What are your build options in 2.4.* vs those in 2.6.*?
The same, as far as possible.
Have all the drivers for your hardware been updated to 2.6.* yet (some things like promise disk controllers only have old drivers) - what does the README in the directory for the drivers for the hardware that doesn't work properly have to say about the subject?
The only external modules I'm using are nvidia and svgalib, both of which are updated. I assumed (perhaps naively) that in-kernel drivers would all be up to date with 2.6.
All this may seem like a hassle, if so there are still 2.4.* kernels coming out.
It is the software that defines the typical user, not the contrary. People used DOS for ages and everybody was happy
Many were very much not happy. My parents refused to use computers before the introduction of windows.
until Windows came around. I was very happy with fvwm(1).
I think the majority of today's "linux users" wouldn't be, even at the time.
You came around at first with RMS' argument saying that we must recognize other people's work, so please put GNU/ before everything. But then all your argumentation was that GNU is the most important part of the system. Can I conclude that that's why we should put GNU/ before everything? But that is "my rule", and whilst you call it silly you are using it to justify your whole point.
I'm not saying GNU is the most important part. The whole system is, pretty much, equally important. A lot more of it is GNU than anything else. That is why it should be called GNU.
Calling Linux "Linux" is not giving all credits to Linus.
By itself, without context, it is - and that is how most people are likely to hear Linux. People hear it and assume the system was written entirely by Linus - a "newbie" distribution may not even have a kernel source directory. Assuming the entire system was written by the GNU would be wrong, but not as wrong as assuming it was Linus, and would at least include the knowledge that a large number of hackers from around the world contributed.
Uhhh, then I take it you don't use too many usb devices because running mknod everytime you want to plug in your usb camera is not what I would consider user friendly.
Huh? The only time I've had to mknod is when I tried devfs. If you have to mknod the same node more than once it's devfs or udev that's causing your problem, not the solution./dev/sda,/dev/sda1, and equivalents all the way up to sdg are all present and correct on my system, not that I plug in more than one at a time.
And having/dev/hd* enumerate all the way out to k30 is also annoying when you have only one hard disk with three partitions on it.
Marginally, but less annoying than having the permissions on them change around when you're not looking (udev), or the whole thing disappear when you reboot (devfs). Besides, if you don't use them you can delete them, then if you find you need them you can just mknod them again.
Mobo is an FIC AZ11EA, 800MHz AMD Duron "Spitfire" proc, 384mb DDR SDRAM (don't tell me it doesn't exist) in 3 sticks, Netgear FA312 (I think), Nvidia RIVA TNT2 M64 (AGP) and then a PCI geforce2 MX/MX400.
Why don't you post it so we can see it? Oh yeah, no ad revenue that way.
Linux configuration files: they are all in different formats, and follow different rules. (case sensitive? headings? comments? whitespace significance?)
A huge exaggeration. There are a few different formats, yes, but this allows config files to be better suited to the program.
Very often configuration file changes do not take affect until you restart whatever program you are configuration.(sic)
This is also true in Windows.Take a look at Apache. A server widely acclaimed for its up-time, and yet you can't even change a single setting without restarting the server!
That's simply untrue. Send it the right signal and it will reread the config file and act on it with no trouble.
Compare this now with Microsoft IIS. A nice GUI that lets you change almost any setting or add an entire web site with a few clicks and you're already live. No restart, no downtime.
Apache is a server program designed for servers, which are often headless, so it simply makes more sense to have it configured with a text file. And when you are experienced - as a server admin is expected to be - it's simply quicker to use a text file. It's not a representative "desktop linux" application.
And you don't need a fat "Apache Unleashed" manual for IIS to figure out how to do this or that.
Funny, I've seen just as many of those books for IIS, and I never needed one for my Apache setup.
Suppose I install Gnome as default and want to install KDE and use that as the default. How easy is that? Well, every time I try, I have to search through several configuration files with 100s of lines to find the one that specifies the default GUI, and then it often doesn't even work.
That's a total lie. Even slackware, the most old-fashioned distribution, gives you a straightforward gui, two menu choices and you can choose which WM to use. Of course, if you prefer to you can use the command line. And even if what you said were true it would still be far easier to change in linux than in windows.
