No, spam is not illegal. State law has been superseded by the CANSPAM act. Fraud is illegal, but spam simply requires a valid labeling of the spam, a non-fraudulent contact address, and a valid "remove" mechanism.
Ignore the "is it advertising" problem. It's not soluble under US law due to free speech issues and the difficulty of determining what is advertising and what is protected speech, and it's a legal rats' nest that will tie up efforts for years. Stick to "bulk unsolicited communications", which makes it clear what is and is not spam based on the volume, not the content.
It should be possible to make a simple, clear policy just like the junk fax law that would restrict bulk unsolicited email, but it's repeatedly been shot down by the Direct Marketing Association. Such a law would encourage ISP's to act broadly and responsibly to protect their customers and to be willing to act against spamming customers, but until then the ISP's don't want to have the legal fights they can expect for cutting off spammers. It's too darned expensive for them!
Actually, DDOS'ing such sites does work. It costs the lazy ISP's real money that can be directly measured and can get their overly cautious policy makers to permit their staff to act. If you think not, then your site is unusual and fortunate in being able to set aggressive anti-spam policies, and I think you're a lucky person in a site that doesn't have anywhere near the size of the more popular free email account sites.
Shutting accounts down *does* cost time and money from staff who are responsible. Often, to a cautious site, that time and money is more than they can afford or are willing to waste, especially with a low profit item such as free accounts. And those free accounts help them bulk up their advertising numbers. Remember, they make business plans based on the amount of traffic and on click-through advertising of such accounts.
I'm sorry, but the abuse desks almost never do anything useful. They are constrained by the lack of manpower, and they are constrained by ISP policy from doing anything that could ever be considered censorship to avoid losing the "common carrier" protections they currently enjoy.
Moreover, for many ISP's, spammers with "pink" contracts pay good money and help keep pay the ISP's bills. Agis.net tried this, and it wasn't until the Cyberpromo spammers had their upstream routers DOS attacked to death that Agis stopped taking Cyberpromo's checks, despite Cyberpromo's demonstrably criminal and fraudulent behavior.
To an ISP on the edge, a paying customer is very valuable.
Occasionally, however, the head IT guy gets over-ridden by management or by available finances. I've been there, saying "we need to spend money on this" and having to make do with much less money, or even with a cut in funding. You need to document the problem in advance to cover your ass, and get it in print and saved offsite to protect yourself from that kind of mistake.
I've done that, too. It helped protect me from a nasty lawsuit because I demonstrated where I had told a consulting client, in print, when the systems would start failing and the resulting legal liabilities, and gotten it signed by the company notary.
Nah, under Linux you can trivially create new files as swap space when needed. It may mean they overflowed available partition space on critical systems, or were unable to administer a heavily loaded fast enough to add swap before it overflowed.
Knowingn nothing else, I'd guess they overflowed a key database partition. A lot of old programmers very foolishly over-partition available disk, trying to outguess the OS about what partition will need how much space and instead of protecting themselves from disastrous overflows, actually causing disastrous overflows. It's an old programmer's habit that's hard to train people out of.
It wasn't perfect, but it was a lot cleaner. Sun has been slowly shifting back to BSD style tools under pressure from their Linux competition: they switched to AT&T at a time when there considering a merger or purchase by AT&T, and the deal fell through and left them with a bunch of fairly poor tools. But the hardware was solid and ran Linux fairly well: We see a similar state with even slightly out of date Mac hardware now.
yaztromoNO@SPAMjsyncmanager.org wrote:
The primary requirement is that the developer has to make the source to the program available upon request to anyone they have distributed the binaries to. Thus, if you want to make things inconvienent to others, you can restrict the binary distribution, and then require someone to make a request in writing for the source, which you can then send via regular postal mail on diskette, CD-ROM, or as an OCR-able print-out (as the GPL does stipulate "machine readable"), for a fee of $100.
You really need to re-read the GPL. You can't charge a fee for it, and you can't prevent them from re-publishing it. This means that one smart-aleck who buys your binaries in a product gets to maintain their website with all the source code, and any attempt to manipulate the GPL in the way you describe is shot to hell.
Nah. MS had several other TCP stacks available for a quite modest purchase price, even some stacks that were noticably better than BSD's. Several technically brilliant but small companies went under when Microsoft folded TCP into Windows 3.1x and took out their markets.
You worked on XFree86, didn't you? I can tell by your cheerful approach to integrating code from anyone else.
More seriously, this is precisely how open source projects should work. It would be interesting to see if your projects remerge in 5 or 10 years.
