They aren't making the GPU more complex, nor adding capabilities. They are leveraging the fact that the GPU is capable of running these simple transforms *without modification of the processor*. RTFA, nowhere do they indicate that this is anything than a stock Nvidia GPU...
So where is the problem? The GPU is not being modified or made more complex or getting any added capabilities. Truth be told, you could simplify and strip some capabilities from a GPU and still have a decent DSP.
Think for a second, any modern video card is capable of displaying 16M discrete colors, right? The equivalent in the sound world are notes, of which in the chromatic scale there are 12. Even running octaves through the entire human audio spectrum still prduces less than 16 million notes. We can even add all those tiny little frequency shifts that individual instruments may have and still be under 16 million notes.
I could see your concern if doing this involved fundamental changes to the GPU, but that is precluded by this method. What the developers have done is found the one piece of silicon in your PC which is closest to being a DSP. In point of fact it is more capable and powerful than a built for purpose DSP. The game designer doesn't give a rat's ass what you do with your GPU when he isn't using it, so why not use it, especially since it does not require modification or addition of any additional silicon.
If you want to knock it, try and hit something relevant. For example, the GPU is not a DSP, that seems to indicate that there is some software magic which is interpreting what the GPU spits out. If you ask me, this is where the flaw is likely to be, not in some non-existent complexity or extra capability...
First and foremost, as the poster points out, the author admits that he is casual in citing sources. Wunderbar! So I'm supposed to accept academic criticism of the academic process by a supposed academic who fails to follow the process and cite his sources? Ummm, pass, thanks-no.
To extend an old adage:
Those who can, do. Those who can't, teach. But those that can neither do, nor teach, cop-out and claim the system is sabotaged, allways has been and is unfixable....
I'd like to thank the author for quitting. The last thing I want my children exposed to is that kind of example from their teachers.
I agree, it is a sad thing that parent's must put their children into public schools, particularly in view of what I've read today. Not because of the author's poorly cited references, or shoddy logic. No because I fear that the author is not the only defeatist teacher out there seeking excuses for his/her failures rather than motivation to succeed. That will damage my children more than any statement made by a coal miner in 1871.
No really, I'm interested to hear why this is a problem. Ever heard of any assembly line? Yeah, that's the thing where a group of simple workers outperforms a group of complex workers. Cars can be built by braking down the task into a number of simple tasks that can be completed without leveraging expensive resources. Similar to the GPU and CPU, the GPU is capable of performing a rather significantly larger number of simple operations compared to the CPU. Similar again to audio editing, which is nothing more than the application of a series of simple transforms.
There are tasks in audio eiditing more suited to the CPU than the GPU, but from what I read, you would be using the GPU for tasks suited to the GPU, and the CPU for tasks suited to it...
Finally, and anecdotally, I've ripped apart more than my fair share of electronic recording equipment, signal processing gear, etc. etc. etc. I've never found an Intel Inside sticker on one, nor anything that would be better than marginal as a general purpose CPU. Usually a bunch of ASICs, specifically DSPs. Your GPU is an ASIC too. These two ASICs have more in common with each other than either does with your AMD/Intel/VIA/Transmeta/what-have-you. The biggest difference is the GPU is evidently less differentiated and capable of a wider set of tasks than you would expect.
Your reaction seems to be one of "cool for it's own sake, but so what?" but this really is a case of using the right PU for the job...
I had a Prof in U who used to make similar statements. On one occasion a student took umbrage with this position and demanded data and proof. The prof indicated he would be happy to do so, as soon as the student was able to give him two problems, one 25% more complex than the other. After 15 minutes of demolishing every example offered by the student, the prof went on to explain that until the student was able to evaluate complexity in a meaningful and coherent fashion, any such proof would be a waste of time. The student was citing examples, which as the prof demonstrated, were consistent with scaling not complexity. A 25% increase in scale, while easy to understand and quantify, is not (in most cases) a significant increase in complexity, and certainly not a 25% increase.
More often I think folk don't understand what a 25% increase in complexity really is. Once you have a valid framework to make this evaluation it is easier to see how this relationship works, well, after you evolve some way of evaluating solution complexity as well...
If a principle is only a principle when "big business does it" then what is the point of the original principle?