True, if you buy SuSE Linux on DVD at the store, you can drop it in and have it all set up for you. Windows XP brags about the same convenience and ease. How many patches must you apply to SuSE right after install for all the security fixes? Dozens. Windows? Dozens. Hmmm... Seems about even there. To be safe, you ought to be behind a firewall in either case. I've never found a recent Linux distro that didn't have patches ready to install as soon as I finished installing to OS.
While flaws obviously do need patching, I've left "stock" linux boxes for month with a direct net connection with no ill effects. There simply aren't the mass-reproducing worms you get on windows, the only way you will get cracked is if some cracker picks you out as a specific target. You can make whatever excuses you like in terms of marketshare, but it's still the case.
Let's compare apples to apples here. Windows is an OS, not a set of applications. You Linux fanatics get on Microsoft for bundling apps with their OS, and then here you get on their case for their lack of apps. So I won't even address the software bundled with each OS, since that is not comparing the OS but the packaging, which can easily be changed for either OS. Let's use our time wisely by looking at apps available to be installed later.
I haven't criticised MS for bundling apps with their OS. The fact is that MS couldn't change to including them, either because they charge a lot of money for them or for political reasons. So you get far more apps with your typical linux install. If MS did offer a version with its apps bundled you might have a point, but they don't.
Virtually every app available for Linux has a Windows counterpart. The reverse is also true. But the fact of the matter is, you open-source Linux fanatics choose to ignore this. OpenOffice is as much available for Windows as it is for Linux. S
So many people write "linux is not ready for the desktop" articles, seemingly without even using it, or else using it like the advanced user they are. In particular with the first point, I've seen many users speaking of the necessity of using the command line when there is a perfectly good GUI way to do it - it's simply far more efficient to use the command line. Linux is at the stage where it is easier than windows more often than not.
This isn't a mistake in bittorent design, bittorrent is not a filesharing program and was never meant to be a filesharing program. As you say, it's good as an FTP replacement, but the mass exodus to bittorrent for music/movies/warez trading was sheer idiocy. Claims of faster downloads (despite being blatantly false for anyone who actually tests it) are the only reason I can find for it, but even then it makes little sense. Everyone knows FTP is the fastest protocol for transferring files, but the days of people trying to use it for their filesharing are long gone.
Therefore, the idea of a DRM "period" is bogus. At the very least, the user should be able to add another period if there are enough copies left in the stacks.
If you read it, they can.
It should not require an Internet line either,
How else would you suggest they check whether there are copies available? I suppose you could go into the library, but then how would they get the license onto your computer?
and it should be able to run on free software not some attackware that executes on my computer in a manner contrary to my wishes.
The software has to be able to prevent you acting according to your wishes to do what it's meant to.
DRM erasing files on your machine after a given period is WRONG. Lateness should engender late fees, so the person can balance opportunity cost at least.
The fees are called fines for a reason. It's not intended that you decide whether to keep the book longer, just like you shouldn't choose to keep the copy and pay the library, they charge you for the times when you do as a punishment.
You can't "lose" a file like you can lose or destroy a book, and books at least can be distributed for massively less money on cd or online. Such cost savings should be figured in when purchasing and when deciding on checkout policies.
True, and I'm sure they are considered.
Even if the library purchases titles with a maximum simultaneous readers clause in it, if enough copies are available it should extend the period so that late fees are waived.
TFA mentions that these audiobooks can be renewed, just like you would with a physical book.
The library should be able to calculate AVERAGE SIMULTANEOUS READERSHIP of a given title to maximize its investment and give readers some of the benefits of digital technology. In other words, it should allow a burst of MORE simultaneous users than contracted, and then balance that out by artificially reducing the number of titles that can be simultaneously read at a later date. This can be amortized over a Very Long Time (tm), which gives the library some time to consider buying more simultaneous liscenses when it really needs them.
What motivation would a publisher have to license in this fashion though? Most physical books the library buys spend a lot of their time sitting on the shelves, why should they be able to buy less books and copy some of them some of the time?
Libraries should demand contracts with publishers which allow them to calculate average simultaneous readership to allow for readership bursts (say due to holidays or related news events). Libraries must also demand the option to easily purchase more liscenses at a later time based on an industry-wide open standard compliant form.
And if the publishers say no? Libraries are very much beholden to publishers, the income they give the publishers is relatively low. They're in no position to be making demands.
