Perhaps you missed it, but Broadcom is selling them the silicon by tacking it on to larger production runs, so they've got as much as they want at quantity pricing.
I had missed that. I don't know much about the semiconductor industry so perhaps someone could confirm - does that mean that while they get them at quantity prices, that pricing is subject to Broadcom receiving sufficient orders for the same part so there are larger production runs to tack it onto?
FWIW every president should have an impeachment committee fired up and researching impeachable offenses starting the day they're elected. These various executives have far too much power to corrupt, delay and stop investigations that are the people's only defense from their crimes.
I wonder how many US presidents would even have lasted long enough to take their oath after election?
What Sony is trying to do would be non-binding in California.
It'd be non-binding in any sensible part of the world, I'd think. The whole point of many laws - particularly those designed to protect individuals - is to redress the balance when there's an otherwise uneven playing field; what's the point in having such laws if they can easily be signed away?
First, you need to give the CEO something they'll like the look of and wrap it up in an EULA that includes a clause to the effect that "licensee may be fed to saltwater crocodiles on national television". That way, not only do you feed them to the crocodiles, they've explicitly agreed to let you do so.
HAHAHA. POSIX states that writing zeros doesn't actually have to do anything but remember that it's supposed to return zeros for those blocks. Thus, you can store a file that's 1TB of zeros on a 100GB drive... Morons everywhere.
One would hope that anyone writing such a utility would have the good sense to fill the file up with something other than zeroes for precisely this reason.
Personally, I wouldn't stake my freedom on a gamble like that. You would be amazed how many applications are written with so little knowledge of the operating system's core API...
So Manning certainly knew about this kind of thing, but either didn't do it or didn't do it correctly. I wonder how difficult it is to mess something like that up?
Extremely easy.
Any modern operating system uses swap space - and while there's usually a way to ask the OS never to swap a program out, it's seldom exposed to the user. It normally relies on the program itself requesting this, and not everything will. Though a program may be exited later, the area of swapfile it used to use is not necessarily freed from disk.
On top of that, a few programs (eg. Gimp) deal with their own memory management to a certain extent and so operate their own swap independently of the OS - they may also keep other temporary files floating around and don't always delete them. Or they may not save a file in the way you expect - when you hit "save", it's not unusual for a program to:
- Create a new file.
- Dump the data into the new file.
- Rename the old file.
- Rename the new file so it has the same name as the old one.
- Delete the old file.
This drastically reduces the risk of the app dying part way through the save process resulting in a corrupted file. It may result in a file that hasn't been saved, so some work may be lost, but it won't lose the lot. Of course this has the side-effect that there's an old file sat on the disk somewhere containing much the same data.
On top of that, very small files will be stored directly in the MFT on Windows. Now the size of file we're talking about is probably not big enough to contain any serious information, but it may well give a forensic investigator a clue as to what's been done.
I can think of a few scenarios in which Manning could easily mess up:
1. Several "secure delete" utilities offer the option to securely delete individual files. Which they will, but as discussed above that may not achieve much. 2. Using a tool to wipe all free space - these usually work by creating a file and filling it with zeroes until the OS eventually returns a disk full error, then deleting the file. I have no idea what - if anything - they'll do with any data still sitting around the MFT. Not to mention the fact that they won't help if there's any incriminating files sitting around that weren't deleted in the first place - and as we've established, it's quite possible for an application to do this totally invisibly to the end user.
Realistically Manning would need to run DBAN or something similar on the entire disk. This will wipe the OS, so the affected computer would need to be reimaged.
I just checked the 10.7 laptop I'm typing on. Checking for and downloading the updates is automatic, actually installing them is something you get prompted to do and there is no tick box to say "Don't bother me, just do it".
We already have Documentum which is supposed to be able to use Firefox and the like but thanks to Mozilla's insistence on their INSANE version escalation practices, every update is an X.0 update meaning Documentum thinks it can't support it.
Would it be considered acceptable to install a reverse proxy that rewrites the User-Agent?
You'd be surprised. I know of at least one industry (which I won't name here) where franchisees have to connect to a website that's private and exclusive to them.
A boot release in the passenger compartment certainly used to be common on saloon cars in Europe. Not sure about anything sold in the last ten years though.
IANAL but AIUI you can't just file a complaint at court and expect a judge to join the dots with existing law. You have to explicitly say "Sony ripped me off contrary to law A". If the suit was started with just "Sony ripped me off!" (which, AFAICT, is exactly what happened), that's not good enough.
