Motorola doesn't know what the hell it's doing. "We're going to use SymbianOS... no, PalmOS... no, WinCE... no, Linux." (I may have some of those out of order.) Their hardware isn't bad but they don't seem to have a clue when it comes to software, and it seems like they can't commit to anything. So far I think they have released one phone with each of SymbianOS and Linux.
There have been several murder trials in the UK where the location log (probably only cell ID) for the victim's phone was used as evidence, including this ongoing trial.
You could try suing them for fraud, since they wrongly claimed you needed a license. However, they are unlikely to have any money left with which to repay the license fees or any other damges.
There is good news on AIDS, though - the drugs to treat it (though not cure it) can now be made for as little as 40 cents per day, making them affordable in third-world countries.
Most drivers are bad, vendor-released or not. Only the popular Linux drivers are really well-written; some of the others are horrible. When I say popular, I mean popular among developers and the companies that pay them, not necessarily among the general user base. Examples of bad drivers:
ide-tape is a joke and ide-cd has major limitations. It turns out that you're better off mapping ATAPI to SCSI using ide-scsi and then using SCSI device class drivers (which is the approach NT takes too). Why do the ATAPI device class drivers exist at all?
I've found and fixed a SCSI driver that originally worked only on little-endian systems and was crudely hacked to work on big-endian systems... only it did terrible things with casts and would crash if compiled with optimisations on.
I used to use an Adaptec ISA SCSI card. These were pretty popular at the low end once. But the driver for these, aha152x, would crash the system in several releases in the 2.0.x series.
Barclaycard in the UK has been doing this to me. Unfortunately I failed to take enough details to be able to make a formal complaint. However, thanks to their zealous sales tactics they're about to lose a customer.
Debian stable is fine; however, I don't see how one can smoothly "upgrade" to that when it has older packages than Knoppix. If it's possible to do that, that's great - but I have never heard that before.
The problem I had in trying debian early on was that the install kernel was 2.2 only and did not support my hardware.
There are multiple builds of the installer and one of them uses a 2.4 kernel. If you're using an official CD you should be able to get that build by entering "bf24" at the multi-boot prompt (where the BIOS prints "boot:"). The installation manual has the details.
It is also a good plan to be up to date when it comes to plugging security holes in software.
Debian back-ports security fixes. You just need to include security.debian.org in your apt sources.
Just because it's hard to find an answer, doesn't mean it's hard to verify it. Consider the canonical example of factorisation - checking the results is trivial.
I think subliminal messages can reinforce or modify the effect of explicit messages given around the same time, but they cannot by themselves influence your behaviour over a longer period. So subliminal advertising inserted randomly into films or TV won't work (and is banned in some places, anyway), but subliminal images in an advertisement might make it more effective.
Ah, now I've found some better explanations and a comparison of Python generators and Ruby iterators. Regardless of their relative merits or precedence, I think C# is following Python not Ruby here, but I may have misunderstood the specification.
I'm not familiar with those other languages, and I didn't claim that Python invented generators because I really don't know. A quick look around gives me the impression that Ruby and Smalltalk don't provide quite the same kind of automatic wrapping that Python and C# are doing. Correct me if I'm wrong.
These "iterators" are actually what were implemented a couple of years ago in Python, where they are called "generators". Python uses more sensible (but less compatible) syntax: "yield" instead of "yield return" and "return" instead of "yield break". So they're an advantage C# will have over Java but they certainly aren't an MS innovation.
They're already using it for the WinCE.NET emulators. The earlier WinCE "emulators" just thunked system calls rather than running WinCE in a VM, so they were quite different in some ways.
It's only been forever since HTTP allowed you to host more than one site per IP,
Well, unless you want an encrypted connection. SSL setup happens before the server knows which name you used to access it, so it can't decide which certificate to present unless there's only one name per address.
There is an "input hook" to which one can attach a handler function in a DLL that will then get all input events. This is how Windows mouse support software makes the mouse-wheel work (somewhat) in old applications.
Surely someone can find out how to do the decryption by running Outlook 2003 under a debugger, then write my own client that does the same thing. Alternately someone could modify Outlook to remove the permission checks. Maybe there's anti-tampering code there, but this has been done so many times before in copy-protection and has generally been cracked before long. Until they can store the secret keys in hardware (which will probably happen soon) this isn't going to work.