High cost efficiency publishing formats should also be provided, for example providing all the works of a given author, or all the works of a given genre, on one CD. That CD is really cheap to print, it is just about liscense fees. Forget software DRM and accounting systems that watch what you read. Just run some experiments, calculate the average simultaneous readership for the titles, and use this to figure out how many titles the library should buy.
Again, would the publishers be willing to do this?
It seems that with all this hub-bub libraries are probably paying WAY too much for digital media. If a library buys one book and 1000 people read it, that is still I think only one book's worth of money going to the author, is it? Libraries need some help here to protect their purchasing power. Amortizing access over time might be a good idea, then you could allow a large number of copies to be created for the big initial rush but not rent out any copies in the future.
You can legally learn English without having to pay for it. Many libraries will help you with this, offering free loan of books to learn from and, if they have sufficient volunteers or the council has decided it's important, free tuition. I have yet to see a library which offers free copies of windows, even just for loan.
It's a way of rewarding really funny posts, because you get karma for "insightful" moderations but not for "funny". This means borderline jokes posted early are a huge karma loser, because they get modded up and down repeatedly, and each downmoderation costs karma with no reward. Modding funny posts insightful is an attempt to redress the balance. (Of course the "correct" way to do it is mod funny posts up underrated)
It's on the device, which I presume acts as a USB mass storage device. You could run the virus directly off it, just like a virus on a floppy disc or CD or network share (windows treats anything with a drive letter the same), but yes it does require user intervention. It happened because this particular virus copies itself to (more or less) random directories, so someone had (presumably) the master disc image mounted and the virus made a copy of itself there.
As I read it, there's simply an executable on the player, so you'd have to run a random program you found on your player to get infected. Still, I can imagine someone thinking "Hmm, creative have put a program on my player, it must be a useful tool for using it".
Don't forget the first word macro virus. Not only did MS ship it on their documentation CDs, it had been written by someone within the company. Creative can at least make the argument that it was an external agent causing the problem, not them - like a cat burglar sneaking into the GM factory and cutting the brake cables.
The data will be valuable so the probe might as well keep on collecting it. Sure, it will be relatively less important as time goes on - we'll know enough about most of the stuff out there to ignore it, but more data is always useful in science.
And I suspect it's simply a fuel saving to have it end up heading inwards, so point it at the earth, it might be useful.
Use a better measure of code "importance" is what I mean, or should have meant. But it's much easier to count LOC contributions than try and asses how much programs from each source contribute to the OS. And though I may be wrong, provided programmers weren't aware I would do this before they wrote the code I suspect the results won't differ very much.
Of course it is, but in this case what makes the difference is the Netbsd kernel, not Linux? You must see in perspective, back when Linux was created. There was no "gnu/netbsd" then. There is a reason to call Linux as "Linux", and it is that it made the difference when it appeared. Before Linux, GNU was useless to most people.
That's only coincidence though. If netbsd had been released earlier, most of us would be using gnu/netbsd, and probably in the same kind of situation. Besides, the early releases of Linux were pretty horrible. It wasn't linux itself that made the difference. It was the hundreds of contributors willing to work on a free kernel as soon as one became openly available who transformed linux into something usable. And they were probably only there because they had all the other tools they needed for a free system - thanks solely to the GNU.
Then, if you agree with my rule, then you must call "KDE" a KDE system and "GNU/Gnome" a gnome system. OK. (Of course, people will never say "GNU/Gnome", no matter how much you insist). Nevertheless, KDE and Gnome where not qualitative advances. Linux was very useable before them.
Gnome already had GNU in it (that's what the first letter stands for), that's a stupid cheap shot. I'd say kde was a leap of the same magnitude as linux, for a "typical user" fvwm95 is simply unusable compared to 1.0. I'm not saying we should call our system KDE, I'm saying your rule is silly.
I don't know what kind of system you're running, but on my laptop, every 2.6 kernel has worked, and everything since 2.6.10 has had perfect ACPI support!
Glad someone's benefiting then. ACPI has been working fine for me since 2.4.20, 2.6 has brought me nothing but instability.
And I run a gentoo- and swsusp2-patched kernel, so it's even stable when patched. (I don't know how much of the stability comes from the patches, but normally patched kernels are considered another source of instability.)