More to the point, the drafters of the GPL utterly rejected that proposition, violation of the GPL is violation of the GPL, regardless of how many employees you do or don't have.
Lastly pally, you seem to think that everyone should just pipe down and be happy they are getting something for nothing. They aren't, this free software doesn't spontaneously code itself, real people do that. When real people contribute their time and effort with no financial compensation and little formal recognition, perhaps their motives are worth examining. When you begin this examination, you start to realize that only certain projects seem to collect these voluntary laborers, and the vast majority of these are GPL. The sophisticated observer at this point will stop, ignore the fact that the software is free, and maybe begin to wonder why this is, and how come there is so much of it. I'll submit to you that maybe, just maybe, it is BECAUSE of the GPL that all that free software is available.
So perhaps we should look a little deeper before making crass observations about just being happy things are free, and maybe look at some of the why's and wherefore's and maybe developing a more sophisticated view of F/OSS than "something for nothing" (Which it most certainly is not, ask some-one who writes code...)
As for X-chat, every individual who has contributed code needs to step up and demand the code be removed. The rest of us should remove the program, cease any participation in the development thereof, and make clear to the developer that we cannot accept his interpretation of the GPL, and that no OSS project can survive in an environment of apathy, which his current actions are virtually guaranteed to create.
In summation, lock both the developer and the original poster in a small closet with RMS.
First, you aren't going to put 15-20 Xeons in place of a single Alpha, if only because the majority of your x86 budget is going to be chewed up by redundant costs. Consider: I can stick that Alpha in a small server closet, with a smaller environmental control system, a smaller UPS/backup generator, and a smaller fire supression system. Best of all, only one system to worry about backups and administration for.
Now, I simply am not going to be able to beat that with 15-20 Xeon boxes, even nice flat 1U rackmounts are going to have higher requirements in all those critical categories. Never mind system/cluster administration, SAN crap, etc. etc.
The point is there are other criteria than simply cost and processing power. If I made all my decisions based on price vs performance without taking into consideration reliability, ease of administration, etc. etc. etc., well let's just say my employer wouldn't be worried about my/. time, since I wouldn't be employed here.
Ultimately diversity is a good thing. In fact diversity is the underlying reason for the success of the x86 architecture. The thesis that there is no need for any other architecture, simply rack up the x86s, is irresponsible, short-sighted, and in fact flies in the face of reasonable analysis. Maybe taking pot shots at x86 chips is bullshit, but denying the value of other architectures, approaches and ideas is so much more bullshit.
Okay, so some people have had their sense of ha-ha programmatically removed...
But, since I have been attacked, I will defend. Allthough meant to be humorous, you will note I clearly delimited that the TCO studies in question were Microsoft-funded, not Pro-Microsoft. Is it possible to have a TCO study that is Pro-Microsoft and not Microsoft-funded? Of course! As you point out, you use the right tool for the job, and in some cases that may mean Microsoft.
As for your first question, do you not understand the quote? The choice is between one of three types of lies, not just the first two. To answer your question a pro-linux tco study (if a lie) would be #2, a damned lie, since it was not funded by Microsoft, and my personal viewpoint makes it a damned lie, as opposed to the ordinary version...
However, the only lie in all three posts so far is your assertion that TCO in general is a lie. Your trivial and irrelevant example bears this out. You could save money on gas by switching to a bike, would you save money overall, probably not, since the time required is prohibitive (and money is nothing more than a unit of time, as in my employer is going to be pissed about the 15bucks worth of time I'm wasting teaching you how to prosecute an argument...) TCO done wrong, (like the Microsoft funded TCO studies cited originally) are worse than useless. Any study which reaches a predetermined result, particularly one in keeping with the interests of the funding body..., need be considered suspect. Over-applying the results of a TCO study, for example to other businesses or sectors of the economy where the study was not taken (incidentally another feature of the Microsoft funded TCO studies...) is equally useless in the best case scenarios, and smacks of fraud in the worst case scenarios.
Ultimately, most people don't know how to evaluate the information being presented to them in a TCO study, nor are able to apply those results meaningfully to their own situation. These folk tend to question the value of the entire process, rather than questioning the process (something liable to yield results!) This bias tends to create situations where they discount an idea because of the method of presentation, comitting the egregious sin of omission, much like they would like to accuse the presentation method of doing...
