We have this situation in the UK. You simply cannot have DSL without a phone line, and you can't compromise with a 'dial in only' one either.. it must be a proper one with all the rental cost etc. (which admittedly isn't large but pushes up the price of DSL).
This is true if you buy DSL from BT or an ISP that relies on BT. Local loop unbundling has happened in some places in the UK, and you can now buy just DSL from a competing telco. Also in some small areas there are phone networks entirely independent of BT. Cable is still bundled though I think NTL will sell you cable Internet service without phone service.
Having DSL also disqualfies you from all the 'low user' rebates even though you never make calls or use the analogue line at all.
To be fair, I suspect the light user scheme is a money-loser for BT. It it is legally required to provide it for the benefit of people that can't afford the full line rental. That surely doesn't include DSL users.
There isn't testing+security. If you're using testing, you have to wait for security fixes to be uploaded into unstable and then propagated into testing, which may be delayed indefinitely due to dependencies. That's one of the things that has put people off from using testing.
Copyright notices are neither here nor there, since copyright is automatic in countries signatory to the Berne Convention. Every search engine operator knowingly copies copyright material without explicit permission. There seems to be a consensus that publication on the web without charge gives implicit permission to copy in this way, or else that it is fair use. The REP provides a means to withhold this permission.
Why should Google cease what seems to be fair use of published material? A news agency could block this by requiring their customers to use the Robots Exclusion Protocol. Their customers probably won't want to do this, because they like the traffic. That's surely the agency's problem, not Google's.
It would be preferable for Windows to have real package management, but given that it doesn't the Firefox installer should be able to deal with an in-place upgrade itself. The fact that it doesn't is a bug in the installer, not in Windows. It's not that difficult to get right.
I thought the EU's complaint against Microsoft was that bundling Media Player with Windows gave them an unfair advantage in the media server market. It seems to me that their advantage remains if the codecs but not the front-end are pre-installed, since any other media player will presumably still be able to play WM* formats using those codecs.
French civil law is based on the Napoleonic Code, not common law. It's quite different.
Re:Codecs same in both: MS-DRM in your living room
on
Blu-Ray vs. HD-DVD
·
· Score: 1
...so is VC-1 (formerly known as VC-9 aka Windows Media Player) and that means MS-Windows DRM and royalties.
VC-1 is just the video codec from Windows Media 9. There's no DRM involved in it; that works at a different level. Of course future video distribution formats will have some kind of DRM, just as DVDs do.
ELF shared libraries have (a) and (b). Use "ldd" to see where libraries are loaded and you'll find they load at the same address almost every time. Note that Windows DLLs also have these features, but most developers never bother to change DLL load addresses from the default, so their programs load slowly as the DLLs collide and have to be relocated.
What interaction with shared libraries? Their code is shared just the same as the code of the executable initially loaded into a process (it is position-independent). However some sections of a shared library may need to be relocated or otherwise modified when loaded, and so those aren't shared. This seems to mean at least 1 unshareable page (normally 4K) for each such section for each process the library is loaded into - hence the concern about the large numbers of shared libraries loaded into apparently simple GNOME programs.
Code segments are simply mmapped to the executable files they come from; the pages are implicitly shareable with any other process that maps the same bit of the file read-only (or copy-on-write).
Libraries don't need stacks and heaps. Threads have stacks, and heaps are usually shared across a whole process. However, each library does normally contain a certain amount of static data and relocation information that can't be shared between processes, and that requires at least a page of VM.
Explain to me how the OS can handle it, in the general case. It seems to me that the best it can do is to restart the application. But if the application isn't designed to recover quickly when restarted then this doesn't help.
Bluetooth uses 2.4GHz frequencies, which according to a 1980's IEEE paper (I have a hardcopy around here sonewhere) is the PERFECT frequency to kill a lab rat, whilst leaving it's body intact.
You mean putting rats in the microwave oven will kill them? Who'da thunk it?
I suppose your favourite operating system is smart enough to handle divide-by-zero errors in applications by working out what the programmer really meant and fixing the bug. Regrettably Windows NT was written by mere humans who aren't capable of this.
I had the same question as you, and a friend attempted to explain the difference to me. Thinking about the twins paradox, I believe the distinction is that the travelling twin changes his frame of reference by accelerating (converting kinetic energy from some other source) whereas the stay-at-home twin stays in the same frame of reference. (Realistically the stay-at-home twin won't stay entirely static, but will use relatively little kinetic energy for sub-orbital travel.) I'm not sure that I understand this correctly though.
Safeway in the UK (which I think is now independent of the US chain) started trialling handheld scanners nearly 10 years ago, if I remember correctly. My local store was one of the first they tried it in.
The original Alchemy chips were developed by a separate company, Alchemy Semiconductors, which AMD bought 3 years ago; they're not entirely an AMD invention.
I didn't mean to say that that was the reason for the slowdown in development of IE. However, since the IE team was disbanded, the people who've been put to work on it now don't have much overlap with the earlier developers (AFAIK) and so won't be that familiar with the code base. It also seems as if the permissions system in IE is in an unmaintainable state, since there is a constant stream of new exploits found in it despite the security work done in XP SP2. Maybe the IE code base and developers aren't in quite such a bad state as the Netscape browser around version 4, but I think some parallels can be drawn.