I've tried gentoo kernels, in fact the one I'm using at the moment is the last gentoo 2.6.11 release, but the other versions are all unstable.
I don't think there's anything wrong with how Linus develops the kernel. I do get tired with all the 2.6.12.x versions, but they're usually a diff of literally 5-10 lines.
The kernel is, for me, no longer stable with 2.6, and the obvious change to attribute this to is Linus' not branching off a 2.7 for development.
I really don't see what's so bad here.
My system worked perfectly with 2.4. It doesn't with 2.6. That's why I have a problem.
Uh... so none of the vanilla sources are stable for your system, and none of the vendor kernels work either?
I can't say certainly none, but I have tried a variety and this is the only one which works
Smells like you have a configuration or hardware problem.
Such a problem would surely be apparent under 2.4 and 2.6.11. You sound like those people who say windows instability is usually due to bad ram. When the same system with the same config is stable with one kernel version but not others, I blame the kernel.
If there ARE changes related to your hardware, submit a bug report
I have. The issue that was causing my most obvious problem in 2.6.0-3 even got fixed, but sadly it was only to reveal another one.
I'm assuming you have a hardware issue, since you don't specify what device/filesystem/etc. is actually having problems
That's because I don't know. My system just freezes. There don't seem to be any untoward log entries or anything.
No, you were using it on the JVM. The fact that there's a reasonably good emulator for all the platforms you list is beside the point. Wheras.net can run properly, natively, on windows, freebsd, and at least one other unix.
My distro is starting to depend on 2.6, with things like NPTL. 2.4 is no longer supported. I'm fortunate in that I found this one 2.6 release that works on my system, but there may be other systems with which I'm not so lucky.
And why use the vanilla-kernels at all? Vendor-kernels are the ones that are considered stable these days. And there IS a "stable"-branch of the kernel (the 2.6.x.y).
All of those I have tried. None are stable.
In that case you should't be using bleeding-edge kernels, stick to the vendor-kernels. I mean, we are being conservative here?
I would, but they don't work. My distro has moved on to 2.6, so I'll try any 2.6 kernel I can to get a stable one.
Yes, but when people haven't been trying to write as many LOC as possible, it gives a reasonably accurate measure in a very quick time. Sure, do a more detailed comparison, but the LOC gives a quick number that will be reasonably accurate.
Slashdot Mirror
I *don't know*. The system freezes.
What are your build options in 2.4.* vs those in 2.6.*?
The same, as far as possible.
Have all the drivers for your hardware been updated to 2.6.* yet (some things like promise disk controllers only have old drivers) - what does the README in the directory for the drivers for the hardware that doesn't work properly have to say about the subject?
The only external modules I'm using are nvidia and svgalib, both of which are updated. I assumed (perhaps naively) that in-kernel drivers would all be up to date with 2.6.
All this may seem like a hassle, if so there are still 2.4.* kernels coming out.
Yes, but my distro no longer supports them.
No, but they strongly recommend it. Words like "obsolete" are being bandied around, and I've known people not help in fora unless you're using devfs.
Many were very much not happy. My parents refused to use computers before the introduction of windows.
until Windows came around. I was very happy with fvwm(1).
I think the majority of today's "linux users" wouldn't be, even at the time.
You came around at first with RMS' argument saying that we must recognize other people's work, so please put GNU/ before everything. But then all your argumentation was that GNU is the most important part of the system. Can I conclude that that's why we should put GNU/ before everything? But that is "my rule", and whilst you call it silly you are using it to justify your whole point.
I'm not saying GNU is the most important part. The whole system is, pretty much, equally important. A lot more of it is GNU than anything else. That is why it should be called GNU.
Calling Linux "Linux" is not giving all credits to Linus.
By itself, without context, it is - and that is how most people are likely to hear Linux. People hear it and assume the system was written entirely by Linus - a "newbie" distribution may not even have a kernel source directory. Assuming the entire system was written by the GNU would be wrong, but not as wrong as assuming it was Linus, and would at least include the knowledge that a large number of hackers from around the world contributed.
Huh? The only time I've had to mknod is when I tried devfs. If you have to mknod the same node more than once it's devfs or udev that's causing your problem, not the solution. /dev/sda, /dev/sda1, and equivalents all the way up to sdg are all present and correct on my system, not that I plug in more than one at a time.