To return to your facile example of the car/bike analysis, apart from lousy phrasing, there certainly could be conditions where riding the bike the 20 miles would be an overall time/money savings. I don't know if you are a 350 pound couch potato that could desperately use some activity to improve your health (and incidentally extend the amount of time you have on this Earth, which unless I'm mistaken would be a profit...) or if you are a health nut who spends $100/month to go to the gym and ride a stationary bike every day, in which case not paying for the privelege also saves those fees.
I suspect that if you think about it for five minutes that your problem is not with TCO studies, but poorly done TCO studies, much like my issue with poor analogies.
Oh please, nothing funny about those organizations, nor is there anything funny about the infantile viewpoints of their main detractors on/. Both are essentially equally reprehensible sides of the same PU-239 minted coin. Both are equally divorced from reality, equally untenable in the real world. Fortunately, in this forum there are people capable of thinking critically about these issues and the recognition of those peers is sufficient unto the moment.
My.sig is actually lauding those people who are able to prosecute an argument on it's merits, not on the basis of wether it lets them get others work without paying for it. Or who are sophisticated enough to deplore RIAA, MPAA, Et. Al. but fight back without resorting to piracy, and incidentally further proving the case which RIAA and MPAA are trying to make. Ultimately, my.sig is for those who think first, and then act, or speak, or make comments about.sigs that they patently do not understand....
...maybe they should spend their money on food and birth control...
And what exactly is the benefit of birth control to the head of a third world family? Remember, we're talking about subsistence farming here. So when you peddle birth control you're trying to sell someone on greater economic prosperity by denying him the only real way in his environment to materially increase his prosperity, more children.
In such an environment children are a resource, not an expense. Birth control is only attractive to a culture where children are an expense, not a resource. Until you materially bring up the overall level of prosperity in these cultures you cannot escape that simple economic reality.
So one is really as useless as the other, the only advantage to the internet appliance is it gives the illusion of greater prosperity, and a view to the wider world. But neither offering materially affects the root problem, until the fundamental inequities in the global distribution of wealth are addressed there is little hope to ending this situation.
Jeez, I'm not even that old, but y'all are making me want to get a walker...
When I went to school it was pretty clear breakup:
Faculty of Education - Apple & VAX/VMS
Faculty of Math/Comp Sci - Unix (SunOS) & VAX/VMS
Faculty of Management - Microsoft
Other Faculties - various or timeshare.
Now the CPSC labs I worked in were all dumb terminals, every last one. Multi-screen terminals (the secret of actually using them came 2nd year...) were on every desk. The only exception were the 4th year students, who were allowed to use X-terminals (HP, I think.)
In Education, where I worked as a lab monitor, the machines were all IIe's and g's typically running Claris Works and LOGO. (Quite an experience, nothing like trying to expain programming to a technophobe when every sentence is virtually guaranteed to include the phrase, "Well you got to tell the turtle...") The VAX was pretty much off-limits for students, and I think it was pretty much maintained by the computer science people, I know it was there, but was never allowed access.
I recently went back for a visit, and things are not so good anymore. The CPSC guys use Linux now, but everywhere else M$ reigns supreme. The Mac zealots still swing by the neck on the lawn of the Education Tower...
What bothers me, is that the University has raised tuition every year since a year or two after I left. Surely at least one of those increases could have been avoided if the University had chosen to eschew M$ costs. No matter how good the price break, it is hard to beat free.
The best part of that is we all remember the halcyon days when Educators, Educational Institutions and Students could get Mac's at *STEEP* discounts.
The funny thing is that it worked *so* well for Apple, they now have less than 3% of the desktop share...
Of course, the funniest thing is that M$, in true fashion, is copying their competitors once again.
But the punchline of this whole fiasco is that M$ is probably moving from the same unproved assumptions that Apple did. Hopefully when the strain of dealing with the OS becomes too much to bear, these folks will also vote with their wallets when the freebies come to an end.
