"All comics must go through articles for deletion"
Bollocks ime a life member of a uk based comunity organisation 18plus or Plus (similar to Rotoract) and I cant create and article about it.
Becuse some non uk Fuckwit with to much time on ther hands deletes the fucking article with no debate
So create an article with references to third-party reliable published sources (either web or print) that discuss your organisation, and post that. If it's deleted, take the deletion to Wikipedia:Deletion review and describe it as out of process for a well sourced article. It should be restored, if you're right.
Wikipedia deletion has its flaws, but they can be worked around, usually. As long as the article you want to write is about something that's important enough that people have published stuff about it.
In that case, the slashdot summary is misleading. It very clearly says "but are the admins listening?"
I think part of the problem is that to a casual wikipedia user, like most of those who have recently jumped on the webcomic deletion problem bandwagon (it's not like the phenomenon of these deletions has only just started), WP:AFD is a confusing place. It's tempting to think that people who comment there are in some way considered more important than you are. There's a lot of politicking going on behind the scenes that people might not be aware of (e.g. changes in notability guidelines), a lot of very technical discussion with frequent references to numerous policies, and it's easy to think that somebody weighing in with "*'''Delete''' does not meet [[WP:N]] due to lack of [[WP:RS]]; impossible to [[WP:V|verify]]. ~~~~" must be an administrator, just because they're clearly _so much more experienced_ than the casual user.
AFD is an intimidating forum, and I'm not sure what can be done about that. But I think we do need to do things to make it more welcoming for casual users.
Not to mention here (Randall answering audience questions at UIUC, including what the raptor entry points for the building are; non-torrent versions here; definitely worth downloading).
The solution is to _never_ put returned items back on the shelf. Then they'd know for sure that this customer was lying to them, rather than just suspecting it. But of course that isn't their policy, returned items do get put back on the shelf, often without being checked, and sometimes people do end up buying boxes full of building materials. So as long as there's the possibility that they've made this mistake, yes, they really do have to offer a refund in cases like this.
that didn't get checked upon return? If not, then I'd have to be as doubtful about that return as the manager was.
You can be as doubtful as you like (and personally I agree with you; this story does smell fishy), but unless you can prove it in a court of law, withholding the replacement drive and not issuing a refund is theft.
In the UK, and it's similar in most of Europe, we have the Sale of Goods Act. If a business sells things, it is responsible for those things being of merchantable quality. If they're not, it's the vendor's problem.
Crazy shit like this still happens, though. Remember the guy who had PC World refuse to fix his laptop with a broken hinge because he'd installed Linux on it? Or, more similarly, I have a friend who years ago purchased a CPU from a local shop, but when he got it home found it wasn't what he'd ordered, but a lower spec CPU wired to be overclockable. He tried to take it back to the shop for a refund, but they claimed that since it wasn't the CPU they sold him (duh!) they couldn't give him one.
so if the proccess this uses if infinitely reversible, does that mean that the usual maximum number of writes associated with flash memory will be gone? or just increased by a huge factor?
That's "almost" infinitely reversible. Elsewhere, they've claimed it will be good for something on the order of 10^13 write cycles, i.e. about a factor of 10^6 higher than Flash.
Someone mentioned that it takes an extraordinarily small amount of energy (1 pJ) to flip bits.
Will this be stable in the field? I mean, EMF should be able to elicit those kinds of potentials fairly easily.
I'm not sure about the energy requirements, but something I just read suggested it'll need about 0.25v to flip a bit. I don't suspect ordinary EMF exposure will be able to build up that kind of potential on such a small structure.
.. my gut says "no". In my experience, announcements of revolutionary technology that are much more than just a few months away from commercialization are typically attempts at funding for research on a proposed project, rather than an actual announcement of an existing viable product.
It seems they've been working on this, with funding, for the last 10 years. This technology has been licensed to actual NVRAM manufacturers (Infineon, according to their web site). I'm inclined to give them the benefit of the doubt on this one.
How about speed, durability, mean time before failure, etc.
