I've looked into greylisting, but haven't tried it. Sure sounds like it works! Thanks for the info.
Greylisting certainly changes the economics of spam, which is what I'm trying to do by holding connections open. Dealing with greylisting may cost too much in terms of time and resources for the average spambot. Let's hope things stay that way!
I don't think the spam that gets to your mailbox has anything to do with what your method of notifying the senders is. That's a testament to your filters, and may I say congratulations!
If that much of your spam is Nigerian, rather than guys sending "from" people's Yahoo accounts, that's weird, but I guess this works for you, then.
Note that the MTA bounce that a legitimate sender would get when you reject at SMTP time would contain the string you included along with your 5xx code, which could be a URL leading to further instructions.
But the recipient, it at least many cases, isn't the person who sent the message!
What good does that do you or them?
If you reject at SMTP time, the sender (if one really exists) gets a notification from his SMTP server, including whatever string your server put in its 5xx response. If it was a spambot, nobody gets anything at all. Which is how it should be.
Simply not including the spam itself doesn't absolve you from contributing to massive amounts of email going to people who have nothing to do with anything. And that is still called backscatter.
I only hold connections open for a long time when the message has tested positive for viruses, has violated an SMTP rule, or achieved a very high spam score. Something simple like an invalid address gets an immediate 550.
Seems like most spam is sent as a joe job. Sending an email only once to each address helps, but it's still a lot of email that shouldn't be sent. Doesn't just that part of it complicate your configuration significantly?
I don't see why it's easier to do checking between the time a message is accepted and delivered. There's no "too late" to drop a connection; you haven't agreed to accept the mail until you send a 250 OK. Have sendmail check for viruses, spaminess, signatures, etc, and then decide whether to send 250 OK or disconnect.
The only advantage I see in your method is that it's more likely the Nigerians can play buzzword bingo with email terms.
It sounds like you send an enormous amount of backscatter, and are probably doing much more harm than good. It would be much better to simply drop the connection at SMTP time, rather than accepting and then generating a bounce. Or do like I do, and hold their connection open for a long time before actually dropping it.
Yeah, if I go back into the same thing, it's pretty good. Looks like it's mostly initial loading. Which, unfortunately, is the worst-case scenario for people wanting to poke around the thing off a Slashdot link.
I'm using Firefox 2 on Gentoo, Athlon 64 3000+, 1.5GB RAM. Every click, I'm waiting for seconds, watching it draw and fill every graphical element. It's really unusable.
Not really. Of course it's bad if people get in over their heads. But that's not an argument for why the tax code should be used to enforce social experiments to fix this. Charity begins (and should end) at home, not at the business end of a gun.
You're right about that being the accepted definition, but it's that definition that I'm taking issue with. The only reason to bring income into the question at all is an underlying assumption of Marxism.
What does he do with the rest of it that he doesn't consume? Would you call that "spending", or would you call it "investing"? Is it something that should be encouraged or discouraged?
Agreed about Amazon owing property tax and not sales tax.
But defining a tax as "progressive" or "regressive" carries the underlying assumption that every tax is an income tax. There's no particular reason to compare the amount paid via sales tax to a person's income; compare it to the amount he consumes. It's not regressive. It's perfectly flat.
The real answer is to reduce the power of government to the point where it simply isn't so critical exactly who holds what office. Right now, it matters a whole lot, because the federal government is basically unrestrained.
Look, this isn't complicated. The 13th had enough states for it to pass, even excluding the South. The 14th didn't. It, and it alone, is an example of a bare majority of Congress changing the Constitution. Which is wrong.
As for which amendment fundamentally changed the nature of the union: the 13th changed labor laws. The 14th created, out of whole cloth, the idea of a "citizen of the United States". It enabled citizens of every state to be directly governed by the federal government, which is precisely against its original purpose, and planted the seed for the effectively omnipotent, nanny-state federal government we have today.
The 14th was such a fundamental change to the nature of the union, that the radical Yankees didn't have enough states to vote for it, even excluding the South. So they used their Congressional majority (remember, the South is not yet back in Congress) to force the southern states to ratify the amendment, thereby ramming it down the throats of both the South and the opposing federal states.
The passage of the 14th amendment was so egregious, it should be null and void. But I suppose that's another story.
However, the language of the 2nd is so much stronger that that of, say, the 1st ("shall not be infringed" vs "Congress shall make no law") that the 2nd could be binding on the states without the need for the 14th.
Does a deal like this lend legitimacy to a ridiculous patent, thus encouraging more patent trolling? Would ignoring them have been better? Maybe this patent is actually meaningful, but I doubt it.
Re:Your car is too fat. Uncle Sam needs to trim it
on
The SUV Is Dethroned
·
· Score: 4, Insightful
Okay, twitter, let me see if I can follow your logic:
The problem was caused by government, government, and then government. Demonstrating the common affliction of irrational faith in government, your solution is now more government!
Slashdot Mirror
I've looked into greylisting, but haven't tried it. Sure sounds like it works! Thanks for the info.
Greylisting certainly changes the economics of spam, which is what I'm trying to do by holding connections open. Dealing with greylisting may cost too much in terms of time and resources for the average spambot. Let's hope things stay that way!