Personally, I don't see per processor charges going away. My favorite scenario is $/processor-type. IE 1-4 single core CPUs cost you X/proc, but 1-4 dual core CPUs cost you Y/proc (where Y is not significantly higher than X)
The problem (as previously pointed out) is that the software companies do need to do more work for MP versions of software, if they can't reasonably expect to recoup such expenditures, why would they bother? But, on the other hand, the chip makers will have a vested interest in moving the multi-cored chips.
Split the difference, keep the software folk writing MP software and the hardware folk adding cores. Ultimately we all end up with a better computing experience.
A lot of folk have pointed out text editors. No matter what else, $EDITOR is your best friend for configuring your Linux system.
I've used various system configuration utilities/packages at various times, but each and every time they've been mv $TOOL/dev/null, usually for one of the following reasons:
- I need to do something just plain out of the ordinary, no config tool could possibly anticipate every contigency for every user. So if you are going to have to $EDITOR $CONFIG_FILE even once, you may as well stick to it and build the habits, and the proficiency to be effective without $TOOL.
- I don't trust them, and generally end up double checking everything they do anyways. In view of that, why take the extra step?
Having said all that, the right configuration tool can be useful to help teach sys config and admin.
An example of that would be sam, (HP/UX) which at least told you what it was doing.
If you like opaque sys admin and config tools, there is this Windows distro from some Microsoft company...
It's all perception, particularly Wall Street and the stock holders.
On one hand, how much more can M$ grow? Not all that much, and really, there is a greater likelihood of M$ losing market share than gaining more at this point.
On the other hand, M$ own business practices have had more to do with killing innovation than any single other factor.
On the gripping hand, cutting costs is something Microsoft can do, and something they pretty much must do in order to ensure that perceptions on The Street and in the minds of shareholders don't change too significantly.
At least until they decide to realize a key fact:
Innovation and Financial Responsibility are not inversely proportional to each other, only your core business decisions are creating that appearance.
Pray tell, what exactly was use of your little exercise there?
Other than exercising an odious habit, I mean.
Are you a grammar Nazi as well as a logic Nazi?
At least the grammar nazi supplies the correct form as part and parcel of his/her odious habit. You have done nothing of the kind.
So how, pray tell is someone to learn from your criticism? In other words, can you take your post and make it constructive criticism?
As it stands, your post is worse than useless. Simply pointing out errors does little, particularly when you cite things like affirming the consequent to someone who you have good reason to believe has no training in these concepts...
If you want to feel superior with your ability to think logically, fine, well and good. But I'll suggest that the warm and fuzzy feeling from feeling superior is a pale shadow of the warm and fuzzies from teaching others...
After all, let's take a page out of the grammar Nazi's book, they engage in their odious habit generally because poor grammar drives them up the wall. At least providing the correct grammar has some chance of ameliorating the trigger condition. On the other hand, if you fail to teach the "right way" then the situation never improves, which tends to make one suspicious of why you even bother in the first place...
Yeah, it's a troll, but no less a troll than logic nazism for it's own sake.
Slashdot Mirror
So where is the problem? The GPU is not being modified or made more complex or getting any added capabilities. Truth be told, you could simplify and strip some capabilities from a GPU and still have a decent DSP.
Think for a second, any modern video card is capable of displaying 16M discrete colors, right? The equivalent in the sound world are notes, of which in the chromatic scale there are 12. Even running octaves through the entire human audio spectrum still prduces less than 16 million notes. We can even add all those tiny little frequency shifts that individual instruments may have and still be under 16 million notes.
I could see your concern if doing this involved fundamental changes to the GPU, but that is precluded by this method. What the developers have done is found the one piece of silicon in your PC which is closest to being a DSP. In point of fact it is more capable and powerful than a built for purpose DSP. The game designer doesn't give a rat's ass what you do with your GPU when he isn't using it, so why not use it, especially since it does not require modification or addition of any additional silicon.
If you want to knock it, try and hit something relevant. For example, the GPU is not a DSP, that seems to indicate that there is some software magic which is interpreting what the GPU spits out. If you ask me, this is where the flaw is likely to be, not in some non-existent complexity or extra capability...
To extend an old adage:
Those who can, do. Those who can't, teach. But those that can neither do, nor teach, cop-out and claim the system is sabotaged, allways has been and is unfixable....
I'd like to thank the author for quitting. The last thing I want my children exposed to is that kind of example from their teachers.