FTA:
Kozicki says the process is like condensing a crystal from a solution, except that the process is almost infinitely reversible. If the PMC is fed a positive charge, the copper atoms return to their previous free-floating state, and the nanowires disassemble.
"Almost infinitely reversible" suggests longer lifetime than flash.
Also, the "free-floating state" of the atoms not used to store 1s suggests that the "vaporware" tag is doubly appropriate.
Actually, those are page titles, not search terms... you can't just go to "http:/en/wikipedia.org/wiki/search terms", you have to go to "http://en.wikipedia.org/wiki/Special:Search?search=search+terms&go=Go".
There probably ARE examples of prior art, but Wikipedia isn't one of them.
"1) Don't download and run random crap - that goes for any OS."
By this, I assume you mean "don't run any third party software not written by Microsoft".
Because the minute you do, the Windows Registry is no longer reliable (if it even is with Windows itself, which is questionable in itself), and eventually either Windows, the third party software, or a combo of the two will hose the Registry, thus bringing Windows to its knees.
I don't know HOW many times that has happened to people I've worked with.
Over the last 9 years, I've been running Windows NT with a wide variety of third party software from a variety of sources, and have never had the problem you describe.
OTOH, I don't download random crap like the fad-du-jour screensaver or P2P software without investigating what junk it might come with first.
"2) Sit behind a decent firewall - that also goes for any OS."
Which lets the Microsoft firewall out. Use a third-party firewall that blocks outbound and inbound connections, and allows greater freedom of configuration - ooops, you just run into problem number one.
Tell me one useful feature blocking outbound connections gives you. Go on. Tell me one, and I'll believe there's a good reason to upgrade from the Windows firewall. But it doesn't seem to me that there is one. At best, it creates a slight impediment to information theft that is trivially circumvented by any attacker with a clue.
The reason Linux works and keeps working is simple: there is NO REGISTRY!
When's the last time you heard anybody on a Linux forum say, "Oh, your drivers are corrupted - reinstall?"
"Drivers" don't get corrupted. The Registry gets corrupted - regularly.
On Linux, you can have buggy drivers or misconfigured drivers. You never have "corrupted" drivers.
Strange. On windows, I've had buggy and misconfigured drivers, but have never had "corrupted" drivers, whether due to the driver file itself or the registry becoming corrupted. In fact, I've never seen the registry become corrupted at all. Not on my installations of Windows, or on the installations of any of my friends, my work colleagues, or my company's clients. It seems to be an extremely solid and well engineered piece of technology.
The same applies to almost everything else in Windows? When's the last time you heard somebody say they had a problem with a Linux "corrupted TCP/IP stack"? How many tools and utilities does Windows provide to "repair a corrupted TCP/IP stack?" How many do you find for Linux?
Again, I've never seen this happen, and bear in mind that I have worked in professional IT support for Windows networks. I'm not sure what the tools in question do exactly, but its worth bearing in mind that changes to settings in Windows networking are persistent (e.g., if I execute a 'route add' statement, that route is not forgotten when I reboot), whereas for Linux you'd need to modify configuration files to have the same effect. These "corrupted TCP/IP stack" recovery tools probably exist simply to remove broken configurations that have been created, probably by user incompetence. It's also worth noting that while MS acknowledge the possibility of a "corrupted TCP/IP stack", it seems this is only considered possible on win98 and me systems.
Exactly. For example, all our old installers are now privilege-escalation exploits: the installer has the option to run the application after install; the application can be used to run arbitrary other applications. Vista requires our installer to run under an administrator account, therefore our installer can be used to run arbitrary applications as an administrator.
The ability to run code as an administrator is not a privilige escalation vulnerability if you have permission to run code (i.e. your installer) as an administrator. It might be annoying and user unfriendly, but it isn't a security vulnerability.