And having /dev/hd* enumerate all the way out to k30 is also annoying when you have only one hard disk with three partitions on it.
Marginally, but less annoying than having the permissions on them change around when you're not looking (udev), or the whole thing disappear when you reboot (devfs). Besides, if you don't use them you can delete them, then if you find you need them you can just mknod them again.
Mobo is an FIC AZ11EA, 800MHz AMD Duron "Spitfire" proc, 384mb DDR SDRAM (don't tell me it doesn't exist) in 3 sticks, Netgear FA312 (I think), Nvidia RIVA TNT2 M64 (AGP) and then a PCI geforce2 MX/MX400.
Because it's better, and better by enough to be worth the money.
Linux configuration files: they are all in different formats, and follow different rules. (case sensitive? headings? comments? whitespace significance?)
A huge exaggeration. There are a few different formats, yes, but this allows config files to be better suited to the program.
Very often configuration file changes do not take affect until you restart whatever program you are configuration.(sic)
This is also true in Windows.Take a look at Apache. A server widely acclaimed for its up-time, and yet you can't even change a single setting without restarting the server!
That's simply untrue. Send it the right signal and it will reread the config file and act on it with no trouble.
Compare this now with Microsoft IIS. A nice GUI that lets you change almost any setting or add an entire web site with a few clicks and you're already live. No restart, no downtime.
Apache is a server program designed for servers, which are often headless, so it simply makes more sense to have it configured with a text file. And when you are experienced - as a server admin is expected to be - it's simply quicker to use a text file. It's not a representative "desktop linux" application.
And you don't need a fat "Apache Unleashed" manual for IIS to figure out how to do this or that.
Funny, I've seen just as many of those books for IIS, and I never needed one for my Apache setup.
Suppose I install Gnome as default and want to install KDE and use that as the default. How easy is that? Well, every time I try, I have to search through several configuration files with 100s of lines to find the one that specifies the default GUI, and then it often doesn't even work.
That's a total lie. Even slackware, the most old-fashioned distribution, gives you a straightforward gui, two menu choices and you can choose which WM to use. Of course, if you prefer to you can use the command line. And even if what you said were true it would still be far easier to change in linux than in windows.
True, if you buy SuSE Linux on DVD at the store, you can drop it in and have it all set up for you. Windows XP brags about the same convenience and ease. How many patches must you apply to SuSE right after install for all the security fixes? Dozens. Windows? Dozens. Hmmm... Seems about even there. To be safe, you ought to be behind a firewall in either case. I've never found a recent Linux distro that didn't have patches ready to install as soon as I finished installing to OS.
While flaws obviously do need patching, I've left "stock" linux boxes for month with a direct net connection with no ill effects. There simply aren't the mass-reproducing worms you get on windows, the only way you will get cracked is if some cracker picks you out as a specific target. You can make whatever excuses you like in terms of marketshare, but it's still the case.
Let's compare apples to apples here. Windows is an OS, not a set of applications. You Linux fanatics get on Microsoft for bundling apps with their OS, and then here you get on their case for their lack of apps. So I won't even address the software bundled with each OS, since that is not comparing the OS but the packaging, which can easily be changed for either OS. Let's use our time wisely by looking at apps available to be installed later.
I haven't criticised MS for bundling apps with their OS. The fact is that MS couldn't change to including them, either because they charge a lot of money for them or for political reasons. So you get far more apps with your typical linux install. If MS did offer a version with its apps bundled you might have a point, but they don't.
Virtually every app available for Linux has a Windows counterpart. The reverse is also true. But the fact of the matter is, you open-source Linux fanatics choose to ignore this. OpenOffice is as much available for Windows as it is for Linux. S
So many people write "linux is not ready for the desktop" articles, seemingly without even using it, or else using it like the advanced user they are. In particular with the first point, I've seen many users speaking of the necessity of using the command line when there is a perfectly good GUI way to do it - it's simply far more efficient to use the command line. Linux is at the stage where it is easier than windows more often than not.
