Depends on turnover, and that's a big part of the problem. WalMart isn't going to buy anything risky, they know that the vast majority of the albums they sell will sell quickly, so the money will be available to buy another album to sell quickly...Profit on an individual item doesn't have to be high if you can turn it over enough times. Meanwhile, the perceived price of an album is whatever WalMart's price is. "Why do I have to pay $6 more for than I pay for Eminem? The independants can't afford to compete with WalMart on the hits, but can't survive only on non-hits.
My brother's take on this was to look up any potential vendor you saw in this month's CS in a 6 month old CS. Obviously wasn't foolproof, but he had decent luck.
The 300 miles between fillups depends on what you need to do to fill it up. I get about 350 miles range, and I'd gladly trade it for something that could go 80 miles on an overnight charge. If you added dual fuel capability, I'd trade my wife's car, too.
The primary tools I use to do my job are written in VB, and depend on Access and Excel. This is a custom app, with years of development. Although it has it's share of problems, it does the job well and most of the development has been tweaking it to match our business processes. Starting over with less proprietary infrastructure would be a good idea long-term, but my manager isn't encouraged to plan that far ahead.
Power brakes have a reserve of several full-travel pumps before power assist is lost, and even then will still work with more effort. Power steering is an assist only, and at any speed over 5mph isn't needed, especially in a car as small as a Spectrum.
I can understand that in a situation like that, not everyone will know enough about cars and have the presence of mind to do the right thing. I don't fault your wife for her handling of that, unless she was racing at high speed for an hour while on the phone to people who should be able to help figure out a solution.
I might believe a glitch that caused momentary unexpected acceleration, but I've got a hard time believing that the entire car is drive-by-wire, or that the throttle, foot brake, parking brake, transmission and ignition all failed in such a way as to prevent the car from being slowed to under 100 miles for 125 miles, even with outside advice. Similar cases have been found to be deliberate.
SpamAssasin is for email, and won't affect anyone trying to browse to your site. At worst, a properly-configured SpamAssasin would see a mention of your URL in an email, resolve it to the same IP as a spammer, and give it a few more points towards the spam threshhold. SpamAssasin (at least as used by my mail admin) scores messages based on various factors rather than giving pass/fail tests, so a suspicious URL in an otherwise non-spammy message wouldn't necessarily send it over the spam threshold.
Cost might be less than $100, but my time involved is worth considerably more than that. On the other hand, even on/off most hard drives will last several years. Does anyone know mean calander time between failures for 24/7 drives vs. say 8/7?
Nope, I didn't mean that one, I meant a different one of course.
Granted, the Linksys remote-admin problem is incredibly stupid, the user is still better off with it than without. At least each brand of router is likely to have a different stupid problem, and if it doesn't work, the average consumer can switch essentially transparently.
This is why the average broadband connection should be behind at least a consumer router, even if it's the only machine connected. Routers are too cheap and easy to skip.
This was my attitude, but apparenlty some of the most recnet spyware is more difficult. I manage to avoid spyware even on my windows boxes, so my experience is in cleaning up other people's computers.
For stuff that Spybot or Adaware can find but can't delete, I create a batch file with lines like
del "c:\program files\whatever\somespyware.dll".
and set a scheduled task to run this on boot. On Win9X machines, I add these lines to the autoexec.bat file.
Haven't run XP myself on the same hardware as anything else, but people I trust tell me that XP runs as well or better than other Windows on 233mhz and up machines. My experience with 98 vs 2000 on a PII 450 would support that--Printing made Winamp stutter on 98, worked fine on 2000 with no other changes.
In this case, the hardware upgrade that makes sense would have been a router. They are cheap enough that they are worthwhile for the firewall functions even on a single-user broadband connection, and they make setting up a new computer that much easier.
Asia gets blocked by some providers not because of the amount of spam, it's because of the proportion of spam to useful mail. I don't know anyone in Asia, so unfortunalty it makes sense for me to use a provider that blocks them.
The other problem is that we don't really have laws that give the government the power to just shut off the top spammers, and I don't know if I want them to have that power--Their list is too likely to include sites like http://www.johnkerryisadouchebagbutimvotingforhima nyway.com/ , based on the same investigative techniques that gave us the compelling evedence of Iraqi WMD.
I think it costs the mobile company a bit to keep a number active, whether minutes are used or not. I don't have a problem with something that is effectively a minimum quarterly charge, especially since the minutes don't expire if you keep the phone active.