I agree, it is a sad thing that parent's must put their children into public schools, particularly in view of what I've read today. Not because of the author's poorly cited references, or shoddy logic. No because I fear that the author is not the only defeatist teacher out there seeking excuses for his/her failures rather than motivation to succeed. That will damage my children more than any statement made by a coal miner in 1871.
More like: unrestricted.flamewar.slashdot.org
Nothing like adding a contentious subject to a contentious site...
Hell I started with the gateway drug, Space Invaders, and I've never looked back since!
No really, I'm interested to hear why this is a problem. Ever heard of any assembly line? Yeah, that's the thing where a group of simple workers outperforms a group of complex workers. Cars can be built by braking down the task into a number of simple tasks that can be completed without leveraging expensive resources. Similar to the GPU and CPU, the GPU is capable of performing a rather significantly larger number of simple operations compared to the CPU. Similar again to audio editing, which is nothing more than the application of a series of simple transforms.
There are tasks in audio eiditing more suited to the CPU than the GPU, but from what I read, you would be using the GPU for tasks suited to the GPU, and the CPU for tasks suited to it...
Finally, and anecdotally, I've ripped apart more than my fair share of electronic recording equipment, signal processing gear, etc. etc. etc. I've never found an Intel Inside sticker on one, nor anything that would be better than marginal as a general purpose CPU. Usually a bunch of ASICs, specifically DSPs. Your GPU is an ASIC too. These two ASICs have more in common with each other than either does with your AMD/Intel/VIA/Transmeta/what-have-you. The biggest difference is the GPU is evidently less differentiated and capable of a wider set of tasks than you would expect.
Your reaction seems to be one of "cool for it's own sake, but so what?" but this really is a case of using the right PU for the job...
until Oct 11 next year for Debian to percolate it out of unstable!
More often I think folk don't understand what a 25% increase in complexity really is. Once you have a valid framework to make this evaluation it is easier to see how this relationship works, well, after you evolve some way of evaluating solution complexity as well...
If a principle is only a principle when "big business does it" then what is the point of the original principle?
More to the point, the drafters of the GPL utterly rejected that proposition, violation of the GPL is violation of the GPL, regardless of how many employees you do or don't have.
Lastly pally, you seem to think that everyone should just pipe down and be happy they are getting something for nothing. They aren't, this free software doesn't spontaneously code itself, real people do that. When real people contribute their time and effort with no financial compensation and little formal recognition, perhaps their motives are worth examining. When you begin this examination, you start to realize that only certain projects seem to collect these voluntary laborers, and the vast majority of these are GPL. The sophisticated observer at this point will stop, ignore the fact that the software is free, and maybe begin to wonder why this is, and how come there is so much of it. I'll submit to you that maybe, just maybe, it is BECAUSE of the GPL that all that free software is available.
So perhaps we should look a little deeper before making crass observations about just being happy things are free, and maybe look at some of the why's and wherefore's and maybe developing a more sophisticated view of F/OSS than "something for nothing" (Which it most certainly is not, ask some-one who writes code...)
As for X-chat, every individual who has contributed code needs to step up and demand the code be removed. The rest of us should remove the program, cease any participation in the development thereof, and make clear to the developer that we cannot accept his interpretation of the GPL, and that no OSS project can survive in an environment of apathy, which his current actions are virtually guaranteed to create.
In summation, lock both the developer and the original poster in a small closet with RMS.
First, you aren't going to put 15-20 Xeons in place of a single Alpha, if only because the majority of your x86 budget is going to be chewed up by redundant costs. Consider: I can stick that Alpha in a small server closet, with a smaller environmental control system, a smaller UPS/backup generator, and a smaller fire supression system. Best of all, only one system to worry about backups and administration for.
Now, I simply am not going to be able to beat that with 15-20 Xeon boxes, even nice flat 1U rackmounts are going to have higher requirements in all those critical categories. Never mind system/cluster administration, SAN crap, etc. etc.
The point is there are other criteria than simply cost and processing power. If I made all my decisions based on price vs performance without taking into consideration reliability, ease of administration, etc. etc. etc., well let's just say my employer wouldn't be worried about my /. time, since I wouldn't be employed here.