This isn't a mistake in bittorent design, bittorrent is not a filesharing program and was never meant to be a filesharing program. As you say, it's good as an FTP replacement, but the mass exodus to bittorrent for music/movies/warez trading was sheer idiocy. Claims of faster downloads (despite being blatantly false for anyone who actually tests it) are the only reason I can find for it, but even then it makes little sense. Everyone knows FTP is the fastest protocol for transferring files, but the days of people trying to use it for their filesharing are long gone.
If you read it, they can.
It should not require an Internet line either,
How else would you suggest they check whether there are copies available? I suppose you could go into the library, but then how would they get the license onto your computer?
and it should be able to run on free software not some attackware that executes on my computer in a manner contrary to my wishes.
The software has to be able to prevent you acting according to your wishes to do what it's meant to.
DRM erasing files on your machine after a given period is WRONG. Lateness should engender late fees, so the person can balance opportunity cost at least.
The fees are called fines for a reason. It's not intended that you decide whether to keep the book longer, just like you shouldn't choose to keep the copy and pay the library, they charge you for the times when you do as a punishment.
You can't "lose" a file like you can lose or destroy a book, and books at least can be distributed for massively less money on cd or online. Such cost savings should be figured in when purchasing and when deciding on checkout policies.
True, and I'm sure they are considered.
Even if the library purchases titles with a maximum simultaneous readers clause in it, if enough copies are available it should extend the period so that late fees are waived.
TFA mentions that these audiobooks can be renewed, just like you would with a physical book.
The library should be able to calculate AVERAGE SIMULTANEOUS READERSHIP of a given title to maximize its investment and give readers some of the benefits of digital technology. In other words, it should allow a burst of MORE simultaneous users than contracted, and then balance that out by artificially reducing the number of titles that can be simultaneously read at a later date. This can be amortized over a Very Long Time (tm), which gives the library some time to consider buying more simultaneous liscenses when it really needs them.
What motivation would a publisher have to license in this fashion though? Most physical books the library buys spend a lot of their time sitting on the shelves, why should they be able to buy less books and copy some of them some of the time?
Libraries should demand contracts with publishers which allow them to calculate average simultaneous readership to allow for readership bursts (say due to holidays or related news events). Libraries must also demand the option to easily purchase more liscenses at a later time based on an industry-wide open standard compliant form.
And if the publishers say no? Libraries are very much beholden to publishers, the income they give the publishers is relatively low. They're in no position to be making demands.
High cost efficiency publishing formats should also be provided, for example providing all the works of a given author, or all the works of a given genre, on one CD. That CD is really cheap to print, it is just about liscense fees. Forget software DRM and accounting systems that watch what you read. Just run some experiments, calculate the average simultaneous readership for the titles, and use this to figure out how many titles the library should buy.
Again, would the publishers be willing to do this?
It seems that with all this hub-bub libraries are probably paying WAY too much for digital media. If a library buys one book and 1000 people read it, that is still I think only one book's worth of money going to the author, is it? Libraries need some help here to protect their purchasing power. Amortizing access over time might be a good idea, then you could allow a large number of copies to be created for the big initial rush but not rent out any copies in the future.
The library has to buy as many copies as it will
You can legally learn English without having to pay for it. Many libraries will help you with this, offering free loan of books to learn from and, if they have sufficient volunteers or the council has decided it's important, free tuition. I have yet to see a library which offers free copies of windows, even just for loan.
It's a way of rewarding really funny posts, because you get karma for "insightful" moderations but not for "funny". This means borderline jokes posted early are a huge karma loser, because they get modded up and down repeatedly, and each downmoderation costs karma with no reward. Modding funny posts insightful is an attempt to redress the balance. (Of course the "correct" way to do it is mod funny posts up underrated)
It's on the device, which I presume acts as a USB mass storage device. You could run the virus directly off it, just like a virus on a floppy disc or CD or network share (windows treats anything with a drive letter the same), but yes it does require user intervention. It happened because this particular virus copies itself to (more or less) random directories, so someone had (presumably) the master disc image mounted and the virus made a copy of itself there.
As I read it, there's simply an executable on the player, so you'd have to run a random program you found on your player to get infected. Still, I can imagine someone thinking "Hmm, creative have put a program on my player, it must be a useful tool for using it".
Don't forget the first word macro virus. Not only did MS ship it on their documentation CDs, it had been written by someone within the company. Creative can at least make the argument that it was an external agent causing the problem, not them - like a cat burglar sneaking into the GM factory and cutting the brake cables.