If the projects are closed source, or not supposed to share closed source inside your company, you may have a legal or ethical quandary. You may also have a billing problem, if good ideas are coming from their project and not from yours that they can use, then maybe they're doing something fundamentally better than you and you should dump your project. Or maybe they just have the better coder or two over there who do these fixes.
No, it isnt acceptable. Its an abuse of the power of the professor, like incest or nepotism. Note that I did date a few of my instructors, but that was well after the course was over and the grade was issued.
And how about putting a bullet through the NIS startup tools? Those require a lengthy delay loop because the network port is not necessarily up yet. You can shave 15 seconds or so off your boot time by integrating the ypbind and autofs init scripts into a single tool that runs in the background and loops until the network is up, then runs ypbind and the automount daemons.
No, they're both dead at the hands of 802.11[fill in your favorite suffix]. Given that Zigbee and Bluetooth both have no security to speak of and never will, and the cost of 802.11[whatever] continues to drop, neither are worth investing or developing in.
Lots of people will spend lots of money with very exciting business plans and do the development for the niche applications used by others, but none of the developers or patent owners will get back the money they wasted on it.
The Chinese government *censors* all media, thez dont control it. There are things its not allowed to be used for, namely criticism of the Chinese government.
But selling connectivity to spammers, and ripping off stupid Americans by selling them Guaranteed Herbal Viagra cures is fine. It doesnt cause political dissent, it brings in foreign currency, and it encourages the growth of network infrastructure, which they badly need.
Re:Concur with the "no more registration required"
on
The Year In Ideas
·
· Score: 1
And if people changed passwords regularly, breaking into people's systems would be much more difficult. But neither is very likely.
You have my old job! Wow!
Was the vendor selected due to backroom deals by the VP's, who owned stock in the vendor or who arranged a "sweetheart deal" that turned out to be wildly more expensive than expected due to poor quality, or who did it as a "foot in the door" to sell stuff to that vendor's parent company that never actually got sold?
I really hated when that happened.
I've worked in such places, both commercially and in research. It's hard to do, and doesn't ride the dot-com or dot-bomb, but everyone sleeps well at night.
Good. Now tell us their name so we can use your recommendation, because without the advertising or your openness, no one else will buy from them and their lying, cheating colleagues will drive them out of business.
Hint: good word-of-mouth is priceless for rewarding good engineers and good companies.
Unfortunately, we can't avoid relying on advertising somewhat. Fortunately, we can check them somewhat, with things like www.tomshardware.com, which I love to pieces for their harsh testing. They haven't steered me wrong yet.
Amen. I've seen medical studies lauding one medication over another when the benefit was clearly due to the extra time and care they gave patients in the study, and another hospital that gave patients on the old medication before exactly the same time and care got less help from the new medicine.
It was nasty at the medical conference that year, because both hospitals got federal funding to study the new medicines and wanted to pursue related medicines, but being told "you're only getting better results because you're actually talking to your patients" point blank in a peer-reviewed paper is pretty embarassing.
No, the 7-bit thing is me and folks tired of badly done internationalization breaking things at odd and difficult to predict moments.
"Internationalization" is for things that need it. But even the most basic tenets of displaying systems or designing interfaces get thrown out the window when you have to deal with backwards languages (such as Hebrew), vertically written languages (such as a lot of Asian languages), and variant versions of Unicode (such as the oddnesses Microsoft has done to it).
When you then have to deal with formerly working software, such as the "dir" or "ls" command, being partly working and partly working in other languages because your basic terminal can't support those funky fonts and makes two wildly distinct filenames look like the same thing with no way to type the difference on a standard keyboard, it's a very serious problem and the approach should be dumped.
Unfortunately, "enterprise applications" are often ported to run on hardware that is brand new, on software that is less than six weeks old, due to new demands. This especially applies to databases, mail servers, and web servers, and features are demanded that are unsupportable with older core operating systems unless you backport.
Considering that plenty of modern motherboards have devices for which there are no device drivers in the 2.4 kernel, and that the performance benefits are quite significant for SMP, yes, and that plenty of old hardware keeps being ported into the newer kernels because no one every bothered to port it into older kenrels, and considering the length of the beta-testing for the even numbered kernels that is done in the odd number kernels (2.5 is the beta kernel, 2.6 is the release kernel), I'm happy to have 2.6 and use it in the field.
Fedora isn't bad. If you need lots of funky developer tools pre-compiled for it, look at the dag and dries repositories at http://dag.wieers.com/home-made/apt/. The stuff is fabulous.
Nothing makes it into Debian stable until it's already deprecated by the developers. Debian stable is *too* stable.