In that all-too-common scenario, what is MS supposed to do? Their main options are:
1. Revert to the old behavior, preventing all new software from benefitting from that functionality.
2. Keep the new, better behavior but endure endless whining from people using fundamentally broken software.
3. Change the program loader to wrap each executable and library with checks like
if progname == 'adobeBrokenv1.3' { use old api }
Or 4. Introduce a new version of the API that supports the extended features, while retaining the old one for backwards compatibility.
MS have been using a combination of 3 and 4 with substantial success for years. WinXP even introduced an entirely new mechanism for achieving this, WinSxS, whereby an application can contain metadata requesting specific versions of the APIs it links to.
A lot of this stems from the fact that, by default in XP all users were Admins (yes, not secure...but that is how it is/was).
No, it isn't. The first user you create during installation is an admin, but the rest aren't unless you explicitly make them so. Try it now. Here's what I do:
Start/Admimistrative Tools/Computer Management Local Users and Groups/Users Action/New user (fill in user name and password, click Create; then click Close) Right click on new user on list, select properties Switch to the "Member Of" tab. The list contains only "Users"
In the UK there was a bit of an uproar when a tabloid printed topless pics of a model called Linsey Dawn McKenzie the day after she turned 16. They had been doing a countdown to her birthday with less risque shoots. The paper insisted that the pics had been taken the minute she turned 16 and printed in the very next edition, though some people claimed at the time that they had been taken earlier, and that all the papers readers (viewers?) were paedos. What a difference a day makes....
Of course, if you have copies of those images now, you are required to destroy them. Minimum age for appearing in pron in the UK changed from 16 to 18 a few years ago, and almost nothing has been done to publicise the fact that thousands, if not millions, of people have images that are now illegal to possess./me regrets having to lose his copy of MO's LDMcK spreads, not to mention the Debee Ashton ones.
The law that was struck down was about age verification and ID requirements for PERFORMERS in the porn. It had nothing to do with the age of the people VIEWING the porn.
Well, yes. But in what way is the summary misleading, though? I mean, it _quite clearly_ states that "the requirements intruded on the privacy and safety of performers", which is quite obviously not about clickthrough age agreements.
Modern GPUs have (I think -- I don't keep up to date) 256 bit wide memory interfaces, running at close to gigahertz speed. This means they can transfer to and from their memory at about 4 times the rate a PC can.
Correction: modern GPUs have a 384 bit wide interface, running at 800MHz *dual-pumped*. Compare with the Intel Core 2's 128-bit wide, 333MHz dual-pumped interface, and you'll see that it's actually over 6 times faster at accessing memory.
Why is the GPU a processor dedicated to nothing but "pretty graphics" so much more powerful than the central multi-purpose processor even at the things like number-crunching?
You need to rephrase your question, because it makes an incorrect assumption. Here:
Why is the GPU a processor dedicated to nothing but "pretty graphics" so much more powerful than the central multi-purpose processor especially at the things like number-crunching?
The answer is obvious if you think about it: those "pretty graphics" are a huge number crunching problem. That's all there is to it. GPU's, however, aren't very good at tasks that don't do exactly the same thing huge numbers of times. This is true of most applications. Including the applications that run on the PC to control what the GPU does in stuff like what the story's talking about.
Is it because the GPU engineers can completely redo the thing from scratch whenever they want to, whereas the CPU-designers are held back by the backwards-compatibility issues?
Partially. Modern GPUs have (I think -- I don't keep up to date) 256 bit wide memory interfaces, running at close to gigahertz speed. This means they can transfer to and from their memory at about 4 times the rate a PC can. This is possible because (1) graphics card manufacturers don't mind the types of memory they use changing on a virtually model-by-model basis and (2) they also don't mind being stuck with non-expandable memory that's soldered directly onto the card right next to the GPU.
It's also because GPU engineers can sacrifice a lot of the flexibility of a PC. So what if the pipeline stalls if all 32 threads aren't doing exactly the same thing at the same time? Most of the time, they will be.
Computer Science teaches, programmers aren't supposed to have to do "tricks" like this -- you code, and the translator (compiler or intepreter) will translate from your programming language to the hardware instructions.
