Actually, it was because people had more and more binary modules with problems that resulted in kernel panics and the like, and would ask about a panic in a mailing list. Before, it would take someone's time to dig around and trace it down, and get more config info before realizing that there was a vendor's proprietary code causing the problem. They then get aggravated they wasted their time chasing a problem they aren't responsible for and can't fix.
With a tainted kernel, developers immediately know that the scope of the problem will quite likely be outside their ability to figure out, and can request a reproduction without a binary module loaded. If it can't be reproduced, contact the binary module provider for support...
People at the tops of these companies are *always* posturing to go somewhere else, and don't plan to be wherever they are currently at for the long haul. With that in mind, short term extreme gain (even up to 3 to 5 years can be called short-term) makes them look really good to those that would move them, and few think about the really long term viability of the strategy (10, 20 years down the line).
The huge flaw I see in the outsourcing trend is that development costs are spent in region y. Meanwhile, company relies on economy in region x to buy the product, because the people of region y cannot afford the price the company wants to charge. The problem then seems obvious. It works for a significant number of companies to cheat, but at some point things will reach a breaking point. If a company pays employees at a rate befitting a certain cost of living typical of a region, but does not price anywhere near consistantly with that regions cost of living, it is unbalanced and simply works due to other companies not cheating and accumulated wealth in their target consumer region.
With this in mind, things should either be priced in accordance with the producing region's standard of living, or simply marketed at the same region that produces it. It seems very wrong for a company to lay people off and then rely on that same group of people to buy their stuff, because their new employees can't afford it.
Despite all this outsourcing, the costs of the results has come down little, so the development savings aren't really being passed on to consumers, but pocketed by higher level management who pat themselves on the back and get promoted.
I know, the same things were said about manufacturing, and the US largely absorbed that outsourcing. Now white-collar jobs are being outsourced and people like me are again screaming the sky is falling. Maybe IT outsourcing will also be absorbed in time. Maybe Biotech jobs can go too, and we'll busy ourselves with something, but the US is starting to mess with jobs that require more and more training and skill, and retraining becomes increasingly difficult. One day, one industry will mark the point where there isn't enough economy left to sustain the consumer market for these companies' products, and then it'll be interesting to see what happens.
Your point makes sense, but at the same time it seems like EULAs for even the smallest applications are made so intentionally long and crammed to the brim with unimportant details that the things the user might be upset with are glossed over as quickly as possible. It seems the intent is to have a lot of innocuous stuff in there to dissuade the 'ardent' users from finisihing the agreement before clicking 'I agree', or at least get them impatient enough to skim and probably miss the part they would care about.
Kind of like going to buy a car and having the dealer saying 'just sign here' after he flips through 7-10 pages of material to get to the signature line, to pressure you into signing the agreemnet without fully knowing the terms. I know there is no urging of the user to click faster, but there is a sort of implicit expectation that the user will at most see a couple of lines and say 'ok, standard stuff, whatever, *cilck*'.
-multiple desktops, Windows has it via at least the PowerToys, probably others, back when I had to use windows, I did get multiple desktops. Oh, and I just remembered the other I used, JSPager, it was very convenient and flexible. -console switching: local console switching is indeed something I'm not sure you can do in OSX or Windows. However, in a pure GUI environment with 'user switching' capability, it becomes a moot point. In fact, the OSX and Windows User-Switching feature is convenient. You can accomplish a similar affect w/ multiple X servers running, but you have a fixed number of VCs allocated to running X and people have to lock their screen and remember which VC they used, not so clean. -unliminted customizability in appearance: In both OSX and XP their are plentiful skinning options. By default as a matter of usability, both environments don't alow rearranging the standard window buttons and shapes, but in Windows their is windowblinds which goes beyond what even XP allows, so yes, other platforms have the same customization. -multiple users and multiple X sessions: If you mean people can ssh in and run apps remotely, MS has Terminal Server which does the analog operation. If you mean multiple X servers can run on the same box, well, I already mentioned that User Switching is actually a cleaner approach, but honestly does not scale to X when you have programs connecting to ':0' not knowing who the owner would be. Perhaps one day their will be an implementation where X server numbers are alloced dynamically in the user switching context. -console sessions: OSX certainly has them. If by console you mean no GUI, it is an either/or proposition in OSX, can't have them running concurrently. If you allow for xterm-like operation, Windows and OSX has it.
