Well, even if Linus does move on before he dies, I imagine Alan Cox would probably be the most likely candidate to takeover development of the mainstream kernel when Linus decides he's burnt-out. While he probably wouldn't be as conservative as Linus is with it, I don't think things would change too drasticaly.
Uh, this is already done, both in the kernel and in many linux distributions (slackware, debian, even redhat has Rawhide)
those -preXX kernel versions are the "unstable" releases (so to speak) of the vanilla kernel tree. the odd-numbered kernels are development kernels that don't even come close to being stable enough to be called unstable.
the -ac kernels are not simply -stable vs -current, a severely unstable patch that might break things isn't going to intentionaly make it into -ac kernels... you could think of -ac kernels as a testbed for future intigration into the mainstream kernel, but that doesn't make it unstable on the level you're talking about with FreeBSD's -current, that level of instability is reserved for the odd-numbered kernel series (i.e. 2.3 and the forthcoming 2.5).
One does wonder though what clueless businesses that already know odd=unstable for the kernel will think when 3.0 arrives:)
There have been some issues with Athlon optimizations, however you don't have to use Athlon optimizations when compiling your kernel.
I've had NO trouble with Athlon optimizations in 2.4.7 and 2.4.8 and now 2.4.8-ac9.
Most of the Athlon problems reported in the past have been problems with VIA's chipsets, which I label as "devil-spawn".
Same here, by sticking to the released kernels I have NEVER had a stability problem outside of the binary nvidia drivers except for a FEW times that were related to VERY specific hardware.
If you experience crashes/oopses that seem to be related to the kernel, PLEASE PLEASE PLEASE report them with all possible information to the linux-kernel list.
"Competing" ?
Without these alternate kernel trees, nothing would ever get done. the -ac trees really aren't a tree I'd recommend for a production server unless it has a fix or driver that the server desperately needs... the -ac's are to test and impliment things in advance to KEEP the "stable" tree "stable", and keep Linus happy that the patches he's putting in the mainstream tree have been tested.
And no, I don't feel the 2.4 series deserves to be called stable, but I damn well use it anyway on my primary desktop box:)
it's designed for embedded systems, but it does exist and could likely be expanded for use in full-scale desktop systems
it makes use of some code from glibc but the diff would likely be gigantic, it's essentialy a fresh write of a C library
one nitpick, LILO isn't strictly neccisary, there's at least 3 other methods of booting linux (grub, loadlin, and a bootloader that version.4 of was recently was announced on linux-kernel), and there may well be more (user-mode linux anyone?:) (then again ext2fs and reiserfs aren't neccisary either if you care to reformat your linux partitions:)
non-gcc C compiliers are avalible
non-gnu C libraries are avalible
non-gnu shells are avalible
non-gnu archiving/compression utilities are avalible
non-gnu "binutils" aren't, strictly speaking, avalible for Linux for x86 to my knowledge, but they could easily be made
non-gnu text vewing programs are avalible
non-gnu "curses" libraries are avalible, though outdated/obsolete
sorry, but nothing you mentioned doesn't have non-gnu alternatives for use in a linux system
Oh? Why should it be? I could construct a Linux system, prehaps for an embedded device, that uses little or no GNU software, and it could get covered by LinuxToday, because it IS Linux.
C'mon people, drop the nit picking when it doesn't really matter.
If someone says "Linux is too big for an embedded system", tell them that they're thinking of GNU/Linux. If someone asks what the difference between GNU and Linux is, tell them. Don't go nit picking over what exactly to call something based around "Linux".
idsoftware has and hopefully will continue to impliment OpenGL exenstively in its engines, I've as yet failed to hear Carmack complain about needing to customize every part of OpenGL for an individual card. Certain cards have certainly required some special work, but this was because the manufacturer either had a severely broken OpenGL, or simply didn't provide OpenGL support sufficient for use in Quake engines (think 3dfx cards and the "minigl" drivers)
How about some documented proof that various companies impliment OpenGL so differently from eachother to require special coding on the part of the game programmers?
A few people seem to misunderstand what Loki does... they don't need to sell as many games to be profitable as a normal software company would, because they don't have the same design and development costs. Their job is to port EXISTING games to GNU/Linux, much of the code in these games remains unmodified, Loki likely never even looks at much of it. The primary job of their programmers are to take care of graphics/display, sound, and input (keyboard/mouse), paths/filename work, and cleaning some of the code up to work with gcc and glibc.