If you rent office space, you pay by the square foot typically. Or some managed offices (where the landlord does more-or-less everything - internet, power, telephones, desks, chairs - all you do is show up and run your business) charge you by the desk.
I think I figured it out myself, but a certain amount of this is guesswork based on observation. Take it with as much salt as you think it needs - if anyone can add to this, please do so.
Sharepoint doesn't give you a lot out of the box, as you've mentioned. What it does give you is a very capable framework with which to put together your own solution.
What I think happens is this:
Person A (who is not in IT - perhaps he's a salesman) works for company B that implements a Sharepoint-based system. Company B is quite joined up - they'll get all the major stakeholders on board when there's a big project. IT will advise on sorting out the underlying infrastructure and making sure it all works; other business units will get involved to make sure the project achieves what it's supposed to. They take their Sharepoint project seriously - they allocate a budget to it, they put together a clear set of requirements based around their business processes (which were already pretty clear) and they hire in or contract out development work. Ultimately they wind up with an extremely competent system that does all they could ever want with pretty good efficiency - and the project is delivered in record time. Yes it's based on Sharepoint but they've really harnessed it to make it work for them.
Person A doesn't know anything about how the project was implemented, however. All he knows is that his employer brought in Sharepoint and it's fantastic.
Anyhow, time marches on and Person A gets offered a new job as a sales director at company C. It's a nice little promotion, so he takes it. What's the first thing he sees? Company C doesn't have anything like the sophisticated computer system he used at company B. So he demands Sharepoint.
Note that person A has a lot of influence in his new job. He's a sales director in charge of a small team who essentially provide the money the company needs to survive. He demands Sharepoint, he gets it. The problem is, nobody at company C - not even person A - actually knows what Sharepoint is. Oh, they know it's some sort of computer-type-system-thingy, but that's as far as they go. Expecting it to be a big project, someone is tasked to look into it and figure out how much it's going to cost..... oh. That's interesting. Apparently, Sharepoint is free. And not free as in "free from some dodgy website", it's a free download from Microsoft themselves.
So instead of putting a formal project together, the IT department simply downloads Sharepoint Foundation, installs it, sends everyone a URL and thinks to themselves "That was easy".
It soon becomes apparent that Sharepoint isn't the all-singing, all-dancing system our blue-eyed boy proclaimed it to be. Sure, it's a competent enough little Intranet and it's a lot less hassle than a traditional fileserver to store your documents on, but the way it was being discussed initially, anyone would think the second coming of the Messiah was in the form of a computer program. It sure isn't that. But nobody quite knows why.
Glad to see I'm not losing my mind. Pretty sure it was that model I encountered - and in this case it was at a client where attempting to control what gets plugged into the port was such a complete non-starter it wasn't worth even broaching the topic. Disable STP, quietly curse whoever QA'd the firmware for that model then get on with life.
We need a Wikipedia for crap hardware that doesn't properly do what it claims to. The only issue is I can see Dell taking umbrage at being inducted into the Hall of Shame before we've even started.
If you mean that the managed switch dies when you connect an unmanaged switch with NO loop: then you have an extremely crappy managed switch. This use case has nothing to do with STP.
That's exactly what I mean; disable STP and it all starts to magically work.
This was a Dell switch, which probably explains rather a lot - rumour has it that particular model is a rebadged Allied Telesyn. Mind you, if Dell were to write to me informing me the sky was blue I'd stick my head out of the window.
I really think that mobile companies should be among the first on IPv6 with IPv4 access via NAT & proxy.
AFAICT the majority of mobile companies - at least in the UK - already are. Plug a USB dongle into your laptop or check the IP address on your phone, there's a good chance it's in RFC1918 address space and they're NAT'ing you.
I think that if something is declared "supported", it is perfectly reasonable to expect it to work. If it turns out it doesn't work, I think the problem is more that the vendor hasn't done as good a job as they should have than that your expectations were too high.
Indeed, but it's the same with all commodity technology - you find various implementations, not all of which work properly.