Ultimately diversity is a good thing. In fact diversity is the underlying reason for the success of the x86 architecture. The thesis that there is no need for any other architecture, simply rack up the x86s, is irresponsible, short-sighted, and in fact flies in the face of reasonable analysis. Maybe taking pot shots at x86 chips is bullshit, but denying the value of other architectures, approaches and ideas is so much more bullshit.
But, since I have been attacked, I will defend. Allthough meant to be humorous, you will note I clearly delimited that the TCO studies in question were Microsoft-funded, not Pro-Microsoft. Is it possible to have a TCO study that is Pro-Microsoft and not Microsoft-funded? Of course! As you point out, you use the right tool for the job, and in some cases that may mean Microsoft.
As for your first question, do you not understand the quote? The choice is between one of three types of lies, not just the first two. To answer your question a pro-linux tco study (if a lie) would be #2, a damned lie, since it was not funded by Microsoft, and my personal viewpoint makes it a damned lie, as opposed to the ordinary version...
However, the only lie in all three posts so far is your assertion that TCO in general is a lie. Your trivial and irrelevant example bears this out. You could save money on gas by switching to a bike, would you save money overall, probably not, since the time required is prohibitive (and money is nothing more than a unit of time, as in my employer is going to be pissed about the 15bucks worth of time I'm wasting teaching you how to prosecute an argument...) TCO done wrong, (like the Microsoft funded TCO studies cited originally) are worse than useless. Any study which reaches a predetermined result, particularly one in keeping with the interests of the funding body..., need be considered suspect. Over-applying the results of a TCO study, for example to other businesses or sectors of the economy where the study was not taken (incidentally another feature of the Microsoft funded TCO studies...) is equally useless in the best case scenarios, and smacks of fraud in the worst case scenarios.
Ultimately, most people don't know how to evaluate the information being presented to them in a TCO study, nor are able to apply those results meaningfully to their own situation. These folk tend to question the value of the entire process, rather than questioning the process (something liable to yield results!) This bias tends to create situations where they discount an idea because of the method of presentation, comitting the egregious sin of omission, much like they would like to accuse the presentation method of doing...
To return to your facile example of the car/bike analysis, apart from lousy phrasing, there certainly could be conditions where riding the bike the 20 miles would be an overall time/money savings. I don't know if you are a 350 pound couch potato that could desperately use some activity to improve your health (and incidentally extend the amount of time you have on this Earth, which unless I'm mistaken would be a profit...) or if you are a health nut who spends $100/month to go to the gym and ride a stationary bike every day, in which case not paying for the privelege also saves those fees.
I suspect that if you think about it for five minutes that your problem is not with TCO studies, but poorly done TCO studies, much like my issue with poor analogies.
2) Damned Lies
3) Microsoft Funded TCO studies
My .sig is actually lauding those people who are able to prosecute an argument on it's merits, not on the basis of wether it lets them get others work without paying for it. Or who are sophisticated enough to deplore RIAA, MPAA, Et. Al. but fight back without resorting to piracy, and incidentally further proving the case which RIAA and MPAA are trying to make. Ultimately, my .sig is for those who think first, and then act, or speak, or make comments about .sigs that they patently do not understand....
And what exactly is the benefit of birth control to the head of a third world family? Remember, we're talking about subsistence farming here. So when you peddle birth control you're trying to sell someone on greater economic prosperity by denying him the only real way in his environment to materially increase his prosperity, more children.
In such an environment children are a resource, not an expense. Birth control is only attractive to a culture where children are an expense, not a resource. Until you materially bring up the overall level of prosperity in these cultures you cannot escape that simple economic reality.
So one is really as useless as the other, the only advantage to the internet appliance is it gives the illusion of greater prosperity, and a view to the wider world. But neither offering materially affects the root problem, until the fundamental inequities in the global distribution of wealth are addressed there is little hope to ending this situation.
I don't see a link to a M$-Word document among the choices...
Maybe they just backed the right group of grads...
When I went to school it was pretty clear breakup:
Faculty of Education - Apple & VAX/VMS
Faculty of Math/Comp Sci - Unix (SunOS) & VAX/VMS
Faculty of Management - Microsoft
Other Faculties - various or timeshare.
