's not necessary to push out broken, untested software as the default in a stable release just to get the code tested and debugged--especially with open-source software. It will take longer, but it will work, and the extra time it takes is worth it because of all the man-hours that are saved from end users going in circles trying to fix regression bugs that shouldn't have been presented to them in the first place. Pushing beta testing off onto end users of stable releases is selfish.
Um...I don't think shipping Amarok is wrong. I think shipping Dolphin and KPackageKit and plasma-desktop that all segfault regularly is wrong, after 5 major releases no less.
Try this: "On Thursday, Google patched..."
It is a bagillion times better than "Google on Thursday patched..."
Yes, I know every single stupid press release there is uses the latter form, but it's incredibly stupid.
I don't care whether a piece of software is free, or developed for free in spare time--pushing out broken software as "stable" default software in a "stable" distribution is a broken model. That's what beta-testing and optional packages are for. I don't care if it takes a while longer to discover and fix the bugs--what's more important is that users of this "stable" distribution get truly stable software they can rely on. Not one user should ever have to spend time fixing regressions in major, fundamental functionality because some impatient developers wanted to use entire user bases as their guinea pigs and some foolish distributors went along with it. An upgrade should never, ever turn out to be a downgrade. There is no excuse for it.
It's stupidity like this that will push me back to Debian, even if it does take a little more work to maintain or set up.
CrashPlan is excellent. $50/year for one computer and unlimited space, indefinitely-kept versioning and deleted files, and a daemon that runs in the background all the time, with a separate GUI frontend.
I wish there were a referral plan so I could get something from this plug, but as of now, there's not. :/ haha Anyway, check it out. For a long time I used Duplicity to a web hosting account, but CrashPlan is easier and more reliable.
Maybe telecoms should be non-profit organizations.
Actually, they decided to waste the bandwidth, because there is zero guarantee of return-on-investment for advertising. There's no guarantee that I see ads, or think about them, or buy because of them. Period.
No money has changed hands, so I am not obligated to watch their ads--nor are they obligated to provide any services. It's like at-will employment: it's at-will, and can be stopped at any time by any party. If they don't like it, let them require registration, or go the Hulu route and make it hard or impossible to block ads--they still won't know if I see, hear, or think about the ads. If it ends up being a waste of their bandwidth, that's completely their decision.
If you choose not to watch ads, then YouTube is paying money (via bandwidth, maintenance, etc.) to deliver you content, and you are giving them nothing back in return. In effect, you are freeloading. This is fine - free market and whatnot - but at least understand that your behaviors are only possible because they're subsidized by other more-profitable viewers. You're the Internet's equivalent of a welfare family. I agree with you in some ways - there should be no right to make money. However, it goes both ways. They have no right to make money, and you have no right to use their service. Unless you're willing to participate in some form of quid pro quo arrangement with a service provider, don't be surprised if you find yourself excluded from their service in favor of profitable users. With free services, meeting them halfway is the name of the game.
There are too many technical problems with your argument. Just because my browser or their Flash applet downloads an ad doesn't even mean that I see it. Just because the ad shows up on my screen doesn't mean that I have paid attention to it. Just because I pay attention to an ad doesn't mean I will visit the advertiser's web site, nor does it mean that I will purchase from them.
To follow your logic, if advertisers support the service, then I should go buy from them, otherwise I'm freeloading by not giving them anything in return.
All advertising is subject to this fundamental problem, whether print ads, billboards, radio, TV, web, streaming web video, or even a guy standing on the street corner in a costume waving a sign: there is no guarantee to the advertiser that anyone will notice or pay attention to his ads; there is no guarantee of return on his investment.
And to that problem I say: tough cheese. I'm not obligated to watch or listen to or think about anyone's ads. I can look away from billboards, skip print ads, turn down radio ads, change channels (or walk away from) TV ads, and as it so happens, can use technical measures to avoid Internet ads. And the advertisers are free to try to use technical countermeasures--oh well, arms races can be fun. No money has changed hands between us, so neither of us is obligated to do anything.