I don't think the spam that gets to your mailbox has anything to do with what your method of notifying the senders is. That's a testament to your filters, and may I say congratulations!
If that much of your spam is Nigerian, rather than guys sending "from" people's Yahoo accounts, that's weird, but I guess this works for you, then.
Note that the MTA bounce that a legitimate sender would get when you reject at SMTP time would contain the string you included along with your 5xx code, which could be a URL leading to further instructions.
But the recipient, it at least many cases, isn't the person who sent the message!
What good does that do you or them?
If you reject at SMTP time, the sender (if one really exists) gets a notification from his SMTP server, including whatever string your server put in its 5xx response. If it was a spambot, nobody gets anything at all. Which is how it should be.
Simply not including the spam itself doesn't absolve you from contributing to massive amounts of email going to people who have nothing to do with anything. And that is still called backscatter.
Thanks for the perspective on this.
I only hold connections open for a long time when the message has tested positive for viruses, has violated an SMTP rule, or achieved a very high spam score. Something simple like an invalid address gets an immediate 550.
Seems like most spam is sent as a joe job. Sending an email only once to each address helps, but it's still a lot of email that shouldn't be sent. Doesn't just that part of it complicate your configuration significantly?
I don't see why it's easier to do checking between the time a message is accepted and delivered. There's no "too late" to drop a connection; you haven't agreed to accept the mail until you send a 250 OK. Have sendmail check for viruses, spaminess, signatures, etc, and then decide whether to send 250 OK or disconnect.
The only advantage I see in your method is that it's more likely the Nigerians can play buzzword bingo with email terms.
I'm glad I live in Texas, not Oregon, where a bunch of bed-wetters will send a man to jail for defending his home.
It sounds like you send an enormous amount of backscatter, and are probably doing much more harm than good. It would be much better to simply drop the connection at SMTP time, rather than accepting and then generating a bounce. Or do like I do, and hold their connection open for a long time before actually dropping it.
Yeah, if I go back into the same thing, it's pretty good. Looks like it's mostly initial loading. Which, unfortunately, is the worst-case scenario for people wanting to poke around the thing off a Slashdot link.
I'm using Firefox 2 on Gentoo, Athlon 64 3000+, 1.5GB RAM. Every click, I'm waiting for seconds, watching it draw and fill every graphical element. It's really unusable.
These legal concepts apply to cars and other tangible goods but not to software. They should.
Would that not destroy hobby software, and much of OSS and Free Software along with it?
Not really. Of course it's bad if people get in over their heads. But that's not an argument for why the tax code should be used to enforce social experiments to fix this. Charity begins (and should end) at home, not at the business end of a gun.
You're right about that being the accepted definition, but it's that definition that I'm taking issue with. The only reason to bring income into the question at all is an underlying assumption of Marxism.
What does he do with the rest of it that he doesn't consume? Would you call that "spending", or would you call it "investing"? Is it something that should be encouraged or discouraged?
Agreed about Amazon owing property tax and not sales tax.
But defining a tax as "progressive" or "regressive" carries the underlying assumption that every tax is an income tax. There's no particular reason to compare the amount paid via sales tax to a person's income; compare it to the amount he consumes. It's not regressive. It's perfectly flat.
The real answer is to reduce the power of government to the point where it simply isn't so critical exactly who holds what office. Right now, it matters a whole lot, because the federal government is basically unrestrained.
Look, this isn't complicated. The 13th had enough states for it to pass, even excluding the South. The 14th didn't. It, and it alone, is an example of a bare majority of Congress changing the Constitution. Which is wrong.
As for which amendment fundamentally changed the nature of the union: the 13th changed labor laws. The 14th created, out of whole cloth, the idea of a "citizen of the United States". It enabled citizens of every state to be directly governed by the federal government, which is precisely against its original purpose, and planted the seed for the effectively omnipotent, nanny-state federal government we have today.
No it wasn't.
The 14th was such a fundamental change to the nature of the union, that the radical Yankees didn't have enough states to vote for it, even excluding the South. So they used their Congressional majority (remember, the South is not yet back in Congress) to force the southern states to ratify the amendment, thereby ramming it down the throats of both the South and the opposing federal states.
The passage of the 14th amendment was so egregious, it should be null and void. But I suppose that's another story.
However, the language of the 2nd is so much stronger that that of, say, the 1st ("shall not be infringed" vs "Congress shall make no law") that the 2nd could be binding on the states without the need for the 14th.
I was replying to a post that got modded Troll, which you probably didn't see. Try to pull it up; it might change your interpretation.
Windows is only $200 if your time is worth nothing.
In the case of Firefox, memory usage has ranged from 25MB to $MEMORY_AVAILABLE. Which sucks no matter how much you have.
The Hauppauge remote talks directly to the card? How stupid is that. Anyway, stick with the PCHDTV cards.
http://www.pchdtv.com/
Not sure about your troubles with remotes, but that doesn't have anything to do with the TV card.
Does a deal like this lend legitimacy to a ridiculous patent, thus encouraging more patent trolling? Would ignoring them have been better? Maybe this patent is actually meaningful, but I doubt it.
Okay, twitter, let me see if I can follow your logic:
The problem was caused by government, government, and then government. Demonstrating the common affliction of irrational faith in government, your solution is now more government!