You can all but bet that once two things have happened:
- Existing inventory of Q9x00 and Q6x00 parts is all but depleted and
- Dunnington's desktop-equivalent part has debuted
that this will indeed be the case. I mean, the yields on such a die would be absolutely shiteful, especially since this is the first monolithic design that intel have fabbed with more than two physical cores on it. If you can disable 2 cores and sell it as a Q9990 XXXXTREME edition, why wouldn't you?
To my mind, the interesting question will be whether we (a-la AMD) see 5-core dunnington-alikes show up. It'll come down to whether intel figure that the extra effort involved in isolating and binning two different sorts of quasi-defective parts (versus doing it once) would be rewarded by having a 5-core part on the market (which would, natch, sell for more than the equivalent, but differently binned, 4 core part.)
No, all that it infers is that doubling the capacity of the drive is physically possible, not that the requisite head assembly would be affordable (or, for that matter, fast) for a desktop disk.
The sort of setup you need to do this level of disk forensics would get you several tens of petabytes of cheap SATA disk.
> When I took my computer forensics class they showed that you could use a hex editor on a zero wiped floppy disk and recover most of the data that was on it previously. (emphasis mine)
If it was zero wiped, you'd see a meg and a bit of 0s. That said, if it's a floppy disk you're trying to wipe, the best way to go about it is a traditional low-level format, in which the drive mechanics physically remagnetise the surface of the media (this is the process which defines tracks and sectors.)
Cider _is_ native code. All the way down. The only real difference (user interface issues aside, and most games don't use the native UI in any event) is that each DirectX call obviously goes through one more stack frame before it hits the hardware (game->cider directX->OpenGL->driver rather than game->directX->driver).
It's been my experience that the speed difference in Cedega (or lack thereof) from a 'native' DirectX implementation is marginal. With Cider, one would assume any differences would be more marginalized again given the fact that the code is in a native binary format and the vast bulk of the application isn't making calls through translated APIs.
It'll be interesting to see if anybody takes the freely available (albeit not open source) Cedega DirectX 9 code (or, for that matter, the open source but somewhat slower Wine DirectX 9 code) and adds the DirectX 10 APIs to it. A system like this would, in theory, allow DirectX 10 apps to run on Windows XP as well as other GL platforms.
You can't patent an instruction set, because an instruction set is an interface, not an implementation. You certainly can't copyright one for largely the same reason. There've been court rulings on this but i'm too lazy to look it up.
Of course, that's not saying that the 2 or 3 most efficient ways of implementing SSE3 etc aren't patented to the hilt, that might be the case. But the situation is nowhere near as dire as people are making out.
> How did anyone get "vastly different GPUs" from this?
Presumably because (for e.g.) a G70-based 7800 and a G92-based 8800GT are vastly different GPUs.
G70, for example, had two sets of fixed-purpose pipeline units (one of which ran your vertex programs, and one of which ran your fragment programs,) a bunch of fixed-function logic, some rasterisation logic, etc.
On the other hand, G80 and above have general purpose 'shader processors' any of which can run any pipeline programs (and, afaik, runs the traditional graphics fixed-function pipeline in software on said SPs), and a minimal amount of glue to hang it together.
About the only thing that current-generation GPUs and previous-generation GPUs have in common is the logo on the box (this applies equally to AMD parts, although the X19xx AMD cards, i'm told, are more similar to a G80-style architecture than a G70-style architecture, which is how the F@H folks managed to get it running on the X19xx before G80 was released.)
Yes, but the parent post was simply asserting that DX10 bought nothing to the table over DX9.
I'm well aware that OpenGL 3.0, as well as Cg 2 (in conjunction with either GL 3.0 or some proprietary NV extension which I think is called gp40), are capable of geometry programs.
My point was simply that DX 10 has geometry program support and DX9 doesn't.
Whether we like it or not, the world needs telemarketers
No it doesn't.
Next time a telemarketer calls during a nice dinner with your family, remember, you don't have to answer the phone, so if you do, you're the one interrupting your dinner, secondly, they most likely have a family that they wish they could be with, but instead they have to call you.
Next time a telemarketer calls during a nice dinner with my family and I answer the phone because it may ACTUALLY BE AN IMPORTANT CALL (newsflash, not everybody that you may want to talk to is capable or desirous of sending CLI), I am completely justified in flying totally off the fucking handle. There is no justification for a FOR-PROFIT BUSINESS calling my house after BUSINESS HOURS. My usual response to telemarketers is "please justify to me in 50 words or less why what you're doing is more ethical than sitting on the dole." The response to this is usually... interesting.