So why did my CS course have a module where we learned how the hardware worked? About memory hierarchies? About SISD, SIMD and MIMD processors? Why does Knuth's The Art of Computer Programming, possibly the most important book ever written on CS, approach problems at an assembly language level? Why, in my CS course, did I learn two different kinds of assembly language (one CISC, one RISC)?
Because CS is concerned with a holistic view of computers. With the fact that they are machines for executing instructions, and what can be done with those instructions. With the fact that it may be more efficient not to specify that much detail, but also the fact that, from time to time, you do need to do that.
Like searching chess positions or recognizing text? I was under the impression it very limited and requires specific types of input with restrictions on which operations can be used.
GPUs aren't very good at algorithms that require a lot of indirection or a lot of jumps. They work because they arrange incoming data in a stream and process large chunks of it in parallel. Both of the applications you suggest are unlikely to benefit hugely from GPUs. Chess position ranking could probably work well; set up a queue of positions in a suitable encoding and stream them through the GPU, get a stream of scores out of the other end. The logic of deciding which positions to send to the GPU is *far* better handled by a traditional CPU.
Text recognition is probably a little trickier. It would work if you're looking for only a few strings in long streams of text, but every time you add another string you're likely to need several additional pipeline stages. You'll run out of GPU pretty quickly.
Disclaimer: not really an expert in this, just someone who's done a little research about processing architectures, and intends to play around with general-purpose GPU apps at some point in time.
I really don't see the need for a faster way to hack my computer.
Speaking as someone who's worked in outsourced IT support, having a faster way to hack my clients' computers is *really useful*. Because if there's one thing that's guaranteed, it's that the users will forget their passwords. Sometimes the admin ones.
Slashdot Mirror
"All comics must go through articles for deletion"
Bollocks ime a life member of a uk based comunity organisation 18plus or Plus (similar to Rotoract) and I cant create and article about it.
Becuse some non uk Fuckwit with to much time on ther hands deletes the fucking article with no debate
So create an article with references to third-party reliable published sources (either web or print) that discuss your organisation, and post that. If it's deleted, take the deletion to Wikipedia:Deletion review and describe it as out of process for a well sourced article. It should be restored, if you're right.
Wikipedia deletion has its flaws, but they can be worked around, usually. As long as the article you want to write is about something that's important enough that people have published stuff about it.
In that case, the slashdot summary is misleading. It very clearly says "but are the admins listening?"
I think part of the problem is that to a casual wikipedia user, like most of those who have recently jumped on the webcomic deletion problem bandwagon (it's not like the phenomenon of these deletions has only just started), WP:AFD is a confusing place. It's tempting to think that people who comment there are in some way considered more important than you are. There's a lot of politicking going on behind the scenes that people might not be aware of (e.g. changes in notability guidelines), a lot of very technical discussion with frequent references to numerous policies, and it's easy to think that somebody weighing in with "*'''Delete''' does not meet [[WP:N]] due to lack of [[WP:RS]]; impossible to [[WP:V|verify]]. ~~~~" must be an administrator, just because they're clearly _so much more experienced_ than the casual user.
AFD is an intimidating forum, and I'm not sure what can be done about that. But I think we do need to do things to make it more welcoming for casual users.
Not to mention here (Randall answering audience questions at UIUC, including what the raptor entry points for the building are; non-torrent versions here; definitely worth downloading).
But I'd have to buy a video camera first... see a problem here? :)
;)
It's okay, just take it out of the box, put a brick or two inside, reseal the box and return it saying you don't want it any more. Problem solved.
The solution is to _never_ put returned items back on the shelf. Then they'd know for sure that this customer was lying to them, rather than just suspecting it. But of course that isn't their policy, returned items do get put back on the shelf, often without being checked, and sometimes people do end up buying boxes full of building materials. So as long as there's the possibility that they've made this mistake, yes, they really do have to offer a refund in cases like this.
that didn't get checked upon return? If not, then I'd have to be as doubtful about that return as the manager was.
