We do this already with apples and potatoes, IIRC. Why not meat?
Well, the theory is that meats contain chemicals (triglycerides, I think) that, when irradiated, sometimes become allegedly-toxic chemicals known as cyclobutanones. See Wikipedia. I don't know if that theory is scientific, or if it's just rubbish put forward retroactively by misguided activists.
Ultimately, it's a matter of risk management. I am in favour of the option that carries the lowest risk of killing me.
The first thing to do would be to talk to your lawyer and find out if the police actually can do what you want in your jurisdiction. If so, file a complaint with the police department's Internal Affairs office. See what happens with that first. It's possible that some individual cops are just being lazy, and that might be all it takes to motivate them.
It's not just water that you have to worry about. Commercial-spec solid-state electronic components are typically rated for operation between 0 and 70 degrees Celsius. Electronic components conduct electricity better (or worse, in the case of many semiconductors) at lower temperatures, so even in a humidity-controlled environment, you could end up melting certain components.
What you need is either computers that are built entirely out of industrial or automotive-spec components that are rated at -40 to 85 degrees Celsius, or you need a temperature-controlled server room that will keep the computers within the commercial-spec range. Both are going to cost money.
Combined with matching optics, could one not use one of those polymer displays to create multiple wavelength signals and send them through one fiber, in theory allowing an indefinite number of signals?
This is more-or-less already being done, although not with this technology. (But new ways of doing it might prove to be more cost-effective.) See Wavelength division multiplexing.
Well, that right there eliminates 99% of commercial software. Why buy Excel when there's Gnumeric? Why buy Word when there's OpenOffice? Why buy MS Visio when there's probably kVisio out there somewhere? It doesn't matter if Excel, Word and Visio are *better* than the open source alternatives... you don't consider that. But because the open source alternatives exist at all, you'll never buy them. Additionally, since it takes only weeks for some open source project to rip-off the next big commercial software product, you'll *never* buy the commercial product since the open source alternative will always exist.
It shouldn't have too, but the FOSS comunity (or at least the vocal ones) would demand it, and anything less would be derided as half assed.
Perhaps, but it's not like binary-only software would be more successful in that community. Look at qmail: It's proprietary, source-available software. How many people would use it if DJB only released binaries for the platforms that he happened to be using when he wrote it?
Well, ultimately what matters to the consumer is what the consumer perceives. If 10% of Mac users think they have spyware, then there's 10% diss-satisfied customers. It doesn't matter if none of them actually do.
Yes, but if spyware isn't the real problem, then anti-spyware software isn't going to help.
Realistically, though, if you took reasonable measures to ensure that the virus wouldn't get out, then I doubt the courts would hold you liable. They might hold the disgruntled employee(s) liable, though.
as long as they don't directly lift material from CR, they can still feel free to say "Consumer Reports' Number One Pick for Antivirus Software"
I think trademark law might prevent that, since it would be an endorsement, but they might be able to cite a particular Consumer Reports article. I'd definitely get my lawyers' advice on something like that, though.
(especially to create scares of vulnerability on OSX, Linux, BSD, etc... where no real vulnerability exists).
The vulnerabilities do exist; they're just not being exploited nearly as much. Of course, run-of-the-mill signature-based antivirus software is equally flawed, as Consumer Reports has shown and security geeks have already known.
As he already said, he simply wanted to try Linux out, so why should he have to spend money on different hardware to do so?
Mac OS X has terrible hardware support: it only runs on Macintosh computers. People know and accept that, so it's a non-issue.
The problem here is that Linux's hardware support is so good (when compared to e.g. Mac OS X) that people have come to expect that it will support every piece of hardware is existance, but that will never happen until hardware manufacturers start using standardized interfaces for everything (and even then, there will still be people using "obsolete" hardware).
Despite what the zealots claim, Linux is not for everyone. No software is. If you want to run Linux, then you can expect that its hardware requirements will differ from those of your particular version of Windows. However, the impact of that is, in my experience, overstated, since once you've decided to run Linux, your future hardware purchases will be made with Linux in mind---and it really doesn't matter if Linux doesn't support hardware that you don't have anyway.