What MS call "Shader model 4" (even though geometry programs aren't, strictly speaking, shaders as they don't necessarily SHADE anything per se) includes mandatory support for geometry programs.
The geometry program sits in the programmable pipeline between the vertex program (which is used for real-time vertex deformation in hardware, world-space to object-space clipping to generate texcoords for the fragment program, etc) and the fragment program (which is used to colour fragments [1] based upon the output of the vertex program and input from one or more texture samplers.)
Unlike "old" vertex programs, a geometry program is able to generate new geometry on the fly. This allows a whole heap of really cool stuff, such as real-time shadowing effects, for essentially free.
So, yeah, much as I hate to admit it (and REALLY hate the Direct3D 'shader' nomenclature concerning pipeline programs,) D3D10 actually has changes with merit from D3D9c.
[1] A fragment is a fancy name for a voxel defined in clip space. After shading and occlusion, the remaining fragments become rasterised as pixels. Thus, the term 'pixel shader' is rather inaccurate.
My understanding is that it was more about assaying fears of a so-called 'plutonium economy', never mind the fact that attempting to get any sort of weapons-grade plutonium out of IFR fuel would almost certainly kill you before you managed to do so.
It was a piece of political expediency at a time when nuclear energy was being demonized for two main reasons (fear of loss-of-coolant and fear of long-lived actinides). The stupid thing, of course, is that the IFR and its ilk are the only nukes that SOLVE that problem, and people are out there now talking about building more PWRs and BWRs to 'fix global warming.'
I suppose as long as someone, somewhere decides to revisit the closed-fuel-cycle fast reactor concept, we don't have too many worries in terms of the waste (because SOMEONE'll use it for fuel long before its 20,000-odd years are up.) The concern, of course, is doing that before it all gets dumped under some mountain somewhere and forgotten about forever.
Love where you're going with your website. Didn't think there were too many other people out there irate with Clinton for killing off EBR2/Clinch River. You just made my friends list.
It's a hard call in some cases. There's an argument that most (any?) multihomed customers should be able to send packets asymmetrically. That said, nobody with even half a brain should be running a NAS/LNS with single-homed customers WITHOUT RPF, although the (admittedly, in.au) number of people i've crossed paths with who are doing it wrong is staggering.
Conversely, if i'm peering with you and six other guys, and happen to be carrying traffic to your network (but not advertising a route back to some of MY peers, for e.g. traffic engineering reasons) across a given link, you have no business dropping that traffic.
RPF is a good idea, but it kind of breaks one of the more fundamental paradigms of the Internets.
BGP routes are filtered when they're exchanged between eBGP routers. If you don't filter, you're an idiot. This I agree with.
You're not talking about BGP route filtering, though; you're talking about some kind of reverse path filtering (making sure that a route to the source address exists on the interface that you received the packet from). In practice, you almost never do this on BGP routers, as RPF makes some (somewhat naive) assumptions about the symmetry of Internetwork traffic.
Most people who have some kind of RPF deployed have it on a customer-facing device, as you generally only have one route per customer (and can make assertions about traffic NOT coming via that route, AND traffic COMING via that route.)
That said, not an awful lot of people even have RPF deployed. Certainly nowhere near 100%. And it only takes one (or a handful of machines) that can forge UDP source addresses for this to be an issue.
looks like Slashdot users don't like reality since i am right about this but i am getting flame baited and trolled.
No, you're being corrected because you are FLAT OUT WRONG.
An EULA does not and can not apply unless it is presented to you BEFORE consideration is exchanged (in the case of purchasing a PSP, this would be at point of sale, BEFORE you gave them your cold hard and the title of said PSP transferred to you.)
Consider an analogy: If i sold lawn mowers, for instance, and you bought one off me with the (IMO perfectly reasonable) assumption that you could replace the blades with any compatible brand, and then went to replace them and found that i'd placed an EULA sticker across the nut holding each blade in place saying "ONLY USE MICHAEL HUNT BRANDED BLADES. THIS IS A LEGALLY BINDING CONTRACT," then you'd be pretty irate.
The fact that Dark Alex's firmware happens to allow playing of warez/backup games/imports/whatever is completely orthogonal to any argument you may or may not make RE: the legality of breaking an EULA. EULAs are not, and can not be retroactively binding. End of line.