Ultimately, a lot of these features are not 'ease-of-use' features, which is what the article discusses, but increased capability for those that can understand beyond ease-of-use operation. On that score there are countless comparisons you could make, and probably make Linux look very good, but this is merely an ease-of-use consideration. I actually think depending on the distributor, it is really already there or almost there, but not beyond that yet.
As much as I love linux (it is my almost exclusive platform for work and home), I think it important to properly know the featureset of the options and make fair comparisons, rather than blindly assume stuff doesn't exist elsewhere without looking for more than 5 seconds and loudly proclaim this. Your argument is weakened when you show lack of knowledge of the other options.
Why, 43 million of course. It is *clearly* stated on the 'I agree' click-through licenses between lines 7,254 and 7,255, and *everyone* reads those agreements thoroughly, right?
Not even necessarily 'hide' the fact, anyone capable of understanding the code already is well aware that any affordable RAID card is just a plain device with a slightly fancier firmware that implements the true stuff at the software level. Hell, anyone looking at a card can tell at a glance it doesn't have hardware RAID by the complexity of the chips on the board.
No, the actual issue is they all believe their software RAID implementation is supreme and won't let anyone know how they do it, when ultimately they all have the same, shitty performance and frequently worse than the open source software RAID of linux already.
Hate to say, but I can't imagine getting off without a chipset even with Opteron doing that. You still need a southbridge, period. Without it, you have no way of having a serial console or video console (you need *something* so you can monitor initial setup), you have no way of connecting any type of hard drive. You need more southbridge functionality before you think of ditching the planar chipset, and AMD64's current approach is quite sensible for the current level of technology, and having a plug-and-play southbridge that has no way of knowing ahead of time the slot configuration of a motherboard is difficult and would just restrict the possiblities of motherboard makers...
Huh, burn out pixels? If the lighting the room aspect involves the display components themselves, it would only be the LCD backlight, and it may very well be a completely independent light. Once manufactured correctly, the most common problem I've seen in LCDs in normal operation is the backlight dying, and if that light is used more, then it will go out more quickly, but killing pixels through this operation isn't really likely at all, it is two totally different things.
I'm sorry, I can't be too sympathetic when they *currently* make $2.75 Million a season. I don't know how long they've been at that rate, but I can guarantee they have each already made more twice than I could ever hope to in my entire life.
However, they are quite possibly on the side of right, as everything is relative. If the executives at FOX pocket millions per episode for themselves, for example, it seems hardly fair that the voice talent is disproportionately paid.
As an aside, anyone know what the artists make with the Simpsons?
Even taking into account that is all that comes off the top of your head, it isn't impressive.
I'm within a 45 minute commute to Cisco, Nortel, IBM, three major universities, hordes of startups, Redhat, and many many other non-computer centric, but highly technical big companies.
And this state is #20 in the ranking. Granted the capital area is overly dense compared to the rest of the state, but that alone is a rather significant chunk of the hi-tech picture.
G5, given, not that impressive, and doubly unimpressive when Apple doesn't take full advantage of it (running 32-bit if I recall still...).
Linux distributions caught up? No, and it never can catch up with Apple's situation, Windows can't either. Linux can catch up with Windows (already has, depending on your tastes and distro) in user friendly, but it is not possible to catch up to OSX. Why? OSX is inherently nice, but the extra mile is that they *know* all the combinations of hardware they support, it is a small population, at a quick glance they know exactly what drivers are needed and exactly how things are configured in that system. There is no need to ask the user for drivers, or how to use things, the OS ships with all required drivers and the system designers have already decided for the user exactly how the hardware will function, for better or for worse. -Proprietary hardware/form factors: Yes, that is true, and I prefer and you prefer flexibility. You prefer flashiness (neon fans? sounds ugly...) and can get that, I prefer simplicity, I can get it. I can customize every detail of a PC, and I like that. Apple's default is a clean design that appeals to most and the target audience don't want choice, they just want whatever Apple gives them and wants it to work. This is a large, sensible target to market to, and their desires can only be met by restricting their choices they don't care about. -Single vendor: ties in to the previous point, really annoying (like the prospect of having to go to the dealer you bought your car from for everything?), but it is simply how they operate and helps them maintain tight control and hardware/software platform integration cleanly. Distribute among vendors and lean towards standards and the difficulty of making the OS 'just work' increases dramatically. -Industrial strength OS? OSX is, has BSD underpinnigs. The big Mac cluster is a clear testament to OSX being industrial strength. Just because it has flashy surface, does not mean the underpinnings are as flimsy as OS9.. You start digging into the OS more and you'll find it very interesting, and you can even ditch Quartz alltogether, use X or console and have a very FreeBSD like experience. Not that you need to for 'industrial strength' operation, but some people need to do/see that before they can believe in the platform.