You can be as doubtful as you like (and personally I agree with you; this story does smell fishy), but unless you can prove it in a court of law, withholding the replacement drive and not issuing a refund is theft.
In the UK, and it's similar in most of Europe, we have the Sale of Goods Act. If a business sells things, it is responsible for those things being of merchantable quality. If they're not, it's the vendor's problem.
Crazy shit like this still happens, though. Remember the guy who had PC World refuse to fix his laptop with a broken hinge because he'd installed Linux on it? Or, more similarly, I have a friend who years ago purchased a CPU from a local shop, but when he got it home found it wasn't what he'd ordered, but a lower spec CPU wired to be overclockable. He tried to take it back to the shop for a refund, but they claimed that since it wasn't the CPU they sold him (duh!) they couldn't give him one.
so if the proccess this uses if infinitely reversible, does that mean that the usual maximum number of writes associated with flash memory will be gone? or just increased by a huge factor?
That's "almost" infinitely reversible. Elsewhere, they've claimed it will be good for something on the order of 10^13 write cycles, i.e. about a factor of 10^6 higher than Flash.
Someone mentioned that it takes an extraordinarily small amount of energy (1 pJ) to flip bits.
Will this be stable in the field? I mean, EMF should be able to elicit those kinds of potentials fairly easily.
I'm not sure about the energy requirements, but something I just read suggested it'll need about 0.25v to flip a bit. I don't suspect ordinary EMF exposure will be able to build up that kind of potential on such a small structure.
.. my gut says "no". In my experience, announcements of revolutionary technology that are much more than just a few months away from commercialization are typically attempts at funding for research on a proposed project, rather than an actual announcement of an existing viable product.
It seems they've been working on this, with funding, for the last 10 years. This technology has been licensed to actual NVRAM manufacturers (Infineon, according to their web site). I'm inclined to give them the benefit of the doubt on this one.
FTA:
"Almost infinitely reversible" suggests longer lifetime than flash.
Also, the "free-floating state" of the atoms not used to store 1s suggests that the "vaporware" tag is doubly appropriate.
1. I am not a microsoft shill.
2. I am not lying.
These problems are not as common as you think.
Sure, Windows ain't perfect, but it has a worse reputation than it deserves.
http://en.wikipedia.org/wiki/Prior_art
Actually, those are page titles, not search terms... you can't just go to "http:/en/wikipedia.org/wiki/search terms", you have to go to "http://en.wikipedia.org/wiki/Special:Search?search=search+terms&go=Go".
There probably ARE examples of prior art, but Wikipedia isn't one of them.
Agreed. The one that springs to mind is php.net.
"1) Don't download and run random crap - that goes for any OS."
By this, I assume you mean "don't run any third party software not written by Microsoft".
Because the minute you do, the Windows Registry is no longer reliable (if it even is with Windows itself, which is questionable in itself), and eventually either Windows, the third party software, or a combo of the two will hose the Registry, thus bringing Windows to its knees.
I don't know HOW many times that has happened to people I've worked with.
Over the last 9 years, I've been running Windows NT with a wide variety of third party software from a variety of sources, and have never had the problem you describe.
OTOH, I don't download random crap like the fad-du-jour screensaver or P2P software without investigating what junk it might come with first.
"2) Sit behind a decent firewall - that also goes for any OS."
Which lets the Microsoft firewall out. Use a third-party firewall that blocks outbound and inbound connections, and allows greater freedom of configuration - ooops, you just run into problem number one.
Tell me one useful feature blocking outbound connections gives you. Go on. Tell me one, and I'll believe there's a good reason to upgrade from the Windows firewall. But it doesn't seem to me that there is one. At best, it creates a slight impediment to information theft that is trivially circumvented by any attacker with a clue.
The reason Linux works and keeps working is simple: there is NO REGISTRY!
When's the last time you heard anybody on a Linux forum say, "Oh, your drivers are corrupted - reinstall?"