Running 4 USB from the big enclosure to the 4 sockets in the server PC isn't much better, because it all goes through the same CPU and PCI bus.
Your 3 SATA cards will still be sharing a PCI bus (and a CPU) as well. Unless the box has PCIe and you get some 1x SATA cards, you're really not going to exceed 100MB/sec to RAM in that kind of setup, and while USB will add latency, it won't slow you down if you use a few individual controllers (since the hard limit of 133MB/sec still applies.)
if you're not in a position to get some 4-port PCIe 1x SATA controllers, 4 individual USB buses is probably the best bet (as long as you make sure you plug each set of disks into a port that doesn't share its bus with any other set of disks,) you'll be laughing.
Of course, if you ARE in a position to get some of the aforementioned 4-port 1x PCIe SATA cards with a linux supported chipset, that's what i'd be doing.
From bitter experience, a good 50% (at least) of the badly configured qmail installs i've come across have been installed as part of Plesk (http://www.swsoft.com), which is one of those hosting control-panel things that 2-bit hosting companies who have no business selling mail hosting use to sell their service to equally 2-bit idiots who have no business reading email.
djb has a lot to answer for. It'd be OK if he'd accept that he was wrong once in a while.
If Aunt Tillie sends me a message (forwarded from Betty, her next door neighbour, which was in turn forwarded from her nephew Boris, who goes to school in another city) which just happens to look like spam (who knows, maybe Boris is telling an amusing anecdote about how one of his friends stumbled across some h3rb4|_ v!agr4 or something,) I'm going to look like a fair dick if the message gets dropped on the floor and Aunt Tillie doesn't at least get notified that the message got eaten.
The 5xx range of status codes exists for this (and other) reasons, there's no reason NOT to use them (by performing content verification inline and either 2xx-ing or 5xx-ing the message between "." and "QUIT".)
As a 9-year veteran of the anti-spam industry (with experience within the regulator, although I've left that behind me now and work in telecoms,) it's a REAL stretch for anybody inside the IT industry to take these kinds of comments seriously.
Anybody who says that 'legitimate' mailservers are sending backscatter instead of 5xx-ing the message in transit is wrong. Mailservers which send backscatter are NOT legitimate, EOL.
...CoS has already been busted (albeit not overly publicly) for releasing a video compilation of death threats, hate mail, etc, which had said death threats in higher res than their supposed 'original' posting on youtube. Suss as....
(can't be screwed finding cites right now, worked for 26 hours straight, and now i'm plain out of it... little help?)
You're also (evidently, by your.sig) some kind of neoconservative. This goes some way to indicating that you hold a degree of unfashionable viewpoints. As such, your unfashionable views on cults are taken by most with a grain of salt. Myself included.
Your second mistake was referring to property belonging to the RTC corporation (variously, also Bridge Publications, IAS inc, &c.) as a 'church'.
Except that if you hand an original got-it-from-Ubuntu (or whatever) CD to a friend, you're not 'distributing' anything (in the context of copyright law,) otherwise used-CD and used-book shops, as well as libraries, would have been MPAAd to death years ago for copyright violation.
Were you to copy the Ubuntu CD for your friend, on the other hand, that would be distribution (and as such, legally actionable by the copyright holder in the absence of the GPL; remember, the GPL/adds/ to your freedom to distribute.) In this case, all that the GPL says is 'if you distribute the software, hand out the source also', as opposed to 'do not make illegal copies of this CD.'
Locking a handset to an IMSI prefix range (MCC/MNC) is supported by most GSM MSes these days so that operators can sell subsidised 'locked' phones. This obviously has nothing to do with the IMEI, which is (as others have said) an informational element only, and is not used for anything other than EIR (equipment ID register) lookups (to prove you didn't flog the phone.)
It's plausible that some carriers (although I honestly cannot see why) make a note of the IMEI when they sell you your phone, but even if the phone is locked, it's not guaranteed to still be you any time after you leave the shop (even locked phones can be used with other SIMs with the same MCC/MNC.
In short, the ONLY use of an IMEI is to verify that a phone hasn't been stolen, and to restrict service/alert law enforcement if it has. It can be used as a serial number, but only in the same way that any other serial number can.