-They should *not* scrap OSX. They finally got it right, don't screw with it. Sure, I'd love to see quartz ported to linux/x86 or even linux/ppc, but there is no value in it for Apple. BSD is equivalent *and* they don't have to worry about GPL details (thought they do contribute back as a good gesture through Darwin) -Moving to industry standard approach like PC world will dilute their image, confuse current market, and chances are the rest of the market don't give a rat's ass about having a G5 as there are AMD64 processors available that can run their x86 apps today and give them 64bit capabilities for tomorrow. -G5 processors at a reasonable price, I assume you mean systems. Yes, you and I think they aren't worth it, but Apple's marketing has determined it is the best balance of market share to profit margin. If it works for them, well, that's fine.
There is a disconnect of interests with the dream....
Users/administrators/etc all want Linux to be a consistant platform, with healthy competition and every provider essentially being a drop in replacement for another, to keep things perfectly competitive.... Competition is good for consumers.
Distributors care about the proliferation of a Linux platform.... *THEIR* Linux platform. If RedHat acheived 95% total market share, would SuSE be any more happier than MS having 95% market share? Benefits of being closer to SuSEs solution and therefore a cleaner transition for those with RedHat experience is offset by the solutions being so close together, there is no point of deviating to SuSE. Particularly among the more successful commercial Linux vendors, there is a minimal, political interest in embracing standards, but they want their config tools to be a) better than the rest and b) not easily usable with anything but their own distro, so as not to lose market share they could hold..
What does that have to do with this? MacOS pre OSX was complete crap for anything but a desktop (and even then, it was 'good' for the desktop in the sense that 95/98 were 'good' for the desktop....) The *BSD/Linux/Unices platforms are the considered alternatives here, not OS=9
Yeah, its amazing.... I changed the partition table without updating the vfat table and put an ext2 filesystem in the second partition.
The vfat partition stayed the same and the ext2 partition was non-zero size... woah....
Its just pesky random file corruption on both partitions you have to worry about...
In all seriousness: *THIS IS VERY VERY VERY DANGEROUS* DO NOT DO THIS *PERIOD*. It may give neat apperances at first, and both filesystems may appear fundamentally functional, but it will *CORRUPT DATA* when the first partition is populated enough to creep into the partition overlay.
Not quite true. This only affects how much the super user has reserved. You'll see that df reports the same for the 'size' column. The Avail column goes up simply because it reports with respect to what a normal user can write. System files owned by root could still be created when no space was available for user owned files. -m is not for file system repair, it is so that no user can make the system unusable for root. Don't set to zero. Even if your private workstation, if something goes awry and consumes your disk space as a user, your system can still log, can still write system tmp files and do that sort of thing allowing the user to fix the situation or else a super user to still log in, work with the system, and rectify the situation while still having persistant storage to work with.
Slashdot Mirror
Actually, it was because people had more and more binary modules with problems that resulted in kernel panics and the like, and would ask about a panic in a mailing list. Before, it would take someone's time to dig around and trace it down, and get more config info before realizing that there was a vendor's proprietary code causing the problem. They then get aggravated they wasted their time chasing a problem they aren't responsible for and can't fix.
With a tainted kernel, developers immediately know that the scope of the problem will quite likely be outside their ability to figure out, and can request a reproduction without a binary module loaded. If it can't be reproduced, contact the binary module provider for support...