As he already said, he simply wanted to try Linux out, so why should he have to spend money on different hardware to do so? What's so special about linux from an and-user POV that they should expend more time, effort, and money on it than any other piece of free trial software?
He shouldn't have to, but reality isn't always what it should be. Besides, missing features in trial software isn't all that unheard of. Think of it as a free trial with wireless networking disabled. If you use a large distribution like Debian, you have 14 CD-ROMs of software that you can install without ever going to the network. When you're satisfied that Debian (or whatever) is worth the investment, you can pay $100 to enable wireless networking, and you get free upgrades forever and there is never any "GNU Product Activation" nonsense to put up with.
My first Linux machine had a Winmodem. I installed Slackware 3.5 (which I downloaded using that Winmodem) onto a UMSDOS filesystem. When I decided that I eventually wanted to replace Windows with Linux, I saved up and bought a real modem. Sure, it would have been nice if Linux would have run perfectly out-of-the-box on the machine, but support for the Winmodem really wasn't all that important in the long run, and I'd much rather see resources put into developing better software than into fighting with the last few uncooperative hardware manufacturers.
Suppose I send out something with a ttl of 5, and along the way it goes through a few not-so well configured routers and gets bounced off of goe-sync sats a dozen times. By you're reasoning my process should never have sent the packet in the first place because it should have known it'd take to long even though no mechanism exists for it to know.
No, that would be ridiculous. What I'm saying is that those "not-so well configured routers" are each supposed to decrement the TTL by some value greater than the actual number of seconds a transmission will take. It's fairly easy to do on a hob-by-hob basis; Anybody who doesn't know that their router is directly connected to a satellite uplink shouldn't be allowed near an Internet router anyway. It complicates the design of routers somewhat, but IPv4 has always assumed moderately complex routers, at least when compared with IPv6.
It really only makes sense that transit time is NOT part of the protocal (as not really thought about when writing it, thus the confusion the use of seconds often causes). Note that it never mentions signal time but refers to processing time. TTL is max # hops/operations on packet, or total time being actually handled or held by a process as measured in seconds(if known) with max value off 255.
That's probably a sensible way to look at it, but it's not what the spec says. Your interpretation limits the TTL to just "processing time" or a "hop count", which contradicts the RFC 791 definition of TTL:
Time to Live
An internet header field which indicates the upper bound on how long this internet datagram may exist.
My position has never been that it was a good idea to treat the TTL as anything other than a hop count (and in fact IPv6 has replaced the TTL with a simple hop count). I simply state that RFC 791, as it stands, requires that routers decrement the IPv4 TTL by no less than the number of seconds that a packet will take to arrive at the next hop.
You posted:
There has been discussion about what happens to the internet when we need to it to reach off planet (Moon/Mars). IP's issue was mostly address space (though newer versions of IP have MUCH bigger address spaces) IIRC, and TCP was the one with lag issues.
Indeed. TCP/IP was designed under the assumption that bits are cheap, roundtrips are short, and processing time is relatively expensive. Interplanetary data links are just the opposite, so we're never going to have FTP to Mars or to Jupiter (though we might have email, but probably not over SMTP or any TCP-based protocol).
Actually nothing you cite says the sender has to be aware of or compensate for speed-of-light delays of this magnitude.
The text I've cited states that the TTL specifies the maximum lifetime, in seconds, of a datagram. If your part of the network causes packets to live longer than their TTL, by whatever means, then you're violating the RFC. I'm not sure how much clearer I can make it.
In any case, it's largely irrelevant, since---as you said---IP was designed for terrestrial networks.
I disagree that Linux's "hardware support is iffy or incomplete", although it may look that way to people who don't know better (and arguably don't know enough to get much benefit out of Linux anyway):
Is Mac OS X's hardware support "iffy or incomplete" because it won't run on most PCs?