Slashdot Mirror
You can all but bet that once two things have happened:
- Existing inventory of Q9x00 and Q6x00 parts is all but depleted
and
- Dunnington's desktop-equivalent part has debuted
that this will indeed be the case. I mean, the yields on such a die would be absolutely shiteful, especially since this is the first monolithic design that intel have fabbed with more than two physical cores on it. If you can disable 2 cores and sell it as a Q9990 XXXXTREME edition, why wouldn't you?
To my mind, the interesting question will be whether we (a-la AMD) see 5-core dunnington-alikes show up. It'll come down to whether intel figure that the extra effort involved in isolating and binning two different sorts of quasi-defective parts (versus doing it once) would be rewarded by having a 5-core part on the market (which would, natch, sell for more than the equivalent, but differently binned, 4 core part.)
No, all that it infers is that doubling the capacity of the drive is physically possible, not that the requisite head assembly would be affordable (or, for that matter, fast) for a desktop disk. The sort of setup you need to do this level of disk forensics would get you several tens of petabytes of cheap SATA disk.
> When I took my computer forensics class they showed that you could use a hex editor on a zero wiped floppy disk and recover most of the data that was on it previously. (emphasis mine)
If it was zero wiped, you'd see a meg and a bit of 0s. That said, if it's a floppy disk you're trying to wipe, the best way to go about it is a traditional low-level format, in which the drive mechanics physically remagnetise the surface of the media (this is the process which defines tracks and sectors.)
Not a lot's going to survive that.
Cider _is_ native code. All the way down. The only real difference (user interface issues aside, and most games don't use the native UI in any event) is that each DirectX call obviously goes through one more stack frame before it hits the hardware (game->cider directX->OpenGL->driver rather than game->directX->driver).
It's been my experience that the speed difference in Cedega (or lack thereof) from a 'native' DirectX implementation is marginal. With Cider, one would assume any differences would be more marginalized again given the fact that the code is in a native binary format and the vast bulk of the application isn't making calls through translated APIs.
It'll be interesting to see if anybody takes the freely available (albeit not open source) Cedega DirectX 9 code (or, for that matter, the open source but somewhat slower Wine DirectX 9 code) and adds the DirectX 10 APIs to it. A system like this would, in theory, allow DirectX 10 apps to run on Windows XP as well as other GL platforms.
You can't patent an instruction set, because an instruction set is an interface, not an implementation. You certainly can't copyright one for largely the same reason. There've been court rulings on this but i'm too lazy to look it up.
Of course, that's not saying that the 2 or 3 most efficient ways of implementing SSE3 etc aren't patented to the hilt, that might be the case. But the situation is nowhere near as dire as people are making out.
> How did anyone get "vastly different GPUs" from this?
Presumably because (for e.g.) a G70-based 7800 and a G92-based 8800GT are vastly different GPUs.
G70, for example, had two sets of fixed-purpose pipeline units (one of which ran your vertex programs, and one of which ran your fragment programs,) a bunch of fixed-function logic, some rasterisation logic, etc.
On the other hand, G80 and above have general purpose 'shader processors' any of which can run any pipeline programs (and, afaik, runs the traditional graphics fixed-function pipeline in software on said SPs), and a minimal amount of glue to hang it together.
About the only thing that current-generation GPUs and previous-generation GPUs have in common is the logo on the box (this applies equally to AMD parts, although the X19xx AMD cards, i'm told, are more similar to a G80-style architecture than a G70-style architecture, which is how the F@H folks managed to get it running on the X19xx before G80 was released.)
float4x4s for the most part, i'd be guessing
Yes, but the parent post was simply asserting that DX10 bought nothing to the table over DX9. I'm well aware that OpenGL 3.0, as well as Cg 2 (in conjunction with either GL 3.0 or some proprietary NV extension which I think is called gp40), are capable of geometry programs. My point was simply that DX 10 has geometry program support and DX9 doesn't.
Whether we like it or not, the world needs telemarketers
No it doesn't.
Next time a telemarketer calls during a nice dinner with your family, remember, you don't have to answer the phone, so if you do, you're the one interrupting your dinner, secondly, they most likely have a family that they wish they could be with, but instead they have to call you.
Next time a telemarketer calls during a nice dinner with my family and I answer the phone because it may ACTUALLY BE AN IMPORTANT CALL (newsflash, not everybody that you may want to talk to is capable or desirous of sending CLI), I am completely justified in flying totally off the fucking handle. There is no justification for a FOR-PROFIT BUSINESS calling my house after BUSINESS HOURS. My usual response to telemarketers is "please justify to me in 50 words or less why what you're doing is more ethical than sitting on the dole." The response to this is usually... interesting.