Why did IE become the dominant browser? I think the writing was on the wall when IE 4 came out, being lighter, more stable and more standard-compliant (at least in the area of styles). I believe MS rewrote most of IE in version 3, whereas Netscape just carried on extending the browser code base well beyond the point where it was understandable to many of the people maintaining it. Now arguably MS is in the position that Netscape was then.
Slashdot Mirror
This is true if you buy DSL from BT or an ISP that relies on BT. Local loop unbundling has happened in some places in the UK, and you can now buy just DSL from a competing telco. Also in some small areas there are phone networks entirely independent of BT. Cable is still bundled though I think NTL will sell you cable Internet service without phone service.
To be fair, I suspect the light user scheme is a money-loser for BT. It it is legally required to provide it for the benefit of people that can't afford the full line rental. That surely doesn't include DSL users.
For the most they aren't taxes on phone service; they're just a cost of doing business that the telco doesn't want to include in its advertised rates.
Or perhaps lack of a coding standard.
Why shouldn't they? Try writing some unit tests.
There isn't testing+security. If you're using testing, you have to wait for security fixes to be uploaded into unstable and then propagated into testing, which may be delayed indefinitely due to dependencies. That's one of the things that has put people off from using testing.
AIUI Google isn't getting news from that web site.
Copyright notices are neither here nor there, since copyright is automatic in countries signatory to the Berne Convention. Every search engine operator knowingly copies copyright material without explicit permission. There seems to be a consensus that publication on the web without charge gives implicit permission to copy in this way, or else that it is fair use. The REP provides a means to withhold this permission.
Why should Google cease what seems to be fair use of published material? A news agency could block this by requiring their customers to use the Robots Exclusion Protocol. Their customers probably won't want to do this, because they like the traffic. That's surely the agency's problem, not Google's.
It would be preferable for Windows to have real package management, but given that it doesn't the Firefox installer should be able to deal with an in-place upgrade itself. The fact that it doesn't is a bug in the installer, not in Windows. It's not that difficult to get right.
I thought the EU's complaint against Microsoft was that bundling Media Player with Windows gave them an unfair advantage in the media server market. It seems to me that their advantage remains if the codecs but not the front-end are pre-installed, since any other media player will presumably still be able to play WM* formats using those codecs.
You can't copy a smart card, at least not without some extremely sophisticated hardware.
French civil law is based on the Napoleonic Code, not common law. It's quite different.
ELF shared libraries have (a) and (b). Use "ldd" to see where libraries are loaded and you'll find they load at the same address almost every time. Note that Windows DLLs also have these features, but most developers never bother to change DLL load addresses from the default, so their programs load slowly as the DLLs collide and have to be relocated.
What interaction with shared libraries? Their code is shared just the same as the code of the executable initially loaded into a process (it is position-independent). However some sections of a shared library may need to be relocated or otherwise modified when loaded, and so those aren't shared. This seems to mean at least 1 unshareable page (normally 4K) for each such section for each process the library is loaded into - hence the concern about the large numbers of shared libraries loaded into apparently simple GNOME programs.
Code segments are simply mmapped to the executable files they come from; the pages are implicitly shareable with any other process that maps the same bit of the file read-only (or copy-on-write).
Libraries don't need stacks and heaps. Threads have stacks, and heaps are usually shared across a whole process. However, each library does normally contain a certain amount of static data and relocation information that can't be shared between processes, and that requires at least a page of VM.
Explain to me how the OS can handle it, in the general case. It seems to me that the best it can do is to restart the application. But if the application isn't designed to recover quickly when restarted then this doesn't help.
I suppose your favourite operating system is smart enough to handle divide-by-zero errors in applications by working out what the programmer really meant and fixing the bug. Regrettably Windows NT was written by mere humans who aren't capable of this.
I had the same question as you, and a friend attempted to explain the difference to me. Thinking about the twins paradox, I believe the distinction is that the travelling twin changes his frame of reference by accelerating (converting kinetic energy from some other source) whereas the stay-at-home twin stays in the same frame of reference. (Realistically the stay-at-home twin won't stay entirely static, but will use relatively little kinetic energy for sub-orbital travel.) I'm not sure that I understand this correctly though.
RTFA. The BIOS is in flash, so you can easily replace it, unlike those hardwired components.
Safeway in the UK (which I think is now independent of the US chain) started trialling handheld scanners nearly 10 years ago, if I remember correctly. My local store was one of the first they tried it in.
The original Alchemy chips were developed by a separate company, Alchemy Semiconductors, which AMD bought 3 years ago; they're not entirely an AMD invention.
I didn't mean to say that that was the reason for the slowdown in development of IE. However, since the IE team was disbanded, the people who've been put to work on it now don't have much overlap with the earlier developers (AFAIK) and so won't be that familiar with the code base. It also seems as if the permissions system in IE is in an unmaintainable state, since there is a constant stream of new exploits found in it despite the security work done in XP SP2. Maybe the IE code base and developers aren't in quite such a bad state as the Netscape browser around version 4, but I think some parallels can be drawn.
Why did IE become the dominant browser? I think the writing was on the wall when IE 4 came out, being lighter, more stable and more standard-compliant (at least in the area of styles). I believe MS rewrote most of IE in version 3, whereas Netscape just carried on extending the browser code base well beyond the point where it was understandable to many of the people maintaining it. Now arguably MS is in the position that Netscape was then.