It's only in the last 2 or so years that prepaid services here (Midwest US) were remotely competitive with contract plans. Before that they were 3-6 times more expensive per minute compared to a 100 minute per month contract, and the prepaid required you to purchase the phone while the contract included it in the price. Now we've got a couple of decent low-use plans, and one unlimited local calls, local coverage only for about the same as a land line. Since that one counts my work as a local call, where the landline was long distance, it winds up cheaper.
It depends on which consumers you're talking about. Millions of people bought 110 and disk film cameras, with picture quality about equal to the current $20 toy digital cameras. The equivilant holds true for music. A bigger issue is user interface. Many non-geeks can't deal well with multi-function devices. My mom will have a hard time switching funtions on a camera/mp3/phone/pda--for her, seperate physical devices are nearly essential.
Some of the prepaids are now reasonable as a payphone replacement. Since the customer pays for the phone, they keep cheap models avaialble--I've seen Virgin Mobile phones for $40 with no fine print. My wife got a bundle that was $160 for a small, simple flip-phone and $120 airtime, enough for the first year at the rate she uses it. Plus if one of these phones breaks, you can replace it with another $40 phone instead of paying an artificial "full price".
Good lord, indeed. I work where they build (or have built, in the case of Olds) versions of all those names. Buick offers thicker glass, and an engine/bodystyle combination that's available individually on the other models, but can only be combined with Buick. That's the complete list of substantial differences. Same chassis, same engines, same transmission, same assembly line, same people putting it together.
If I understand the article, you're stuck with their phones, their access points(!) their PBX and their service. While requiring certain phones and PBX makes sense, I don't see the point of requiring all-new 802.11 access points especially, and that makes me wonder if some artificial restrictions have been designed in to force lock-in when (not if) competitiors come up with similar systems.
Proably one of the biggest reasons Americans don't want Diesel passenger cars is the first ones most of us saw in quantity were the ones GM made. Mostly large cars, converted gasoline engines, horrible reliability, and while the mileage was better than the same displacement gasoline engine, the performance was far lower, so there wasn't a real advantage over a smaller, more economical gasoline engine. At least one used-car price guide had a note "If diesel, deduct 50%".
No, it's more like telling your neighbours that you're going to have a party and to contact you first if it gets too loud, they do, and then you get upset that when you don't stop, they call the police anyway.
Multiple valves, high RPM and turbocharging are all methods to increase horsepower without increasing engine size, or in this case methods to shrink engine size and increase economy without reducing the horsepower. 8 valve engines are more for economy of initial purchase price, rather than extended use.
Smaller pistons can reach high RPM with less stress than larger/heavier ones. The durability-limiting factor isn't rotary speed, it's the force generated by the piston and valves changing directions.
Even worse, does Microsoft manufacture every single of a million lines of an OS themselves? Hell no! Certainly not. There's an abundence of supplier supplying software for an OS, from the networking stack to the defragmenter in the control panel.
my math was off. but my point remains the same. be it 50 million or 50 thousand out of a billion, it is still worth it for spammers to track bounces. i'm not even saying all do. I'm just sayingt that known good addresses are worth far more to profitable spammers and so good addresses are worth digging for using a variety of techniques.
Spammers need lots of good addresses, but in general it's easier to get more good addresses than it is to weed out the bad ones from a list. harvesting? are you now contradicting yourself? if it's not important for spammmers to have "known-good" addresses(which is my point), why bother with harvesting?
Spammers need good addresses, not known-good. I'll bet that if you offered a spammer the choice of 10,000 known-good addresses, or a million addresses with 15,000 of them good, they'd take the million. harvesting is a generic term for obtaining known good addresses.
It's a term for obtaining likely-good addresses. Many harvested addresses are bogus, just not as many as randomly-generated addresses. you can harvest from websites, newsgroups, bounces, fake opt-outs , whatever...
the point of harvesting is to have a db of known-good addresses that you can blast for as long as possible.
well my 1 in 20 was "conservative" to say the least. i wouldn't doubt a ratio way in excess of 1 in 100 plausible.
think about a dictionary attack. how many attempts, given the domain name, would it take to find trichardson@stuff.com ?
A simple dictionary attack will get a high hit rate against aol.com, msn.com, any of the top ISP's. For smaller domains with fewer valid addresses, use the top 500,000 most popular usernames and combine these with each domain in turn. This will be far better than random. i think 1 in 100 is way too optimistic.
we're talking about people who continue even when they get less then a 5% response to their spam. if i was a spammer i'd definitely want to minimize the time & effort for that 5% response...and one signficant way is to make sure your spam is going to a live person.