Some hardware is designed well; Other hardware is "designed for Windows". Is it reasonable to blame Linux developers for not supporting hardware that doesn't come with everything that is needed in order to use it (the missing stuff being interface documentation)?
Linux is ready for the desktop if you want it on your desktop. If you don't; why should anyone care?
Debian has been my desktop OS for most of the last decade. It's partly because I'm a free software zealot, but mostly because I am simply more productive with it than with the allegedly "Desktop-ready" alternatives (specifically, Windows 98/2K/XP, Mandriva, Linspire, and even other Debian-based distros like Xandros and Ubuntu).
I can't help but groan when I see yet another article telling me why "Linux isn't ready for the desktop", since I'm seeing the article on my Linux desktop.
Maybe the difference between me and these other people is that I actually use my computer to do work, and I can't afford to waste my time putzing around with a system that still hasn't managed to get multi-process scheduling working properly.
Also, my iPod only works if I mount it as/dev/sdc6. Don't know why that is, but the dev said he'd put it on his TODO list.
Well it's been working perfectly in Debian for years, including integrating with the Debian Unified Playlist Policy, so you can use any playlist-generating program with it.
Of course, there's no documentation, unless you install the gnupod-doc-nonfree package from non-free, because the documentation is licensed under the GFDL, with a Front Cover Text that says "This software requires Mandrake's misguided security policy" (even though it doesn't on Debian), and an Invariant Section titled "Apple is Teh Rocksor". (Actually, the documentation was in main for a few months, since somebody on the ftp-masters team decided to upload it unilaterally without consulting debian-legal, and it took a General Resolution to get it out. Apparently, the basis for uploading it to main in the first place was some FAQ on Sun's website, which said that "Everything is fine; Nothing is ruined." But I digress...)
Keyboard manufacturers: DON'T MESS WITH MY KEYBOARD. If you want to differentiate your keyboard, you can start by adding extra keys, which can be mapped to various functions. DON'T screw with existing keys.
Typing on a non-standard keyboard is extremely irritating. The worst keyboard I've used is one where the function keys were, BY DEFAULT, assigned to be those "Internet keys" (Help, Open, Save, Email, Web, etc.). You had to press some "Fn Lock" key in order to make them behave like proper function keys should, and every time you rebooted, the "Fn Lock" status got cleared. The same keyboard also had a double-height Delete key, and rearranged the Insert, Home, End, PageUp and PageDown keys, so I'd always end up pressing the wrong one. Oh, and I think it might have even screwed with the positions of the cursor keys, too, but I don't remember that one for sure.
The position of the CapsLock key is annoying, and I can see the justification for swapping it with the left Control key, but the manufacturers will inevitably screw it up, so it would be best if they just leave the keyboard alone, and let the software guys handle remapping and such.
Providing alternate keycaps would be nice, though, so that the labels printed on the keys could be changed to match the software keymap.
Slashdot Mirror
Not necessarily, but for the purposes of this discussion, sure.
What does that have to do with the previous sentence? Care to back that up with any facts? Or reasoning based on facts?
So far, the only argument in your favour seems to be, "it's truthy!"
Well, the theory is that meats contain chemicals (triglycerides, I think) that, when irradiated, sometimes become allegedly-toxic chemicals known as cyclobutanones. See Wikipedia. I don't know if that theory is scientific, or if it's just rubbish put forward retroactively by misguided activists.
Ultimately, it's a matter of risk management. I am in favour of the option that carries the lowest risk of killing me.
Hogwash! The truthiness of the matter is that "on error resume next" is evil. Period.
With the moral decay that we're seeing on Slashdot, soon people will be committing the most heinous crime of all against nature: using "goto".
It's Sodom and Gomorrah all over again! The end is near!!
The first thing to do would be to talk to your lawyer and find out if the police actually can do what you want in your jurisdiction. If so, file a complaint with the police department's Internal Affairs office. See what happens with that first. It's possible that some individual cops are just being lazy, and that might be all it takes to motivate them.