None of this takes anywhere near as much time as the initial design and development of the game, thus it also doesn't cost nearly as much.
A significant portion of the work for games is artwork, this also is eliminated from Loki's work, the graphics/artwork/models are already there and done.
Well, there's always braille machines, as I understand it, they're pretty OS-independant so long as it's text mode.
Zipslack, the UMSDOS-based version of Slackware designed to fit on a 100MB ZIP disk has a variant that uses a peice of software to "speak" the words through a sound card.
IBM has Via Voice for Linux for dictating, not GPL/compatible of course, but it is there.
There are magnifier utilities for X for those that are visualy impared but not totally blind.
Things like Microsoft's "stickykeys" and "mousekeys" I don't know if those exist, but both could probably be implimented in user-space, allowing easy porting to BSD and such, or they could be implimented as a kernel patch.
There are solutions avalible, some just aren't as obvious as others.
Chris Hiner already mentioned Via Voice from IBM, I'm just going to say that I'm running on the assumption that in 1997, a reporter wasn't going to be using linux to write their stories:)
Repetitive Stress Injury is a generic term for conditions like Carpal Tunnel Syndrome, it's used CONSTANTLY in the medical field, and on slashdot itself.
what are they supposed to do? buy Dragon Naturally Speaking
if they'd truely evaluated it in 1997, they would have come to the conclusion that she could continue to do her job with assistance from software and a microphone
they didn't evaluate anything, they just decided they wanted to get her off their hands
it's actually not as funny as you may think, programmers DO work in standard programming languages like C or Perl with dictation software, it simply requires some customization of how things are recognized. Most, if not all VR software can be customized sufficiently to do coding.
You don't get it, it's not emotional/mental stress, it's physical stress, the same thing that makes a stick break if you bend it enough. Typing for years on keyboards can cause Repetitive Stress Injury, this is where the repetitive nature of the task over long periods causes stress on the nerves, muscles, etc. in the hands wrists and arms.
RSI has also been seen in professions completely unrelated to computers, such as a person working in a factory doing the same thing for years, even a blacksmith making horseshoes for years could aquire RSI.
you could try something like learning that "Stress" doesn't reffer to EMOTIONAL stress in repetitive stress injury
it refers to stress put on the PHYSICAL components of the hands/wrists/arms, STRESS is what makes a stick break if you bend it enough.
10 years ago we didn't have viable voice recognition software. Four years ago when this reporter was fired, we DID. The employer spent a bunch of money on chair and workstation modifications instead of spending it on voice recognition software.
Consider, a blind person can dress him/herself, do laundry, make a bed, even shop in some cases. They cannot drive, of course, however they can get around by walking or bus. And in some cases, someone with carpal tunnel or similar may not be able to drive safely, as the pain it causes their wrists can be extreme.
There was one comment on slashdot recently about a man who owned a small computer company, who was completely blind, but walked around without a cane or any sort of guide, built, troubleshooted, took apart, added to, computers by touch, using brail or text-to-speech for interaction with the computers, etc.
However, if they had fired this reporter because she'd gone blind, she would likely have won without trouble.
Voice recognition is COMPLETELY viable, even in 1997 it was usable, esspecialy if the user had some use of their hands to allow for manual corrections when neccisary. All they would have had to do was spend a few hundred dollars on Dragon Naturally Speaking, and a few bucks on a microphone, and everything would have been fine.
Why not? There's many programmers out there that use voice recognition for a substantial part of their coding. Marketing people can easily use voice recognition. Prehaps the only major roles that can't be adequately performed is QA and system administration work, and even those CAN be performed in the right enviorment.
You're throwing away good talent because they can't type for extended periods... that's just sad.
Bad idea, you risk having stupid netadmins that impliment DHCP so poorly that it's difficult or impossible to get it working on a non-windows machine. Better to keep it as-is and have all the routers block DHCP messages.
Slashdot Mirror
Well, even if Linus does move on before he dies, I imagine Alan Cox would probably be the most likely candidate to takeover development of the mainstream kernel when Linus decides he's burnt-out. While he probably wouldn't be as conservative as Linus is with it, I don't think things would change too drasticaly.
ahh yes, the usability that doesn't allow you such things as right clicking on the taskbar to change settings...
if you're going to troll by calling a Windows system more usable than linux, at least use something PLAUSABLE, like win98, in your examples...