I did that, but the fan I got from AMD is crap. I don't switch the computer off often but when I do the fan usually runs slow when I switch it on again and I have to spend 5 minutes fiddling with it until it decides to spin at full speed. (It isn't being throttled - the BIOS "system health" monitor shows me the CPU is getting hotter and hotter while the fan keeps running slow.)
Slashdot Mirror
Obligatory Sidetalkin' link.
Motorola doesn't know what the hell it's doing. "We're going to use SymbianOS... no, PalmOS... no, WinCE... no, Linux." (I may have some of those out of order.) Their hardware isn't bad but they don't seem to have a clue when it comes to software, and it seems like they can't commit to anything. So far I think they have released one phone with each of SymbianOS and Linux.
There have been several murder trials in the UK where the location log (probably only cell ID) for the victim's phone was used as evidence, including this ongoing trial.
You could try suing them for fraud, since they wrongly claimed you needed a license. However, they are unlikely to have any money left with which to repay the license fees or any other damges.
No, it's so uncool that it makes sidetalking look cool by comparison.
Karma: whore
Or not, as the case may be.
There is good news on AIDS, though - the drugs to treat it (though not cure it) can now be made for as little as 40 cents per day, making them affordable in third-world countries.
Most drivers are bad, vendor-released or not. Only the popular Linux drivers are really well-written; some of the others are horrible. When I say popular, I mean popular among developers and the companies that pay them, not necessarily among the general user base. Examples of bad drivers:
Barclaycard in the UK has been doing this to me. Unfortunately I failed to take enough details to be able to make a formal complaint. However, thanks to their zealous sales tactics they're about to lose a customer.
Debian stable is fine; however, I don't see how one can smoothly "upgrade" to that when it has older packages than Knoppix. If it's possible to do that, that's great - but I have never heard that before.
Because it's not like anyone cares about stability, security or support, right?
There are multiple builds of the installer and one of them uses a 2.4 kernel. If you're using an official CD you should be able to get that build by entering "bf24" at the multi-boot prompt (where the BIOS prints "boot:"). The installation manual has the details.
Debian back-ports security fixes. You just need to include security.debian.org in your apt sources.
Doesn't your first amendment refer to the rights of the people, not the rights of the government?
Just because it's hard to find an answer, doesn't mean it's hard to verify it. Consider the canonical example of factorisation - checking the results is trivial.
In adverts for "Girls Gone Wild 6".
I think subliminal messages can reinforce or modify the effect of explicit messages given around the same time, but they cannot by themselves influence your behaviour over a longer period. So subliminal advertising inserted randomly into films or TV won't work (and is banned in some places, anyway), but subliminal images in an advertisement might make it more effective.
Ah, now I've found some better explanations and a comparison of Python generators and Ruby iterators. Regardless of their relative merits or precedence, I think C# is following Python not Ruby here, but I may have misunderstood the specification.
I'm not familiar with those other languages, and I didn't claim that Python invented generators because I really don't know. A quick look around gives me the impression that Ruby and Smalltalk don't provide quite the same kind of automatic wrapping that Python and C# are doing. Correct me if I'm wrong.
These "iterators" are actually what were implemented a couple of years ago in Python, where they are called "generators". Python uses more sensible (but less compatible) syntax: "yield" instead of "yield return" and "return" instead of "yield break". So they're an advantage C# will have over Java but they certainly aren't an MS innovation.
They're already using it for the WinCE.NET emulators. The earlier WinCE "emulators" just thunked system calls rather than running WinCE in a VM, so they were quite different in some ways.
Well, unless you want an encrypted connection. SSL setup happens before the server knows which name you used to access it, so it can't decide which certificate to present unless there's only one name per address.
There is an "input hook" to which one can attach a handler function in a DLL that will then get all input events. This is how Windows mouse support software makes the mouse-wheel work (somewhat) in old applications.
Surely someone can find out how to do the decryption by running Outlook 2003 under a debugger, then write my own client that does the same thing. Alternately someone could modify Outlook to remove the permission checks. Maybe there's anti-tampering code there, but this has been done so many times before in copy-protection and has generally been cracked before long. Until they can store the secret keys in hardware (which will probably happen soon) this isn't going to work.
Electronic voting machines may not let you spoil your ballot.
I did that, but the fan I got from AMD is crap. I don't switch the computer off often but when I do the fan usually runs slow when I switch it on again and I have to spend 5 minutes fiddling with it until it decides to spin at full speed. (It isn't being throttled - the BIOS "system health" monitor shows me the CPU is getting hotter and hotter while the fan keeps running slow.)