It's not just water that you have to worry about. Commercial-spec solid-state electronic components are typically rated for operation between 0 and 70 degrees Celsius. Electronic components conduct electricity better (or worse, in the case of many semiconductors) at lower temperatures, so even in a humidity-controlled environment, you could end up melting certain components.
What you need is either computers that are built entirely out of industrial or automotive-spec components that are rated at -40 to 85 degrees Celsius, or you need a temperature-controlled server room that will keep the computers within the commercial-spec range. Both are going to cost money.
This is more-or-less already being done, although not with this technology. (But new ways of doing it might prove to be more cost-effective.) See Wavelength division multiplexing.
Welcome to competition. The horror!
Perhaps, but it's not like binary-only software would be more successful in that community. Look at qmail: It's proprietary, source-available software. How many people would use it if DJB only released binaries for the platforms that he happened to be using when he wrote it?
No! Keep the Windows keys! I use them at least 20 times a day, and probably more like 100.
My Windows key opens up an xterm. :)
Yes, but if spyware isn't the real problem, then anti-spyware software isn't going to help.
Or wipe them using a platform that the virus code can't run on. If you're writing the virus, you'll know what its capabilities are.
Realistically, though, if you took reasonable measures to ensure that the virus wouldn't get out, then I doubt the courts would hold you liable. They might hold the disgruntled employee(s) liable, though.
Heh. Does the second #9 replace the first #9?
Seriously, though, what if someone got a copy of the virus through the lab's TEMPEST emissions?
I think trademark law might prevent that, since it would be an endorsement, but they might be able to cite a particular Consumer Reports article. I'd definitely get my lawyers' advice on something like that, though.
The vulnerabilities do exist; they're just not being exploited nearly as much. Of course, run-of-the-mill signature-based antivirus software is equally flawed, as Consumer Reports has shown and security geeks have already known.
Mac OS X has terrible hardware support: it only runs on Macintosh computers. People know and accept that, so it's a non-issue.
The problem here is that Linux's hardware support is so good (when compared to e.g. Mac OS X) that people have come to expect that it will support every piece of hardware is existance, but that will never happen until hardware manufacturers start using standardized interfaces for everything (and even then, there will still be people using "obsolete" hardware).
Despite what the zealots claim, Linux is not for everyone. No software is. If you want to run Linux, then you can expect that its hardware requirements will differ from those of your particular version of Windows. However, the impact of that is, in my experience, overstated, since once you've decided to run Linux, your future hardware purchases will be made with Linux in mind---and it really doesn't matter if Linux doesn't support hardware that you don't have anyway.
He shouldn't have to, but reality isn't always what it should be. Besides, missing features in trial software isn't all that unheard of. Think of it as a free trial with wireless networking disabled. If you use a large distribution like Debian, you have 14 CD-ROMs of software that you can install without ever going to the network. When you're satisfied that Debian (or whatever) is worth the investment, you can pay $100 to enable wireless networking, and you get free upgrades forever and there is never any "GNU Product Activation" nonsense to put up with.
My first Linux machine had a Winmodem. I installed Slackware 3.5 (which I downloaded using that Winmodem) onto a UMSDOS filesystem. When I decided that I eventually wanted to replace Windows with Linux, I saved up and bought a real modem. Sure, it would have been nice if Linux would have run perfectly out-of-the-box on the machine, but support for the Winmodem really wasn't all that important in the long run, and I'd much rather see resources put into developing better software than into fighting with the last few uncooperative hardware manufacturers.
No, that would be ridiculous. What I'm saying is that those "not-so well configured routers" are each supposed to decrement the TTL by some value greater than the actual number of seconds a transmission will take. It's fairly easy to do on a hob-by-hob basis; Anybody who doesn't know that their router is directly connected to a satellite uplink shouldn't be allowed near an Internet router anyway. It complicates the design of routers somewhat, but IPv4 has always assumed moderately complex routers, at least when compared with IPv6.