Uh, this is already done, both in the kernel and in many linux distributions (slackware, debian, even redhat has Rawhide)
:)
those -preXX kernel versions are the "unstable" releases (so to speak) of the vanilla kernel tree. the odd-numbered kernels are development kernels that don't even come close to being stable enough to be called unstable.
the -ac kernels are not simply -stable vs -current, a severely unstable patch that might break things isn't going to intentionaly make it into -ac kernels... you could think of -ac kernels as a testbed for future intigration into the mainstream kernel, but that doesn't make it unstable on the level you're talking about with FreeBSD's -current, that level of instability is reserved for the odd-numbered kernel series (i.e. 2.3 and the forthcoming 2.5).
One does wonder though what clueless businesses that already know odd=unstable for the kernel will think when 3.0 arrives
There have been some issues with Athlon optimizations, however you don't have to use Athlon optimizations when compiling your kernel.
I've had NO trouble with Athlon optimizations in 2.4.7 and 2.4.8 and now 2.4.8-ac9.
Most of the Athlon problems reported in the past have been problems with VIA's chipsets, which I label as "devil-spawn".
Same here, by sticking to the released kernels I have NEVER had a stability problem outside of the binary nvidia drivers except for a FEW times that were related to VERY specific hardware.
If you experience crashes/oopses that seem to be related to the kernel, PLEASE PLEASE PLEASE report them with all possible information to the linux-kernel list.
"Competing" ?
:)
Without these alternate kernel trees, nothing would ever get done. the -ac trees really aren't a tree I'd recommend for a production server unless it has a fix or driver that the server desperately needs... the -ac's are to test and impliment things in advance to KEEP the "stable" tree "stable", and keep Linus happy that the patches he's putting in the mainstream tree have been tested.
And no, I don't feel the 2.4 series deserves to be called stable, but I damn well use it anyway on my primary desktop box
Linus, GIVE US 2.5 TO PLAY WITH!
or, thanks to a freshmeat search, you can go to sudo's homepage and get tarballs
http://www.courtesan.com/sudo/
That was the point, we don't need to go off half-cocked whenever someone calls it Linux instead of GNU/Linux
glibc isn't really neccisary
.4 of was recently was announced on linux-kernel), and there may well be more (user-mode linux anyone? :) (then again ext2fs and reiserfs aren't neccisary either if you care to reformat your linux partitions :)
http://cvs.uclinux.org/uClibc.html
it's designed for embedded systems, but it does exist and could likely be expanded for use in full-scale desktop systems
it makes use of some code from glibc but the diff would likely be gigantic, it's essentialy a fresh write of a C library
one nitpick, LILO isn't strictly neccisary, there's at least 3 other methods of booting linux (grub, loadlin, and a bootloader that version
none of those are neccisary
non-gcc C compiliers are avalible
non-gnu C libraries are avalible
non-gnu shells are avalible
non-gnu archiving/compression utilities are avalible
non-gnu "binutils" aren't, strictly speaking, avalible for Linux for x86 to my knowledge, but they could easily be made
non-gnu text vewing programs are avalible
non-gnu "curses" libraries are avalible, though outdated/obsolete
sorry, but nothing you mentioned doesn't have non-gnu alternatives for use in a linux system
Oh? Why should it be? I could construct a Linux system, prehaps for an embedded device, that uses little or no GNU software, and it could get covered by LinuxToday, because it IS Linux.
C'mon people, drop the nit picking when it doesn't really matter.
If someone says "Linux is too big for an embedded system", tell them that they're thinking of GNU/Linux. If someone asks what the difference between GNU and Linux is, tell them. Don't go nit picking over what exactly to call something based around "Linux".
OK, so Alice's creators are idiots.
idsoftware has and hopefully will continue to impliment OpenGL exenstively in its engines, I've as yet failed to hear Carmack complain about needing to customize every part of OpenGL for an individual card. Certain cards have certainly required some special work, but this was because the manufacturer either had a severely broken OpenGL, or simply didn't provide OpenGL support sufficient for use in Quake engines (think 3dfx cards and the "minigl" drivers)
How about some documented proof that various companies impliment OpenGL so differently from eachother to require special coding on the part of the game programmers?