The same was true 10 or 15 years ago with booting from CD. Same was true 5-6 years ago with PXE. Same's true with CIDR - I've come across equipment like printers that can't handle the idea - you have to give them a class A, B or C subnet mask. Same with STP (spanning tree) - I've met switches that just plain don't work if you turn on STP then plug in a cheapie unmanaged switch - and I don't mean the port plugged into the cheapie switch doesn't work, I mean the entire expensive managed switch doesn't work. Only a couple of weeks ago I met a server BIOS providing software RAID (yeuch) that needed the drives set to RAID in the BIOS for it to work. But if power to the server was lost, that specific BIOS setting would go. Every other BIOS setting would be just fine and you'd get no error at bootup; you'd just find your disks magically appeared differently on boot.
If Google's network team honestly thought that any product with "IPv6 supported" on the label meant "Every aspect of IPv6 fully supported, tested, interoperable with other vendor's implementation - basically it'll work as well as you'd expect IPv4 to work in something released in the last five years", they're displaying incredible naiveté.
Let's take a step back, stop foaming at the mouth and try to look at this dispassionately. Please note before reading that IANAL.
It's already been documented that Jobs was Not Happy with the direction Android was heading in, and wanted nothing less than to kill the platform altogether, to hell with any collateral damage this might do the industry as a whole. I think it's vanishingly unlikely that Jobs surrounded himself with people who would take a live and let-live attitude, so the fact that he's passed on is likely neither here nor there.
Now, it's unthinkable that Apple might have failed to notice Microsoft's ability to extort royalty payments out of Android handset vendors. So it looks like they're trying to do the same thing. This makes a lot of sense for two reasons:
1. Anyone looking to base a new product on Android will now think twice - is "free" such a bargain if you've got to license so many patents from so many places? Might be cheaper to build your own or license a commercial product and get some sort of patent indemnification at the same time. 2. Anyone who's already got a bunch of Android products on the market will be paying Apple.
But there's a problem. The cellphone industry is absolutely swimming in patents - some of which must be licensed on reasonable, non-discriminatory terms, some of which have no such restriction, some of which are in enough of a grey area that the holder could easily eat up a lot of lawyer time. Apple can certainly expect counter-claims if they sue every Android handset maker on the planet - and regardless of the outcome, each will mean years thrashing out a solution. Years of injunctions to prevent sale of the latest product, appeals against those injunctions, the works.
Solution? Well, if a company that doesn't make anything does the suing - a patent troll, if you like - there's much less scope for retaliatory counter-claims. Set up such a company, transfer the appropriate patents to it and let that company do the suing. An elegant solution engineered explicitly to game the legal system. But I'm not sure how the legal system will react to such a gaming - I can see it attracting antitrust action.
I've been thinking this for some time, and the more I see articles like this (and for that matter Android phones), the more convinced I am that a rapidly evolving platform put together by a company famed for an attitude of "foist a beta onto users, we can always update it later" and the cellphone industry as it stands simply do not mix.
Why don't they mix? Well, think back to a dumb phone - more specifically back to when that was more-or-less your only option.
Every manufacturer had at least 4-6 models available to the general public at any point in time and each model was generally quite different from any others from the same manufacturer available at the same time; each model would be sold for a relatively high price on contract for maybe 9-12 months before being offered at a knockdown price on pay-as-you-go for maybe another 9-12 months before being discontinued. A few were developed down to a price to begin with so they'd only ever be available on pay-as-you-go. Firmware updates were applied by phone stores and were only ever released to correct really bad bugs. 99% of the manufacturer's work on the firmware was completed before the phone was released to the general public; the development team would be on to other things by the time the phone was actually released and wouldn't be taken off to work on bugs in existing firmware unless really necessary. Brand loyalty was virtually non-existent, so there was little point in worrying too much about providing a stellar user experience as long as you didn't actively annoy the user.
AFAICT, most phone manufacturers haven't really changed. Samsung do this, so do Sony Ericsson, so do HTC. But when your business model is based around building phones in this fashion, you can't really make drastic firmware upgrades available for existing models. You'd need to significantly increase the development team simply to get Android working on phones that are already shipping. To what benefit? By the time you've done so on a shipping model, it's just about to be discontinued, its replacement is around the corner and such work will get a return on investment of approximately zero.
Slashdot Mirror
Perhaps you missed it, but Broadcom is selling them the silicon by tacking it on to larger production runs, so they've got as much as they want at quantity pricing.
I had missed that. I don't know much about the semiconductor industry so perhaps someone could confirm - does that mean that while they get them at quantity prices, that pricing is subject to Broadcom receiving sufficient orders for the same part so there are larger production runs to tack it onto?