That must have been *some* early beta there...
All we need is cold fusion.... then we'd be set....
No no, they wouldn't do something that horrible.
BTW I heard VIA discovered good solder could be derived from anthrax, I'm sure its unrelated...
More complex than that.
People at the tops of these companies are *always* posturing to go somewhere else, and don't plan to be wherever they are currently at for the long haul. With that in mind, short term extreme gain (even up to 3 to 5 years can be called short-term) makes them look really good to those that would move them, and few think about the really long term viability of the strategy (10, 20 years down the line).
The huge flaw I see in the outsourcing trend is that development costs are spent in region y. Meanwhile, company relies on economy in region x to buy the product, because the people of region y cannot afford the price the company wants to charge. The problem then seems obvious. It works for a significant number of companies to cheat, but at some point things will reach a breaking point. If a company pays employees at a rate befitting a certain cost of living typical of a region, but does not price anywhere near consistantly with that regions cost of living, it is unbalanced and simply works due to other companies not cheating and accumulated wealth in their target consumer region.
With this in mind, things should either be priced in accordance with the producing region's standard of living, or simply marketed at the same region that produces it. It seems very wrong for a company to lay people off and then rely on that same group of people to buy their stuff, because their new employees can't afford it.
Despite all this outsourcing, the costs of the results has come down little, so the development savings aren't really being passed on to consumers, but pocketed by higher level management who pat themselves on the back and get promoted.
I know, the same things were said about manufacturing, and the US largely absorbed that outsourcing. Now white-collar jobs are being outsourced and people like me are again screaming the sky is falling. Maybe IT outsourcing will also be absorbed in time. Maybe Biotech jobs can go too, and we'll busy ourselves with something, but the US is starting to mess with jobs that require more and more training and skill, and retraining becomes increasingly difficult. One day, one industry will mark the point where there isn't enough economy left to sustain the consumer market for these companies' products, and then it'll be interesting to see what happens.
Your point makes sense, but at the same time it seems like EULAs for even the smallest applications are made so intentionally long and crammed to the brim with unimportant details that the things the user might be upset with are glossed over as quickly as possible. It seems the intent is to have a lot of innocuous stuff in there to dissuade the 'ardent' users from finisihing the agreement before clicking 'I agree', or at least get them impatient enough to skim and probably miss the part they would care about.
Kind of like going to buy a car and having the dealer saying 'just sign here' after he flips through 7-10 pages of material to get to the signature line, to pressure you into signing the agreemnet without fully knowing the terms. I know there is no urging of the user to click faster, but there is a sort of implicit expectation that the user will at most see a couple of lines and say 'ok, standard stuff, whatever, *cilck*'.
-multiple desktops, Windows has it via at least the PowerToys, probably others, back when I had to use windows, I did get multiple desktops. Oh, and I just remembered the other I used, JSPager, it was very convenient and flexible.
-console switching:
local console switching is indeed something I'm not sure you can do in OSX or Windows. However, in a pure GUI environment with 'user switching' capability, it becomes a moot point. In fact, the OSX and Windows User-Switching feature is convenient. You can accomplish a similar affect w/ multiple X servers running, but you have a fixed number of VCs allocated to running X and people have to lock their screen and remember which VC they used, not so clean.
-unliminted customizability in appearance:
In both OSX and XP their are plentiful skinning options. By default as a matter of usability, both environments don't alow rearranging the standard window buttons and shapes, but in Windows their is windowblinds which goes beyond what even XP allows, so yes, other platforms have the same customization.
-multiple users and multiple X sessions:
If you mean people can ssh in and run apps remotely, MS has Terminal Server which does the analog operation. If you mean multiple X servers can run on the same box, well, I already mentioned that User Switching is actually a cleaner approach, but honestly does not scale to X when you have programs connecting to ':0' not knowing who the owner would be. Perhaps one day their will be an implementation where X server numbers are alloced dynamically in the user switching context.
-console sessions:
OSX certainly has them. If by console you mean no GUI, it is an either/or proposition in OSX, can't have them running concurrently. If you allow for
xterm-like operation, Windows and OSX has it.