Slashdot Mirror
No, spam is not illegal. State law has been superseded by the CANSPAM act. Fraud is illegal, but spam simply requires a valid labeling of the spam, a non-fraudulent contact address, and a valid "remove" mechanism. Ignore the "is it advertising" problem. It's not soluble under US law due to free speech issues and the difficulty of determining what is advertising and what is protected speech, and it's a legal rats' nest that will tie up efforts for years. Stick to "bulk unsolicited communications", which makes it clear what is and is not spam based on the volume, not the content.
It should be possible to make a simple, clear policy just like the junk fax law that would restrict bulk unsolicited email, but it's repeatedly been shot down by the Direct Marketing Association. Such a law would encourage ISP's to act broadly and responsibly to protect their customers and to be willing to act against spamming customers, but until then the ISP's don't want to have the legal fights they can expect for cutting off spammers. It's too darned expensive for them!
Actually, DDOS'ing such sites does work. It costs the lazy ISP's real money that can be directly measured and can get their overly cautious policy makers to permit their staff to act. If you think not, then your site is unusual and fortunate in being able to set aggressive anti-spam policies, and I think you're a lucky person in a site that doesn't have anywhere near the size of the more popular free email account sites.
Shutting accounts down *does* cost time and money from staff who are responsible. Often, to a cautious site, that time and money is more than they can afford or are willing to waste, especially with a low profit item such as free accounts. And those free accounts help them bulk up their advertising numbers. Remember, they make business plans based on the amount of traffic and on click-through advertising of such accounts.
I'm sorry, but the abuse desks almost never do anything useful. They are constrained by the lack of manpower, and they are constrained by ISP policy from doing anything that could ever be considered censorship to avoid losing the "common carrier" protections they currently enjoy.
Moreover, for many ISP's, spammers with "pink" contracts pay good money and help keep pay the ISP's bills. Agis.net tried this, and it wasn't until the Cyberpromo spammers had their upstream routers DOS attacked to death that Agis stopped taking Cyberpromo's checks, despite Cyberpromo's demonstrably criminal and fraudulent behavior.
To an ISP on the edge, a paying customer is very valuable.
Occasionally, however, the head IT guy gets over-ridden by management or by available finances. I've been there, saying "we need to spend money on this" and having to make do with much less money, or even with a cut in funding. You need to document the problem in advance to cover your ass, and get it in print and saved offsite to protect yourself from that kind of mistake. I've done that, too. It helped protect me from a nasty lawsuit because I demonstrated where I had told a consulting client, in print, when the systems would start failing and the resulting legal liabilities, and gotten it signed by the company notary.
Nah, under Linux you can trivially create new files as swap space when needed. It may mean they overflowed available partition space on critical systems, or were unable to administer a heavily loaded fast enough to add swap before it overflowed.
Knowingn nothing else, I'd guess they overflowed a key database partition. A lot of old programmers very foolishly over-partition available disk, trying to outguess the OS about what partition will need how much space and instead of protecting themselves from disastrous overflows, actually causing disastrous overflows. It's an old programmer's habit that's hard to train people out of.
It wasn't perfect, but it was a lot cleaner. Sun has been slowly shifting back to BSD style tools under pressure from their Linux competition: they switched to AT&T at a time when there considering a merger or purchase by AT&T, and the deal fell through and left them with a bunch of fairly poor tools. But the hardware was solid and ran Linux fairly well: We see a similar state with even slightly out of date Mac hardware now.
yaztromoNO@SPAMjsyncmanager.org wrote: The primary requirement is that the developer has to make the source to the program available upon request to anyone they have distributed the binaries to. Thus, if you want to make things inconvienent to others, you can restrict the binary distribution, and then require someone to make a request in writing for the source, which you can then send via regular postal mail on diskette, CD-ROM, or as an OCR-able print-out (as the GPL does stipulate "machine readable"), for a fee of $100. You really need to re-read the GPL. You can't charge a fee for it, and you can't prevent them from re-publishing it. This means that one smart-aleck who buys your binaries in a product gets to maintain their website with all the source code, and any attempt to manipulate the GPL in the way you describe is shot to hell.
Nah. MS had several other TCP stacks available for a quite modest purchase price, even some stacks that were noticably better than BSD's. Several technically brilliant but small companies went under when Microsoft folded TCP into Windows 3.1x and took out their markets.
You worked on XFree86, didn't you? I can tell by your cheerful approach to integrating code from anyone else.
More seriously, this is precisely how open source projects should work. It would be interesting to see if your projects remerge in 5 or 10 years.