How about we play a little game of our own, called "provide some documented proof" that Quake3 depends that heavily on DirectX.
A few people seem to misunderstand what Loki does... they don't need to sell as many games to be profitable as a normal software company would, because they don't have the same design and development costs. Their job is to port EXISTING games to GNU/Linux, much of the code in these games remains unmodified, Loki likely never even looks at much of it. The primary job of their programmers are to take care of graphics/display, sound, and input (keyboard/mouse), paths/filename work, and cleaning some of the code up to work with gcc and glibc.
None of this takes anywhere near as much time as the initial design and development of the game, thus it also doesn't cost nearly as much.
A significant portion of the work for games is artwork, this also is eliminated from Loki's work, the graphics/artwork/models are already there and done.
Well, there's always braille machines, as I understand it, they're pretty OS-independant so long as it's text mode.
Zipslack, the UMSDOS-based version of Slackware designed to fit on a 100MB ZIP disk has a variant that uses a peice of software to "speak" the words through a sound card.
IBM has Via Voice for Linux for dictating, not GPL/compatible of course, but it is there.
There are magnifier utilities for X for those that are visualy impared but not totally blind.
Things like Microsoft's "stickykeys" and "mousekeys" I don't know if those exist, but both could probably be implimented in user-space, allowing easy porting to BSD and such, or they could be implimented as a kernel patch.
There are solutions avalible, some just aren't as obvious as others.
Chris Hiner already mentioned Via Voice from IBM, I'm just going to say that I'm running on the assumption that in 1997, a reporter wasn't going to be using linux to write their stories :)
Repetitive Stress Injury is a generic term for conditions like Carpal Tunnel Syndrome, it's used CONSTANTLY in the medical field, and on slashdot itself.
what are they supposed to do? buy Dragon Naturally Speaking
if they'd truely evaluated it in 1997, they would have come to the conclusion that she could continue to do her job with assistance from software and a microphone
they didn't evaluate anything, they just decided they wanted to get her off their hands
it's actually not as funny as you may think, programmers DO work in standard programming languages like C or Perl with dictation software, it simply requires some customization of how things are recognized. Most, if not all VR software can be customized sufficiently to do coding.
You don't get it, it's not emotional/mental stress, it's physical stress, the same thing that makes a stick break if you bend it enough. Typing for years on keyboards can cause Repetitive Stress Injury, this is where the repetitive nature of the task over long periods causes stress on the nerves, muscles, etc. in the hands wrists and arms.
RSI has also been seen in professions completely unrelated to computers, such as a person working in a factory doing the same thing for years, even a blacksmith making horseshoes for years could aquire RSI.
you could try something like learning that "Stress" doesn't reffer to EMOTIONAL stress in repetitive stress injury
it refers to stress put on the PHYSICAL components of the hands/wrists/arms, STRESS is what makes a stick break if you bend it enough.
10 years ago we didn't have viable voice recognition software. Four years ago when this reporter was fired, we DID. The employer spent a bunch of money on chair and workstation modifications instead of spending it on voice recognition software.
Consider, a blind person can dress him/herself, do laundry, make a bed, even shop in some cases. They cannot drive, of course, however they can get around by walking or bus. And in some cases, someone with carpal tunnel or similar may not be able to drive safely, as the pain it causes their wrists can be extreme.
There was one comment on slashdot recently about a man who owned a small computer company, who was completely blind, but walked around without a cane or any sort of guide, built, troubleshooted, took apart, added to, computers by touch, using brail or text-to-speech for interaction with the computers, etc.
However, if they had fired this reporter because she'd gone blind, she would likely have won without trouble.
Voice recognition is COMPLETELY viable, even in 1997 it was usable, esspecialy if the user had some use of their hands to allow for manual corrections when neccisary. All they would have had to do was spend a few hundred dollars on Dragon Naturally Speaking, and a few bucks on a microphone, and everything would have been fine.
Why not? There's many programmers out there that use voice recognition for a substantial part of their coding. Marketing people can easily use voice recognition. Prehaps the only major roles that can't be adequately performed is QA and system administration work, and even those CAN be performed in the right enviorment.
You're throwing away good talent because they can't type for extended periods... that's just sad.
Bad idea, you risk having stupid netadmins that impliment DHCP so poorly that it's difficult or impossible to get it working on a non-windows machine. Better to keep it as-is and have all the routers block DHCP messages.