Ultimately, a lot of these features are not 'ease-of-use' features, which is what the article discusses, but increased capability for those that can understand beyond ease-of-use operation. On that score there are countless comparisons you could make, and probably make Linux look very good, but this is merely an ease-of-use consideration. I actually think depending on the distributor, it is really already there or almost there, but not beyond that yet.
As much as I love linux (it is my almost exclusive platform for work and home), I think it important to properly know the featureset of the options and make fair comparisons, rather than blindly assume stuff doesn't exist elsewhere without looking for more than 5 seconds and loudly proclaim this. Your argument is weakened when you show lack of knowledge of the other options.
http://libvncserver.sourceforge.net/
It comes with x11vnc (forwards the X11 console),
and also with LinuxVNC (forwards the text console).
Also, at least with RHEL3 I've notice their X is compiled with a VNC X11 extension that does the same thing as x11vnc without a separate application.
Why, 43 million of course. It is *clearly* stated on the 'I agree' click-through licenses between lines 7,254 and 7,255, and *everyone* reads those agreements thoroughly, right?
Not even necessarily 'hide' the fact, anyone capable of understanding the code already is well aware that any affordable RAID card is just a plain device with a slightly fancier firmware that implements the true stuff at the software level. Hell, anyone looking at a card can tell at a glance it doesn't have hardware RAID by the complexity of the chips on the board.
No, the actual issue is they all believe their software RAID implementation is supreme and won't let anyone know how they do it, when ultimately they all have the same, shitty performance and frequently worse than the open source software RAID of linux already.
Hate to say, but I can't imagine getting off without a chipset even with Opteron doing that. You still need a southbridge, period. Without it, you have no way of having a serial console or video console (you need *something* so you can monitor initial setup), you have no way of connecting any type of hard drive. You need more southbridge functionality before you think of ditching the planar chipset, and AMD64's current approach is quite sensible for the current level of technology, and having a plug-and-play southbridge that has no way of knowing ahead of time the slot configuration of a motherboard is difficult and would just restrict the possiblities of motherboard makers...
mushroom mushroom.......
The browser was like *beep beep beep* and it ate my bookmarks...
And they were really good bookmarks too...
Huh, burn out pixels? If the lighting the room aspect involves the display components themselves, it would only be the LCD backlight, and it may very well be a completely independent light. Once manufactured correctly, the most common problem I've seen in LCDs in normal operation is the backlight dying, and if that light is used more, then it will go out more quickly, but killing pixels through this operation isn't really likely at all, it is two totally different things.
I'm sorry, I can't be too sympathetic when they *currently* make $2.75 Million a season. I don't know how long they've been at that rate, but I can guarantee they have each already made more twice than I could ever hope to in my entire life.
However, they are quite possibly on the side of right, as everything is relative. If the executives at FOX pocket millions per episode for themselves, for example, it seems hardly fair that the voice talent is disproportionately paid.
As an aside, anyone know what the artists make with the Simpsons?
And of course, when we note the evil bit being set, we just shoot the avian carrier out of the sky to drop the packet...
Even taking into account that is all that comes off the top of your head, it isn't impressive.
I'm within a 45 minute commute to Cisco, Nortel, IBM, three major universities, hordes of startups, Redhat, and many many other non-computer centric, but highly technical big companies.
And this state is #20 in the ranking. Granted the capital area is overly dense compared to the rest of the state, but that alone is a rather significant chunk of the hi-tech picture.
G5, given, not that impressive, and doubly unimpressive when Apple doesn't take full advantage of it (running 32-bit if I recall still...).
Linux distributions caught up? No, and it never can catch up with Apple's situation, Windows can't either. Linux can catch up with Windows (already has, depending on your tastes and distro) in user friendly, but it is not possible to catch up to OSX. Why? OSX is inherently nice, but the extra mile is that they *know* all the combinations of hardware they support, it is a small population, at a quick glance they know exactly what drivers are needed and exactly how things are configured in that system. There is no need to ask the user for drivers, or how to use things, the OS ships with all required drivers and the system designers have already decided for the user exactly how the hardware will function, for better or for worse.