"Drivers" don't get corrupted. The Registry gets corrupted - regularly.
On Linux, you can have buggy drivers or misconfigured drivers. You never have "corrupted" drivers.
Strange. On windows, I've had buggy and misconfigured drivers, but have never had "corrupted" drivers, whether due to the driver file itself or the registry becoming corrupted. In fact, I've never seen the registry become corrupted at all. Not on my installations of Windows, or on the installations of any of my friends, my work colleagues, or my company's clients. It seems to be an extremely solid and well engineered piece of technology.
The same applies to almost everything else in Windows? When's the last time you heard somebody say they had a problem with a Linux "corrupted TCP/IP stack"? How many tools and utilities does Windows provide to "repair a corrupted TCP/IP stack?" How many do you find for Linux?
Again, I've never seen this happen, and bear in mind that I have worked in professional IT support for Windows networks. I'm not sure what the tools in question do exactly, but its worth bearing in mind that changes to settings in Windows networking are persistent (e.g., if I execute a 'route add' statement, that route is not forgotten when I reboot), whereas for Linux you'd need to modify configuration files to have the same effect. These "corrupted TCP/IP stack" recovery tools probably exist simply to remove broken configurations that have been created, probably by user incompetence. It's also worth noting that while MS acknowledge the possibility of a "corrupted TCP/IP stack", it seems this is only considered possible on win98 and me systems.
Exactly. For example, all our old installers are now privilege-escalation exploits: the installer has the option to run the application after install; the application can be used to run arbitrary other applications. Vista requires our installer to run under an administrator account, therefore our installer can be used to run arbitrary applications as an administrator.
The ability to run code as an administrator is not a privilige escalation vulnerability if you have permission to run code (i.e. your installer) as an administrator. It might be annoying and user unfriendly, but it isn't a security vulnerability.
In that all-too-common scenario, what is MS supposed to do? Their main options are:
1. Revert to the old behavior, preventing all new software from benefitting from that functionality.
2. Keep the new, better behavior but endure endless whining from people using fundamentally broken software.
3. Change the program loader to wrap each executable and library with checks like
if progname == 'adobeBrokenv1.3' { use old api }
Or 4. Introduce a new version of the API that supports the extended features, while retaining the old one for backwards compatibility.
MS have been using a combination of 3 and 4 with substantial success for years. WinXP even introduced an entirely new mechanism for achieving this, WinSxS, whereby an application can contain metadata requesting specific versions of the APIs it links to.
Why change strategy now?
A lot of this stems from the fact that, by default in XP all users were Admins (yes, not secure...but that is how it is/was).
No, it isn't. The first user you create during installation is an admin, but the rest aren't unless you explicitly make them so. Try it now. Here's what I do:
Start/Admimistrative Tools/Computer Management
Local Users and Groups/Users
Action/New user
(fill in user name and password, click Create; then click Close)
Right click on new user on list, select properties
Switch to the "Member Of" tab. The list contains only "Users"
In the UK there was a bit of an uproar when a tabloid printed topless pics of a model called Linsey Dawn McKenzie the day after she turned 16. They had been doing a countdown to her birthday with less risque shoots. The paper insisted that the pics had been taken the minute she turned 16 and printed in the very next edition, though some people claimed at the time that they had been taken earlier, and that all the papers readers (viewers?) were paedos. What a difference a day makes....
/me regrets having to lose his copy of MO's LDMcK spreads, not to mention the Debee Ashton ones.
Of course, if you have copies of those images now, you are required to destroy them. Minimum age for appearing in pron in the UK changed from 16 to 18 a few years ago, and almost nothing has been done to publicise the fact that thousands, if not millions, of people have images that are now illegal to possess.
See, I do [link]photography, sometimes of naked people[/link] for fun
;)
Way to get your site slashdotted. Well done.
The law that was struck down was about age verification and ID requirements for PERFORMERS in the porn. It had nothing to do with the age of the people VIEWING the porn.