Of course they would love to find ways to force us to see and think about ads (i.e. to brainwash us), but I'm under no obligation to let them. If they don't want people "freeloading" by skipping ads, then they should go Hulu's route, or require registration, or sell access--and if those measures drive away visitors such as myself, so be it. I can live without YouTube--the world got along fine without it for a few thousand years, and the Internet's full of alternatives.
And so we are reminded: things are not always what they seem.
Haha, flamebait? I guess we need a "+1 Irony."
No--didn't you notice where I mentioned "the sum of their uncoordinated expectations"?
There is no (or very little) hierarchy, but there ought to be more. Quality is not valued highly enough in KDE right now. My impression is that it's more of a playground for many of the devs--code that should still be considered beta, or even alpha, is pushed out in "stable" releases. Bugs and regressions and feature-disappearances are not taken seriously enough. As much as I prefer KDE over GNOME, it's no wonder distributors are putting GNOME up as the primary UI, because--sadly--it's more reliable.
Sure, they do it for free, so they're entitled to do what they want. But consider Debian: its developers and maintainers work for free as well, but they put quality at the top of their list. It's done when it's done, and it's not released with known, serious bugs. A little bit of self-sacrifice--giving new, fun code a lower priority than maintaining and fixing existing code--is called for, IMHO.
By the way, developer resources are attributed (did you mean distributed) arbitrarily, aren't they? Each dev does what he wants, when he wants--isn't that arbitrary?
Well certainly Comcast could eat the costs. They are almost certainly capable of doing this. Just like Wal Mart could eat the cost of certain levels of shrinkage, and Microsoft could probably give VS2010 away for free and still make money overall. That doesn't mean that they should.
Actually, I think that's exactly what they should do. How many gabillions of dollars do they need to make? Especially considering their practical monopolies, they ought to eat the costs, rather than passing them on to their already-abused customers. But, of course, that would cut into profits, and growth must continue indefinitely.
Telecoms and ISPs ought to be non-profit corporations or coops. Then they'd actually use fees to serve customers well, rather than lining pockets of CEOs and entertainment corporations. But people ought not to kill each other, too. And so the world turns.... :(
Why do you say that?
I think the opposite could be true: if the right is exclusive to the author, it means that it is exclusive to the author. As the GP said, the Constitution says what may be done, not what may not be done. Nowhere does it mention transference, or descendants, or corporations.
Remember to consider the Spirit, not merely the Letter.
If that's true, someone mod the parent up.
DSL is mostly another type of monopoly, since it's run by the phone company. WiMAX and 3GPP LTE are a whole 'nother ballgame because of technical limitations, and may not be able to fully compete with wired services.
Obviously, a geographically-enabled monopoly does not qualify as a free market. If Comcast had competition, and its customers had more choice of ISP, it would be at least somewhat of a free market, and it would change things considerably. Since Comcast has a monopoly in many areas, and since it's providing competing content--a conflict of interest--it needs more regulation.
Your argument is fallacious.
the equipment necessary to stream the kind of bandwidth Netflix needs to a significant portion of their customers simply can not be purchased and maintained for the current price of a residential broadband connection.
Do we know that for a fact? I am skeptical. Bandwidth usage globally is increasing, and the rate of increase is increasing, and it's only going to get worse. Every ISP in the world has to deal with this every day, every year, and so on. Comcast is a huge company. If carrying Netflix is putting them in the red, why doesn't it do the same to small, local cable ISPs, who only have a few thousand customers? Why aren't the local ISPs' upstream providers doing the same thing? What about ISPs in Europe and Japan, where they provide comparatively enormous amounts of bandwidth to users? Why aren't they going bankrupt when they're sending 10x the bandwidth Comcast provides to each customer?
I may be wrong, but I suspect it's not a matter of losing money carrying Netflix content, but simply a matter of corporate greed.