FWIW every president should have an impeachment committee fired up and researching impeachable offenses starting the day they're elected. These various executives have far too much power to corrupt, delay and stop investigations that are the people's only defense from their crimes.
I wonder how many US presidents would even have lasted long enough to take their oath after election?
What Sony is trying to do would be non-binding in California.
It'd be non-binding in any sensible part of the world, I'd think. The whole point of many laws - particularly those designed to protect individuals - is to redress the balance when there's an otherwise uneven playing field; what's the point in having such laws if they can easily be signed away?
You're not going far enough.
First, you need to give the CEO something they'll like the look of and wrap it up in an EULA that includes a clause to the effect that "licensee may be fed to saltwater crocodiles on national television". That way, not only do you feed them to the crocodiles, they've explicitly agreed to let you do so.
HAHAHA. POSIX states that writing zeros doesn't actually have to do anything but remember that it's supposed to return zeros for those blocks. Thus, you can store a file that's 1TB of zeros on a 100GB drive... Morons everywhere.
One would hope that anyone writing such a utility would have the good sense to fill the file up with something other than zeroes for precisely this reason.
Personally, I wouldn't stake my freedom on a gamble like that. You would be amazed how many applications are written with so little knowledge of the operating system's core API...
So Manning certainly knew about this kind of thing, but either didn't do it or didn't do it correctly. I wonder how difficult it is to mess something like that up?
Extremely easy.
Any modern operating system uses swap space - and while there's usually a way to ask the OS never to swap a program out, it's seldom exposed to the user. It normally relies on the program itself requesting this, and not everything will. Though a program may be exited later, the area of swapfile it used to use is not necessarily freed from disk.
On top of that, a few programs (eg. Gimp) deal with their own memory management to a certain extent and so operate their own swap independently of the OS - they may also keep other temporary files floating around and don't always delete them. Or they may not save a file in the way you expect - when you hit "save", it's not unusual for a program to:
- Create a new file.
- Dump the data into the new file.
- Rename the old file.
- Rename the new file so it has the same name as the old one.
- Delete the old file.
This drastically reduces the risk of the app dying part way through the save process resulting in a corrupted file. It may result in a file that hasn't been saved, so some work may be lost, but it won't lose the lot. Of course this has the side-effect that there's an old file sat on the disk somewhere containing much the same data.
On top of that, very small files will be stored directly in the MFT on Windows. Now the size of file we're talking about is probably not big enough to contain any serious information, but it may well give a forensic investigator a clue as to what's been done.
I can think of a few scenarios in which Manning could easily mess up:
1. Several "secure delete" utilities offer the option to securely delete individual files. Which they will, but as discussed above that may not achieve much.
2. Using a tool to wipe all free space - these usually work by creating a file and filling it with zeroes until the OS eventually returns a disk full error, then deleting the file. I have no idea what - if anything - they'll do with any data still sitting around the MFT. Not to mention the fact that they won't help if there's any incriminating files sitting around that weren't deleted in the first place - and as we've established, it's quite possible for an application to do this totally invisibly to the end user.
Realistically Manning would need to run DBAN or something similar on the entire disk. This will wipe the OS, so the affected computer would need to be reimaged.
I just checked the 10.7 laptop I'm typing on. Checking for and downloading the updates is automatic, actually installing them is something you get prompted to do and there is no tick box to say "Don't bother me, just do it".
We already have Documentum which is supposed to be able to use Firefox and the like but thanks to Mozilla's insistence on their INSANE version escalation practices, every update is an X.0 update meaning Documentum thinks it can't support it.
Would it be considered acceptable to install a reverse proxy that rewrites the User-Agent?
You'd be surprised. I know of at least one industry (which I won't name here) where franchisees have to connect to a website that's private and exclusive to them.
Guess what? IE only. Even today.
If they've got any sense at all they'll be using WSUS - can Microsoft override this? I wouldn't think so.
Seriously, all those great ideas about hiding places you've had? Like under the driver's seat or in the spare wheel well?
You're not the first person to have those great ideas. In fact, you're probably about the 4,000,000th.
Liquid bleach will probably do exactly that, it's quite nasty to most rubbers over a period of time.
A boot release in the passenger compartment certainly used to be common on saloon cars in Europe. Not sure about anything sold in the last ten years though.