Well, yes. But in what way is the summary misleading, though? I mean, it _quite clearly_ states that "the requirements intruded on the privacy and safety of performers", which is quite obviously not about clickthrough age agreements.
Note "128 processing units"... Am I supposed to assume they mean pipelines or just being clueless?
Well, given that a pipeline _is_ a processing unit, I don't see why either should be the case.
Modern GPUs have (I think -- I don't keep up to date) 256 bit wide memory interfaces, running at close to gigahertz speed. This means they can transfer to and from their memory at about 4 times the rate a PC can.
Correction: modern GPUs have a 384 bit wide interface, running at 800MHz *dual-pumped*. Compare with the Intel Core 2's 128-bit wide, 333MHz dual-pumped interface, and you'll see that it's actually over 6 times faster at accessing memory.
Why is the GPU a processor dedicated to nothing but "pretty graphics" so much more powerful than the central multi-purpose processor even at the things like number-crunching?
You need to rephrase your question, because it makes an incorrect assumption. Here:
Why is the GPU a processor dedicated to nothing but "pretty graphics" so much more powerful than the central multi-purpose processor especially at the things like number-crunching?
The answer is obvious if you think about it: those "pretty graphics" are a huge number crunching problem. That's all there is to it. GPU's, however, aren't very good at tasks that don't do exactly the same thing huge numbers of times. This is true of most applications. Including the applications that run on the PC to control what the GPU does in stuff like what the story's talking about.
Is it because the GPU engineers can completely redo the thing from scratch whenever they want to, whereas the CPU-designers are held back by the backwards-compatibility issues?
Partially. Modern GPUs have (I think -- I don't keep up to date) 256 bit wide memory interfaces, running at close to gigahertz speed. This means they can transfer to and from their memory at about 4 times the rate a PC can. This is possible because (1) graphics card manufacturers don't mind the types of memory they use changing on a virtually model-by-model basis and (2) they also don't mind being stuck with non-expandable memory that's soldered directly onto the card right next to the GPU.
It's also because GPU engineers can sacrifice a lot of the flexibility of a PC. So what if the pipeline stalls if all 32 threads aren't doing exactly the same thing at the same time? Most of the time, they will be.
Computer Science teaches, programmers aren't supposed to have to do "tricks" like this -- you code, and the translator (compiler or intepreter) will translate from your programming language to the hardware instructions.
So why did my CS course have a module where we learned how the hardware worked? About memory hierarchies? About SISD, SIMD and MIMD processors? Why does Knuth's The Art of Computer Programming, possibly the most important book ever written on CS, approach problems at an assembly language level? Why, in my CS course, did I learn two different kinds of assembly language (one CISC, one RISC)?
Because CS is concerned with a holistic view of computers. With the fact that they are machines for executing instructions, and what can be done with those instructions. With the fact that it may be more efficient not to specify that much detail, but also the fact that, from time to time, you do need to do that.
Like searching chess positions or recognizing text? I was under the impression it very limited and requires specific types of input with restrictions on which operations can be used.
GPUs aren't very good at algorithms that require a lot of indirection or a lot of jumps. They work because they arrange incoming data in a stream and process large chunks of it in parallel. Both of the applications you suggest are unlikely to benefit hugely from GPUs. Chess position ranking could probably work well; set up a queue of positions in a suitable encoding and stream them through the GPU, get a stream of scores out of the other end. The logic of deciding which positions to send to the GPU is *far* better handled by a traditional CPU.
Text recognition is probably a little trickier. It would work if you're looking for only a few strings in long streams of text, but every time you add another string you're likely to need several additional pipeline stages. You'll run out of GPU pretty quickly.
Disclaimer: not really an expert in this, just someone who's done a little research about processing architectures, and intends to play around with general-purpose GPU apps at some point in time.
I really don't see the need for a faster way to hack my computer.
Speaking as someone who's worked in outsourced IT support, having a faster way to hack my clients' computers is *really useful*. Because if there's one thing that's guaranteed, it's that the users will forget their passwords. Sometimes the admin ones.