Now the CPSC labs I worked in were all dumb terminals, every last one. Multi-screen terminals (the secret of actually using them came 2nd year...) were on every desk. The only exception were the 4th year students, who were allowed to use X-terminals (HP, I think.)
In Education, where I worked as a lab monitor, the machines were all IIe's and g's typically running Claris Works and LOGO. (Quite an experience, nothing like trying to expain programming to a technophobe when every sentence is virtually guaranteed to include the phrase, "Well you got to tell the turtle...") The VAX was pretty much off-limits for students, and I think it was pretty much maintained by the computer science people, I know it was there, but was never allowed access.
I recently went back for a visit, and things are not so good anymore. The CPSC guys use Linux now, but everywhere else M$ reigns supreme. The Mac zealots still swing by the neck on the lawn of the Education Tower...
What bothers me, is that the University has raised tuition every year since a year or two after I left. Surely at least one of those increases could have been avoided if the University had chosen to eschew M$ costs. No matter how good the price break, it is hard to beat free.
The funny thing is that it worked *so* well for Apple, they now have less than 3% of the desktop share...
Of course, the funniest thing is that M$, in true fashion, is copying their competitors once again.
But the punchline of this whole fiasco is that M$ is probably moving from the same unproved assumptions that Apple did. Hopefully when the strain of dealing with the OS becomes too much to bear, these folks will also vote with their wallets when the freebies come to an end.
I was more interested in what y'all thought.
Personally, I don't see per processor charges going away. My favorite scenario is $/processor-type. IE 1-4 single core CPUs cost you X/proc, but 1-4 dual core CPUs cost you Y/proc (where Y is not significantly higher than X)
The problem (as previously pointed out) is that the software companies do need to do more work for MP versions of software, if they can't reasonably expect to recoup such expenditures, why would they bother? But, on the other hand, the chip makers will have a vested interest in moving the multi-cored chips.
Split the difference, keep the software folk writing MP software and the hardware folk adding cores. Ultimately we all end up with a better computing experience.
I've used various system configuration utilities/packages at various times, but each and every time they've been mv $TOOL /dev/null, usually for one of the following reasons:
Having said all that, the right configuration tool can be useful to help teach sys config and admin.
An example of that would be sam, (HP/UX) which at least told you what it was doing.
If you like opaque sys admin and config tools, there is this Windows distro from some Microsoft company...
Hell, not only do I have the games, I still have BASIC, PILOT, and probably LOGO somewhere...
Say it with me now...
It's frigging excellent geek Karma...
Let's face it, the only guy getting better Karma than you is the maintainer on the OpenBSD Vax port...
After all, you're just donating endangered hardware, he is actively developping for it...
But, it's not wasted work kids, not as long as we have a VAX emulator!
It's all perception, particularly Wall Street and the stock holders.
On one hand, how much more can M$ grow? Not all that much, and really, there is a greater likelihood of M$ losing market share than gaining more at this point.
On the other hand, M$ own business practices have had more to do with killing innovation than any single other factor.
On the gripping hand, cutting costs is something Microsoft can do, and something they pretty much must do in order to ensure that perceptions on The Street and in the minds of shareholders don't change too significantly.
At least until they decide to realize a key fact:
Other than exercising an odious habit, I mean.
Are you a grammar Nazi as well as a logic Nazi?
At least the grammar nazi supplies the correct form as part and parcel of his/her odious habit. You have done nothing of the kind.
So how, pray tell is someone to learn from your criticism? In other words, can you take your post and make it constructive criticism?
As it stands, your post is worse than useless. Simply pointing out errors does little, particularly when you cite things like affirming the consequent to someone who you have good reason to believe has no training in these concepts...
If you want to feel superior with your ability to think logically, fine, well and good. But I'll suggest that the warm and fuzzy feeling from feeling superior is a pale shadow of the warm and fuzzies from teaching others...
After all, let's take a page out of the grammar Nazi's book, they engage in their odious habit generally because poor grammar drives them up the wall. At least providing the correct grammar has some chance of ameliorating the trigger condition. On the other hand, if you fail to teach the "right way" then the situation never improves, which tends to make one suspicious of why you even bother in the first place...
Yeah, it's a troll, but no less a troll than logic nazism for it's own sake.