That's probably a sensible way to look at it, but it's not what the spec says. Your interpretation limits the TTL to just "processing time" or a "hop count", which contradicts the RFC 791 definition of TTL:
My position has never been that it was a good idea to treat the TTL as anything other than a hop count (and in fact IPv6 has replaced the TTL with a simple hop count). I simply state that RFC 791, as it stands, requires that routers decrement the IPv4 TTL by no less than the number of seconds that a packet will take to arrive at the next hop.
You posted:
Indeed. TCP/IP was designed under the assumption that bits are cheap, roundtrips are short, and processing time is relatively expensive. Interplanetary data links are just the opposite, so we're never going to have FTP to Mars or to Jupiter (though we might have email, but probably not over SMTP or any TCP-based protocol).
The text I've cited states that the TTL specifies the maximum lifetime, in seconds, of a datagram. If your part of the network causes packets to live longer than their TTL, by whatever means, then you're violating the RFC. I'm not sure how much clearer I can make it.
In any case, it's largely irrelevant, since---as you said---IP was designed for terrestrial networks.
Only compared to the others. Linux has a lot of ugliness (e.g. ptrace, procfs), which is probably why local root holes keep cropping up.
I disagree that Linux's "hardware support is iffy or incomplete", although it may look that way to people who don't know better (and arguably don't know enough to get much benefit out of Linux anyway):
Linux is ready for the desktop if you want it on your desktop. If you don't; why should anyone care?
Sigh. You paid $0 for the software. You can afford to get a new wireless card that works with said software.
Not everyone who qualifies as "you" lives in the U.S.
Debian has been my desktop OS for most of the last decade. It's partly because I'm a free software zealot, but mostly because I am simply more productive with it than with the allegedly "Desktop-ready" alternatives (specifically, Windows 98/2K/XP, Mandriva, Linspire, and even other Debian-based distros like Xandros and Ubuntu).
I can't help but groan when I see yet another article telling me why "Linux isn't ready for the desktop", since I'm seeing the article on my Linux desktop.
Maybe the difference between me and these other people is that I actually use my computer to do work, and I can't afford to waste my time putzing around with a system that still hasn't managed to get multi-process scheduling working properly.
Well it's been working perfectly in Debian for years, including integrating with the Debian Unified Playlist Policy, so you can use any playlist-generating program with it.
Of course, there's no documentation, unless you install the gnupod-doc-nonfree package from non-free, because the documentation is licensed under the GFDL, with a Front Cover Text that says "This software requires Mandrake's misguided security policy" (even though it doesn't on Debian), and an Invariant Section titled "Apple is Teh Rocksor". (Actually, the documentation was in main for a few months, since somebody on the ftp-masters team decided to upload it unilaterally without consulting debian-legal, and it took a General Resolution to get it out. Apparently, the basis for uploading it to main in the first place was some FAQ on Sun's website, which said that "Everything is fine; Nothing is ruined." But I digress...)
Keyboard manufacturers: DON'T MESS WITH MY KEYBOARD. If you want to differentiate your keyboard, you can start by adding extra keys, which can be mapped to various functions. DON'T screw with existing keys.
Typing on a non-standard keyboard is extremely irritating. The worst keyboard I've used is one where the function keys were, BY DEFAULT, assigned to be those "Internet keys" (Help, Open, Save, Email, Web, etc.). You had to press some "Fn Lock" key in order to make them behave like proper function keys should, and every time you rebooted, the "Fn Lock" status got cleared. The same keyboard also had a double-height Delete key, and rearranged the Insert, Home, End, PageUp and PageDown keys, so I'd always end up pressing the wrong one. Oh, and I think it might have even screwed with the positions of the cursor keys, too, but I don't remember that one for sure.
The position of the CapsLock key is annoying, and I can see the justification for swapping it with the left Control key, but the manufacturers will inevitably screw it up, so it would be best if they just leave the keyboard alone, and let the software guys handle remapping and such.
Providing alternate keycaps would be nice, though, so that the labels printed on the keys could be changed to match the software keymap.