IANAL but AIUI you can't just file a complaint at court and expect a judge to join the dots with existing law. You have to explicitly say "Sony ripped me off contrary to law A". If the suit was started with just "Sony ripped me off!" (which, AFAICT, is exactly what happened), that's not good enough.
Not entirely true. Apparently the whole of Russia is so delighted with their government they have, to a man, voted them all back in again.
If you rent office space, you pay by the square foot typically. Or some managed offices (where the landlord does more-or-less everything - internet, power, telephones, desks, chairs - all you do is show up and run your business) charge you by the desk.
In any event, that sounds about right.
I think I figured it out myself, but a certain amount of this is guesswork based on observation. Take it with as much salt as you think it needs - if anyone can add to this, please do so.
Sharepoint doesn't give you a lot out of the box, as you've mentioned. What it does give you is a very capable framework with which to put together your own solution.
What I think happens is this:
Person A (who is not in IT - perhaps he's a salesman) works for company B that implements a Sharepoint-based system. Company B is quite joined up - they'll get all the major stakeholders on board when there's a big project. IT will advise on sorting out the underlying infrastructure and making sure it all works; other business units will get involved to make sure the project achieves what it's supposed to. They take their Sharepoint project seriously - they allocate a budget to it, they put together a clear set of requirements based around their business processes (which were already pretty clear) and they hire in or contract out development work. Ultimately they wind up with an extremely competent system that does all they could ever want with pretty good efficiency - and the project is delivered in record time. Yes it's based on Sharepoint but they've really harnessed it to make it work for them.
Person A doesn't know anything about how the project was implemented, however. All he knows is that his employer brought in Sharepoint and it's fantastic.
Anyhow, time marches on and Person A gets offered a new job as a sales director at company C. It's a nice little promotion, so he takes it. What's the first thing he sees? Company C doesn't have anything like the sophisticated computer system he used at company B. So he demands Sharepoint.
Note that person A has a lot of influence in his new job. He's a sales director in charge of a small team who essentially provide the money the company needs to survive. He demands Sharepoint, he gets it. The problem is, nobody at company C - not even person A - actually knows what Sharepoint is. Oh, they know it's some sort of computer-type-system-thingy, but that's as far as they go. Expecting it to be a big project, someone is tasked to look into it and figure out how much it's going to cost..... oh. That's interesting. Apparently, Sharepoint is free. And not free as in "free from some dodgy website", it's a free download from Microsoft themselves.
So instead of putting a formal project together, the IT department simply downloads Sharepoint Foundation, installs it, sends everyone a URL and thinks to themselves "That was easy".
It soon becomes apparent that Sharepoint isn't the all-singing, all-dancing system our blue-eyed boy proclaimed it to be. Sure, it's a competent enough little Intranet and it's a lot less hassle than a traditional fileserver to store your documents on, but the way it was being discussed initially, anyone would think the second coming of the Messiah was in the form of a computer program. It sure isn't that. But nobody quite knows why.
Glad to see I'm not losing my mind. Pretty sure it was that model I encountered - and in this case it was at a client where attempting to control what gets plugged into the port was such a complete non-starter it wasn't worth even broaching the topic. Disable STP, quietly curse whoever QA'd the firmware for that model then get on with life.
We need a Wikipedia for crap hardware that doesn't properly do what it claims to. The only issue is I can see Dell taking umbrage at being inducted into the Hall of Shame before we've even started.
You can make money, you just need large volumes of business on a smaller margin.
This is entirely true, but it's damn difficult to service large volumes of business when you're just starting up.
Said I'd seen. Didn't say I'd bought. Most of that was inherited equipment.
If you mean that the managed switch dies when you connect an unmanaged switch with NO loop: then you have an extremely crappy managed switch. This use case has nothing to do with STP.
That's exactly what I mean; disable STP and it all starts to magically work.
This was a Dell switch, which probably explains rather a lot - rumour has it that particular model is a rebadged Allied Telesyn. Mind you, if Dell were to write to me informing me the sky was blue I'd stick my head out of the window.
I really think that mobile companies should be among the first on IPv6 with IPv4 access via NAT & proxy.
AFAICT the majority of mobile companies - at least in the UK - already are. Plug a USB dongle into your laptop or check the IP address on your phone, there's a good chance it's in RFC1918 address space and they're NAT'ing you.
I think that if something is declared "supported", it is perfectly reasonable to expect it to work. If it turns out it doesn't work, I think the problem is more that the vendor hasn't done as good a job as they should have than that your expectations were too high.