Two words: Geometry Program.
What MS call "Shader model 4" (even though geometry programs aren't, strictly speaking, shaders as they don't necessarily SHADE anything per se) includes mandatory support for geometry programs.
The geometry program sits in the programmable pipeline between the vertex program (which is used for real-time vertex deformation in hardware, world-space to object-space clipping to generate texcoords for the fragment program, etc) and the fragment program (which is used to colour fragments [1] based upon the output of the vertex program and input from one or more texture samplers.)
Unlike "old" vertex programs, a geometry program is able to generate new geometry on the fly. This allows a whole heap of really cool stuff, such as real-time shadowing effects, for essentially free.
So, yeah, much as I hate to admit it (and REALLY hate the Direct3D 'shader' nomenclature concerning pipeline programs,) D3D10 actually has changes with merit from D3D9c.
[1] A fragment is a fancy name for a voxel defined in clip space. After shading and occlusion, the remaining fragments become rasterised as pixels. Thus, the term 'pixel shader' is rather inaccurate.
My understanding is that it was more about assaying fears of a so-called 'plutonium economy', never mind the fact that attempting to get any sort of weapons-grade plutonium out of IFR fuel would almost certainly kill you before you managed to do so.
It was a piece of political expediency at a time when nuclear energy was being demonized for two main reasons (fear of loss-of-coolant and fear of long-lived actinides). The stupid thing, of course, is that the IFR and its ilk are the only nukes that SOLVE that problem, and people are out there now talking about building more PWRs and BWRs to 'fix global warming.'
I suppose as long as someone, somewhere decides to revisit the closed-fuel-cycle fast reactor concept, we don't have too many worries in terms of the waste (because SOMEONE'll use it for fuel long before its 20,000-odd years are up.) The concern, of course, is doing that before it all gets dumped under some mountain somewhere and forgotten about forever.
Love where you're going with your website. Didn't think there were too many other people out there irate with Clinton for killing off EBR2/Clinch River. You just made my friends list.
It's a hard call in some cases. There's an argument that most (any?) multihomed customers should be able to send packets asymmetrically. That said, nobody with even half a brain should be running a NAS/LNS with single-homed customers WITHOUT RPF, although the (admittedly, in .au) number of people i've crossed paths with who are doing it wrong is staggering.
Conversely, if i'm peering with you and six other guys, and happen to be carrying traffic to your network (but not advertising a route back to some of MY peers, for e.g. traffic engineering reasons) across a given link, you have no business dropping that traffic.
RPF is a good idea, but it kind of breaks one of the more fundamental paradigms of the Internets.
BGP routes are filtered when they're exchanged between eBGP routers. If you don't filter, you're an idiot. This I agree with.
You're not talking about BGP route filtering, though; you're talking about some kind of reverse path filtering (making sure that a route to the source address exists on the interface that you received the packet from). In practice, you almost never do this on BGP routers, as RPF makes some (somewhat naive) assumptions about the symmetry of Internetwork traffic.
Most people who have some kind of RPF deployed have it on a customer-facing device, as you generally only have one route per customer (and can make assertions about traffic NOT coming via that route, AND traffic COMING via that route.)
That said, not an awful lot of people even have RPF deployed. Certainly nowhere near 100%. And it only takes one (or a handful of machines) that can forge UDP source addresses for this to be an issue.
PS, you're an idiot.
looks like Slashdot users don't like reality since i am right about this but i am getting flame baited and trolled.
No, you're being corrected because you are FLAT OUT WRONG.
An EULA does not and can not apply unless it is presented to you BEFORE consideration is exchanged (in the case of purchasing a PSP, this would be at point of sale, BEFORE you gave them your cold hard and the title of said PSP transferred to you.)
Consider an analogy: If i sold lawn mowers, for instance, and you bought one off me with the (IMO perfectly reasonable) assumption that you could replace the blades with any compatible brand, and then went to replace them and found that i'd placed an EULA sticker across the nut holding each blade in place saying "ONLY USE MICHAEL HUNT BRANDED BLADES. THIS IS A LEGALLY BINDING CONTRACT," then you'd be pretty irate.
The fact that Dark Alex's firmware happens to allow playing of warez/backup games/imports/whatever is completely orthogonal to any argument you may or may not make RE: the legality of breaking an EULA. EULAs are not, and can not be retroactively binding. End of line.