-Proprietary hardware/form factors: Yes, that is true, and I prefer and you prefer flexibility. You prefer flashiness (neon fans? sounds ugly...) and can get that, I prefer simplicity, I can get it. I can customize every detail of a PC, and I like that. Apple's default is a clean design that appeals to most and the target audience don't want choice, they just want whatever Apple gives them and wants it to work. This is a large, sensible target to market to, and their desires can only be met by restricting their choices they don't care about.
-Single vendor: ties in to the previous point, really annoying (like the prospect of having to go to the dealer you bought your car from for everything?), but it is simply how they operate and helps them maintain tight control and hardware/software platform integration cleanly. Distribute among vendors and lean towards standards and the difficulty of making the OS 'just work' increases dramatically.
-Industrial strength OS? OSX is, has BSD underpinnigs. The big Mac cluster is a clear testament to OSX being industrial strength. Just because it has flashy surface, does not mean the underpinnings are as flimsy as OS9.. You start digging into the OS more and you'll find it very interesting, and you can even ditch Quartz alltogether, use X or console and have a very FreeBSD like experience. Not that you need to for 'industrial strength' operation, but some people need to do/see that before they can believe in the platform.
-They should *not* scrap OSX. They finally got it right, don't screw with it. Sure, I'd love to see quartz ported to linux/x86 or even linux/ppc, but there is no value in it for Apple. BSD is equivalent *and* they don't have to worry about GPL details (thought they do contribute back as a good gesture through Darwin)
-Moving to industry standard approach like PC world will dilute their image, confuse current market, and chances are the rest of the market don't give a rat's ass about having a G5 as there are AMD64 processors available that can run their x86 apps today and give them 64bit capabilities for tomorrow.
-G5 processors at a reasonable price, I assume you mean systems. Yes, you and I think they aren't worth it, but Apple's marketing has determined it is the best balance of market share to profit margin. If it works for them, well, that's fine.
There is a disconnect of interests with the dream....
Users/administrators/etc all want Linux to be a consistant platform, with healthy competition and every provider essentially being a drop in replacement for another, to keep things perfectly competitive.... Competition is good for consumers.
Distributors care about the proliferation of a Linux platform.... *THEIR* Linux platform. If RedHat acheived 95% total market share, would SuSE be any more happier than MS having 95% market share? Benefits of being closer to SuSEs solution and therefore a cleaner transition for those with RedHat experience is offset by the solutions being so close together, there is no point of deviating to SuSE. Particularly among the more successful commercial Linux vendors, there is a minimal, political interest in embracing standards, but they want their config tools to be a) better than the rest and b) not easily usable with anything but their own distro, so as not to lose market share they could hold..
Yes, even rpm without up2date or yast will do this...
FYI, there will never be a RedHat 10 (at least in the forseeable future), they are now at RedHat Enterprise 3, and dropped the 'non-enterprise' line.
What does that have to do with this? MacOS pre OSX was complete crap for anything but a desktop (and even then, it was 'good' for the desktop in the sense that 95/98 were 'good' for the desktop....)
The *BSD/Linux/Unices platforms are the considered alternatives here, not OS=9
Its called Hotmail. No screaming bloody murder, more like laughing our asses off...
Yeah, its amazing.... I changed the partition table without updating the vfat table and put an ext2 filesystem in the second partition.
The vfat partition stayed the same and the ext2 partition was non-zero size... woah....
Its just pesky random file corruption on both partitions you have to worry about...
In all seriousness:
*THIS IS VERY VERY VERY DANGEROUS* DO NOT DO THIS *PERIOD*. It may give neat apperances at first, and both filesystems may appear fundamentally functional, but it will *CORRUPT DATA* when the first partition is populated enough to creep into the partition overlay.
Not quite true. This only affects how much the super user has reserved. You'll see that df reports the same for the 'size' column. The Avail column goes up simply because it reports with respect to what a normal user can write. System files owned by root could still be created when no space was available for user owned files. -m is not for file system repair, it is so that no user can make the system unusable for root. Don't set to zero. Even if your private workstation, if something goes awry and consumes your disk space as a user, your system can still log, can still write system tmp files and do that sort of thing allowing the user to fix the situation or else a super user to still log in, work with the system, and rectify the situation while still having persistant storage to work with.