If the projects are closed source, or not supposed to share closed source inside your company, you may have a legal or ethical quandary. You may also have a billing problem, if good ideas are coming from their project and not from yours that they can use, then maybe they're doing something fundamentally better than you and you should dump your project. Or maybe they just have the better coder or two over there who do these fixes.
No, it isnt acceptable. Its an abuse of the power of the professor, like incest or nepotism. Note that I did date a few of my instructors, but that was well after the course was over and the grade was issued.
No, NT is based on VMS. Look into the old David Cutler lawsuits with DEC for details.
And how about putting a bullet through the NIS startup tools? Those require a lengthy delay loop because the network port is not necessarily up yet. You can shave 15 seconds or so off your boot time by integrating the ypbind and autofs init scripts into a single tool that runs in the background and loops until the network is up, then runs ypbind and the automount daemons.
No, they're both dead at the hands of 802.11[fill in your favorite suffix]. Given that Zigbee and Bluetooth both have no security to speak of and never will, and the cost of 802.11[whatever] continues to drop, neither are worth investing or developing in.
Lots of people will spend lots of money with very exciting business plans and do the development for the niche applications used by others, but none of the developers or patent owners will get back the money they wasted on it.
The Chinese government *censors* all media, thez dont control it. There are things its not allowed to be used for, namely criticism of the Chinese government.
But selling connectivity to spammers, and ripping off stupid Americans by selling them Guaranteed Herbal Viagra cures is fine. It doesnt cause political dissent, it brings in foreign currency, and it encourages the growth of network infrastructure, which they badly need.
And if people changed passwords regularly, breaking into people's systems would be much more difficult. But neither is very likely.
Call the company, tell them that their name came up on Slashdot and you'd like to say nice things about them by name. They might give you permission.
You have my old job! Wow! Was the vendor selected due to backroom deals by the VP's, who owned stock in the vendor or who arranged a "sweetheart deal" that turned out to be wildly more expensive than expected due to poor quality, or who did it as a "foot in the door" to sell stuff to that vendor's parent company that never actually got sold? I really hated when that happened.
I've worked in such places, both commercially and in research. It's hard to do, and doesn't ride the dot-com or dot-bomb, but everyone sleeps well at night.
Good. Now tell us their name so we can use your recommendation, because without the advertising or your openness, no one else will buy from them and their lying, cheating colleagues will drive them out of business. Hint: good word-of-mouth is priceless for rewarding good engineers and good companies.
Unfortunately, we can't avoid relying on advertising somewhat. Fortunately, we can check them somewhat, with things like www.tomshardware.com, which I love to pieces for their harsh testing. They haven't steered me wrong yet.
Amen. I've seen medical studies lauding one medication over another when the benefit was clearly due to the extra time and care they gave patients in the study, and another hospital that gave patients on the old medication before exactly the same time and care got less help from the new medicine.
It was nasty at the medical conference that year, because both hospitals got federal funding to study the new medicines and wanted to pursue related medicines, but being told "you're only getting better results because you're actually talking to your patients" point blank in a peer-reviewed paper is pretty embarassing.
No, the 7-bit thing is me and folks tired of badly done internationalization breaking things at odd and difficult to predict moments.
"Internationalization" is for things that need it. But even the most basic tenets of displaying systems or designing interfaces get thrown out the window when you have to deal with backwards languages (such as Hebrew), vertically written languages (such as a lot of Asian languages), and variant versions of Unicode (such as the oddnesses Microsoft has done to it).
When you then have to deal with formerly working software, such as the "dir" or "ls" command, being partly working and partly working in other languages because your basic terminal can't support those funky fonts and makes two wildly distinct filenames look like the same thing with no way to type the difference on a standard keyboard, it's a very serious problem and the approach should be dumped.
Unfortunately, "enterprise applications" are often ported to run on hardware that is brand new, on software that is less than six weeks old, due to new demands. This especially applies to databases, mail servers, and web servers, and features are demanded that are unsupportable with older core operating systems unless you backport.
Considering that plenty of modern motherboards have devices for which there are no device drivers in the 2.4 kernel, and that the performance benefits are quite significant for SMP, yes, and that plenty of old hardware keeps being ported into the newer kernels because no one every bothered to port it into older kenrels, and considering the length of the beta-testing for the even numbered kernels that is done in the odd number kernels (2.5 is the beta kernel, 2.6 is the release kernel), I'm happy to have 2.6 and use it in the field.
Fedora isn't bad. If you need lots of funky developer tools pre-compiled for it, look at the dag and dries repositories at http://dag.wieers.com/home-made/apt/. The stuff is fabulous. Nothing makes it into Debian stable until it's already deprecated by the developers. Debian stable is *too* stable.