Running 4 USB from the big enclosure to the 4 sockets in the server PC isn't much better, because it all goes through the same CPU and PCI bus.
Your 3 SATA cards will still be sharing a PCI bus (and a CPU) as well. Unless the box has PCIe and you get some 1x SATA cards, you're really not going to exceed 100MB/sec to RAM in that kind of setup, and while USB will add latency, it won't slow you down if you use a few individual controllers (since the hard limit of 133MB/sec still applies.)
if you're not in a position to get some 4-port PCIe 1x SATA controllers, 4 individual USB buses is probably the best bet (as long as you make sure you plug each set of disks into a port that doesn't share its bus with any other set of disks,) you'll be laughing.
Of course, if you ARE in a position to get some of the aforementioned 4-port 1x PCIe SATA cards with a linux supported chipset, that's what i'd be doing.
From bitter experience, a good 50% (at least) of the badly configured qmail installs i've come across have been installed as part of Plesk (http://www.swsoft.com), which is one of those hosting control-panel things that 2-bit hosting companies who have no business selling mail hosting use to sell their service to equally 2-bit idiots who have no business reading email.
djb has a lot to answer for. It'd be OK if he'd accept that he was wrong once in a while.
There's various regulators around the world that police various Anti-Spam laws.
If you _REALLY_ care, a quick check of my posting history would probably tell you which one.
PS: You're a fucking idiot.
If Aunt Tillie sends me a message (forwarded from Betty, her next door neighbour, which was in turn forwarded from her nephew Boris, who goes to school in another city) which just happens to look like spam (who knows, maybe Boris is telling an amusing anecdote about how one of his friends stumbled across some h3rb4|_ v!agr4 or something,) I'm going to look like a fair dick if the message gets dropped on the floor and Aunt Tillie doesn't at least get notified that the message got eaten.
The 5xx range of status codes exists for this (and other) reasons, there's no reason NOT to use them (by performing content verification inline and either 2xx-ing or 5xx-ing the message between "." and "QUIT".)
As a 9-year veteran of the anti-spam industry (with experience within the regulator, although I've left that behind me now and work in telecoms,) it's a REAL stretch for anybody inside the IT industry to take these kinds of comments seriously.
Anybody who says that 'legitimate' mailservers are sending backscatter instead of 5xx-ing the message in transit is wrong. Mailservers which send backscatter are NOT legitimate, EOL.
- A pissed off mail admin.
...CoS has already been busted (albeit not overly publicly) for releasing a video compilation of death threats, hate mail, etc, which had said death threats in higher res than their supposed 'original' posting on youtube. Suss as....
(can't be screwed finding cites right now, worked for 26 hours straight, and now i'm plain out of it... little help?)
You're also (evidently, by your .sig) some kind of neoconservative. This goes some way to indicating that you hold a degree of unfashionable viewpoints. As such, your unfashionable views on cults are taken by most with a grain of salt. Myself included.
Your second mistake was referring to property belonging to the RTC corporation (variously, also Bridge Publications, IAS inc, &c.) as a 'church'.
QED
Except that if you hand an original got-it-from-Ubuntu (or whatever) CD to a friend, you're not 'distributing' anything (in the context of copyright law,) otherwise used-CD and used-book shops, as well as libraries, would have been MPAAd to death years ago for copyright violation.
/adds/ to your freedom to distribute.) In this case, all that the GPL says is 'if you distribute the software, hand out the source also', as opposed to 'do not make illegal copies of this CD.'
Were you to copy the Ubuntu CD for your friend, on the other hand, that would be distribution (and as such, legally actionable by the copyright holder in the absence of the GPL; remember, the GPL
No, they don't.
Locking a handset to an IMSI prefix range (MCC/MNC) is supported by most GSM MSes these days so that operators can sell subsidised 'locked' phones. This obviously has nothing to do with the IMEI, which is (as others have said) an informational element only, and is not used for anything other than EIR (equipment ID register) lookups (to prove you didn't flog the phone.)
It's plausible that some carriers (although I honestly cannot see why) make a note of the IMEI when they sell you your phone, but even if the phone is locked, it's not guaranteed to still be you any time after you leave the shop (even locked phones can be used with other SIMs with the same MCC/MNC.
In short, the ONLY use of an IMEI is to verify that a phone hasn't been stolen, and to restrict service/alert law enforcement if it has. It can be used as a serial number, but only in the same way that any other serial number can.
You're a tool.