Nope. Even with KDE 4.5.1, plasma-desktop (the primary UI process!), Dolphin (the primary file manager!), and especially KPackageKit (the primary package manager!) segfault with some regularity.
It's not a system problem--it's unfixed bugs in KDE, bugs which I don't encounter when using 3.5.10.
No, he's not ignorant. His observations are spot-on from an end-user's point-of-view. This illustrates how the KDE devs are scratching their own itches. That's expected in open-source development, but KDE is a huge project, and it's released to end-users with the expectation that they will use it day-in, day-out. It's delusional to expect end-users to put up with segfaults and utter failures in software after five major iterations. But that seems to be the expectation of many KDE devs--or, at least, the sum of all their uncoordinated expectations.
What KDE needs is an overriding commitment to quality: it should be job #1. Bugs first, features (and ripping-out-and-replacing huge chunks) second.
Even KDE 4.5.1 is not rock solid. I still have segfaults in Dolphin, plasma-desktop, and KPackageKit. KDE 3.5.10 is more stable and reliable, and that's why I still use it on my laptop. It's sad, really.
You are absolutely right. Akonadi should be tested thoroughly for at least a year with no changes other than bugfixes before being put into a stable release.
Sad but true. Even in 4.5.1 I still get segfaults from Dolphin and plasma-desktop from time to time. Shouldn't those have been ironed out by, oh, I don't know, 4.2, at least?
Government wasn't out to punish every infraction. They were out to teach if they thought you were capable of learning.
Sorry, but that's the dumbest response I've seen in a while. Where you drew that logical fallacy from is beyond me. It's some of the worst logic I've ever seen.
KDE should be approaching around version 4.3 now, not 4.6. Why? It still has alpha-level software in full releases. Example: yesterday I filed a bug on Nepomuk because it fails to follow moved files (files moved in Dolphin, no less), and it loses the assigned tags and ratings. It's completely undependable, and therefore completely useless. I might as well put the tags and ratings in the filename. Nepomuk is missing basic functionality--it should be considered alpha-level software--yet it's presented to the user in a full "KDE SC release" as if it's feature-complete, reliable, and ready to be trusted and used to its full extent.
KDE needs to go the way of Debian and release "when it's ready." If I were a betting man I'd put money on people losing email to this new Akonadi backend because I know many bugs are going to be discovered and fixed only after 4.6 is released and put into the hands of unsuspecting users. Akonadi and the KDE PIM software should be tested to smithereens before it's even released as a first beta! There's no way I'd trust this new backend with my personal email archive! I lost email to old versions of KMail in 3.5, and that was a codebase that'd been worked on for years! There's absolutely no way Akonadi should be considered ready for primetime and released until it's been heavily tested and verified to function correctly and reliably, preferably with extensive unit tests. I don't care that KDE is developed by people for free in their spare time; they're expecting users to trust this new software with their important data, so they should make absolutely sure it's as tested and vetted as possible before releasing it to average users.
I still use KDE 3.5 on my laptop because even in KDE 4.5 I can't get the Plasma taskbar to look as clean, readable, and usable as Kicker in 3.5. I'd have to spend a week hacking a Plasma theme with custom PNGs just to get a clean taskbar with a plain translucent background that doesn't waste screen space on big, ugly borders and shiny, barely-readable buttons.
Not to mention that KPackageKit constantly crashes in Kubuntu 10.10, popping up a segfault error on top of whatever you're doing whenever it feels like it. (Don't even ask how many dupes of this bug are reported on bugs.kde.org.)
The bottom line, IMHO, is this: KDE shouldn't release anything (going back to 4.x) until it has feature-parity with previous versions, and until it is as tested and reliable as possible--in other words, "when it's ready." Anything else merely frustrates users and hurts its reputation--and sometimes risks data loss, as well.
But, of course, "voices of reason" like this (IMHO) have been sounding off since 4.0 and have always fallen on deaf ears. *sigh*
Chuck? :)