And I suspect it's simply a fuel saving to have it end up heading inwards, so point it at the earth, it might be useful.
Use a better measure of code "importance" is what I mean, or should have meant. But it's much easier to count LOC contributions than try and asses how much programs from each source contribute to the OS. And though I may be wrong, provided programmers weren't aware I would do this before they wrote the code I suspect the results won't differ very much.
That's only coincidence though. If netbsd had been released earlier, most of us would be using gnu/netbsd, and probably in the same kind of situation. Besides, the early releases of Linux were pretty horrible. It wasn't linux itself that made the difference. It was the hundreds of contributors willing to work on a free kernel as soon as one became openly available who transformed linux into something usable. And they were probably only there because they had all the other tools they needed for a free system - thanks solely to the GNU.
Then, if you agree with my rule, then you must call "KDE" a KDE system and "GNU/Gnome" a gnome system. OK. (Of course, people will never say "GNU/Gnome", no matter how much you insist). Nevertheless, KDE and Gnome where not qualitative advances. Linux was very useable before them.
Gnome already had GNU in it (that's what the first letter stands for), that's a stupid cheap shot. I'd say kde was a leap of the same magnitude as linux, for a "typical user" fvwm95 is simply unusable compared to 1.0. I'm not saying we should call our system KDE, I'm saying your rule is silly.
If Linux refers only to Linus
Maybe you should have read my post.
Glad someone's benefiting then. ACPI has been working fine for me since 2.4.20, 2.6 has brought me nothing but instability.
And I run a gentoo- and swsusp2-patched kernel, so it's even stable when patched. (I don't know how much of the stability comes from the patches, but normally patched kernels are considered another source of instability.)
I've tried gentoo kernels, in fact the one I'm using at the moment is the last gentoo 2.6.11 release, but the other versions are all unstable.
I don't think there's anything wrong with how Linus develops the kernel. I do get tired with all the 2.6.12.x versions, but they're usually a diff of literally 5-10 lines.
The kernel is, for me, no longer stable with 2.6, and the obvious change to attribute this to is Linus' not branching off a 2.7 for development.
I really don't see what's so bad here.
My system worked perfectly with 2.4. It doesn't with 2.6. That's why I have a problem.
I can't say certainly none, but I have tried a variety and this is the only one which works
Smells like you have a configuration or hardware problem.
Such a problem would surely be apparent under 2.4 and 2.6.11. You sound like those people who say windows instability is usually due to bad ram. When the same system with the same config is stable with one kernel version but not others, I blame the kernel.
If there ARE changes related to your hardware, submit a bug report
I have. The issue that was causing my most obvious problem in 2.6.0-3 even got fixed, but sadly it was only to reveal another one.
I'm assuming you have a hardware issue, since you don't specify what device/filesystem/etc. is actually having problems
That's because I don't know. My system just freezes. There don't seem to be any untoward log entries or anything.
nor if you are using any vendor-supplied drivers
I've tried removing these. Makes no difference.
No, you were using it on the JVM. The fact that there's a reasonably good emulator for all the platforms you list is beside the point. Wheras .net can run properly, natively, on windows, freebsd, and at least one other unix.
I'm not, I'm opposed to having to switch around my OS every other week. And I saw no benefit of devfs or udev over a simple static /dev.
My distro is starting to depend on 2.6, with things like NPTL. 2.4 is no longer supported. I'm fortunate in that I found this one 2.6 release that works on my system, but there may be other systems with which I'm not so lucky.
And why use the vanilla-kernels at all? Vendor-kernels are the ones that are considered stable these days. And there IS a "stable"-branch of the kernel (the 2.6.x.y).
All of those I have tried. None are stable.
In that case you should't be using bleeding-edge kernels, stick to the vendor-kernels. I mean, we are being conservative here?
I would, but they don't work. My distro has moved on to 2.6, so I'll try any 2.6 kernel I can to get a stable one.
Yes, but when people haven't been trying to write as many LOC as possible, it gives a reasonably accurate measure in a very quick time. Sure, do a more detailed comparison, but the LOC gives a quick number that will be reasonably accurate.
Ah, but how would it be to install 1.5? AIUI it's taken a lot of work to get to the current stage with the older versions.