Indeed, but it's the same with all commodity technology - you find various implementations, not all of which work properly.
The same was true 10 or 15 years ago with booting from CD. Same was true 5-6 years ago with PXE. Same's true with CIDR - I've come across equipment like printers that can't handle the idea - you have to give them a class A, B or C subnet mask. Same with STP (spanning tree) - I've met switches that just plain don't work if you turn on STP then plug in a cheapie unmanaged switch - and I don't mean the port plugged into the cheapie switch doesn't work, I mean the entire expensive managed switch doesn't work. Only a couple of weeks ago I met a server BIOS providing software RAID (yeuch) that needed the drives set to RAID in the BIOS for it to work. But if power to the server was lost, that specific BIOS setting would go. Every other BIOS setting would be just fine and you'd get no error at bootup; you'd just find your disks magically appeared differently on boot.
If Google's network team honestly thought that any product with "IPv6 supported" on the label meant "Every aspect of IPv6 fully supported, tested, interoperable with other vendor's implementation - basically it'll work as well as you'd expect IPv4 to work in something released in the last five years", they're displaying incredible naiveté.
Let's take a step back, stop foaming at the mouth and try to look at this dispassionately. Please note before reading that IANAL.
It's already been documented that Jobs was Not Happy with the direction Android was heading in, and wanted nothing less than to kill the platform altogether, to hell with any collateral damage this might do the industry as a whole. I think it's vanishingly unlikely that Jobs surrounded himself with people who would take a live and let-live attitude, so the fact that he's passed on is likely neither here nor there.
Now, it's unthinkable that Apple might have failed to notice Microsoft's ability to extort royalty payments out of Android handset vendors. So it looks like they're trying to do the same thing. This makes a lot of sense for two reasons:
1. Anyone looking to base a new product on Android will now think twice - is "free" such a bargain if you've got to license so many patents from so many places? Might be cheaper to build your own or license a commercial product and get some sort of patent indemnification at the same time.
2. Anyone who's already got a bunch of Android products on the market will be paying Apple.
But there's a problem. The cellphone industry is absolutely swimming in patents - some of which must be licensed on reasonable, non-discriminatory terms, some of which have no such restriction, some of which are in enough of a grey area that the holder could easily eat up a lot of lawyer time. Apple can certainly expect counter-claims if they sue every Android handset maker on the planet - and regardless of the outcome, each will mean years thrashing out a solution. Years of injunctions to prevent sale of the latest product, appeals against those injunctions, the works.
Solution? Well, if a company that doesn't make anything does the suing - a patent troll, if you like - there's much less scope for retaliatory counter-claims. Set up such a company, transfer the appropriate patents to it and let that company do the suing. An elegant solution engineered explicitly to game the legal system. But I'm not sure how the legal system will react to such a gaming - I can see it attracting antitrust action.
I've been thinking this for some time, and the more I see articles like this (and for that matter Android phones), the more convinced I am that a rapidly evolving platform put together by a company famed for an attitude of "foist a beta onto users, we can always update it later" and the cellphone industry as it stands simply do not mix.
Why don't they mix? Well, think back to a dumb phone - more specifically back to when that was more-or-less your only option.
Every manufacturer had at least 4-6 models available to the general public at any point in time and each model was generally quite different from any others from the same manufacturer available at the same time; each model would be sold for a relatively high price on contract for maybe 9-12 months before being offered at a knockdown price on pay-as-you-go for maybe another 9-12 months before being discontinued. A few were developed down to a price to begin with so they'd only ever be available on pay-as-you-go. Firmware updates were applied by phone stores and were only ever released to correct really bad bugs. 99% of the manufacturer's work on the firmware was completed before the phone was released to the general public; the development team would be on to other things by the time the phone was actually released and wouldn't be taken off to work on bugs in existing firmware unless really necessary. Brand loyalty was virtually non-existent, so there was little point in worrying too much about providing a stellar user experience as long as you didn't actively annoy the user.
AFAICT, most phone manufacturers haven't really changed. Samsung do this, so do Sony Ericsson, so do HTC. But when your business model is based around building phones in this fashion, you can't really make drastic firmware upgrades available for existing models. You'd need to significantly increase the development team simply to get Android working on phones that are already shipping. To what benefit? By the time you've done so on a shipping model, it's just about to be discontinued, its replacement is around the corner and such work will get a return on investment of approximately zero.