Your math is off again. Direct mail would be happy with a 5% response rate, the best estimates I've seen of spammer response rates are below 0.05%. Spammers need to get out to as many total suckers as possible within the limits of their time and bandwidth. While decreasing the number of undeliverables is a good thing, a spammer would reach the point of diminishing returns long before doing all that would be necessary to pay attention to bounces.
Either your math is off, or my math is off. My math says 1/20th of 1 billion (us) is 50 million.
I don't think spamming makes sense, blindly or semi-blindly, but I think it's likely that spammers can generate addresses via harvesting and semi-intellegent dictionary attacks with more than 1 in 20 valid addresses. If it were hard to get the ratio under 1 in 100 valid addresses I might agree that tracking bounces is a good idea.
Another thing to consider--Due to IP based filters, bounces will vary based on the sending IP address.
Slashdot Mirror
Depends on turnover, and that's a big part of the problem. WalMart isn't going to buy anything risky, they know that the vast majority of the albums they sell will sell quickly, so the money will be available to buy another album to sell quickly...Profit on an individual item doesn't have to be high if you can turn it over enough times. Meanwhile, the perceived price of an album is whatever WalMart's price is. "Why do I have to pay $6 more for than I pay for Eminem? The independants can't afford to compete with WalMart on the hits, but can't survive only on non-hits.
My brother's take on this was to look up any potential vendor you saw in this month's CS in a 6 month old CS. Obviously wasn't foolproof, but he had decent luck.
The 300 miles between fillups depends on what you need to do to fill it up. I get about 350 miles range, and I'd gladly trade it for something that could go 80 miles on an overnight charge. If you added dual fuel capability, I'd trade my wife's car, too.
The primary tools I use to do my job are written in VB, and depend on Access and Excel. This is a custom app, with years of development. Although it has it's share of problems, it does the job well and most of the development has been tweaking it to match our business processes. Starting over with less proprietary infrastructure would be a good idea long-term, but my manager isn't encouraged to plan that far ahead.
Power brakes have a reserve of several full-travel pumps before power assist is lost, and even then will still work with more effort. Power steering is an assist only, and at any speed over 5mph isn't needed, especially in a car as small as a Spectrum.
I can understand that in a situation like that, not everyone will know enough about cars and have the presence of mind to do the right thing. I don't fault your wife for her handling of that, unless she was racing at high speed for an hour while on the phone to people who should be able to help figure out a solution.
I might believe a glitch that caused momentary unexpected acceleration, but I've got a hard time believing that the entire car is drive-by-wire, or that the throttle, foot brake, parking brake, transmission and ignition all failed in such a way as to prevent the car from being slowed to under 100 miles for 125 miles, even with outside advice. Similar cases have been found to be deliberate.
1. You want to test and burn in the machine, make sure everything actually works
2. You don't know a trustworthy source of barebones systems. Not all the local whitebox dealers are good.
3. You want a laptop
4. Your boss wants corporate-standard hardware
SpamAssasin is for email, and won't affect anyone trying to browse to your site. At worst, a properly-configured SpamAssasin would see a mention of your URL in an email, resolve it to the same IP as a spammer, and give it a few more points towards the spam threshhold. SpamAssasin (at least as used by my mail admin) scores messages based on various factors rather than giving pass/fail tests, so a suspicious URL in an otherwise non-spammy message wouldn't necessarily send it over the spam threshold.
Cost might be less than $100, but my time involved is worth considerably more than that. On the other hand, even on/off most hard drives will last several years. Does anyone know mean calander time between failures for 24/7 drives vs. say 8/7?
Nope, I didn't mean that one, I meant a different one of course.
Granted, the Linksys remote-admin problem is incredibly stupid, the user is still better off with it than without. At least each brand of router is likely to have a different stupid problem, and if it doesn't work, the average consumer can switch essentially transparently.
This is why the average broadband connection should be behind at least a consumer router, even if it's the only machine connected. Routers are too cheap and easy to skip.
This was my attitude, but apparenlty some of the most recnet spyware is more difficult. I manage to avoid spyware even on my windows boxes, so my experience is in cleaning up other people's computers.
For stuff that Spybot or Adaware can find but can't delete, I create a batch file with lines like
del "c:\program files\whatever\somespyware.dll".
and set a scheduled task to run this on boot. On Win9X machines, I add these lines to the autoexec.bat file.
Haven't run XP myself on the same hardware as anything else, but people I trust tell me that XP runs as well or better than other Windows on 233mhz and up machines. My experience with 98 vs 2000 on a PII 450 would support that--Printing made Winamp stutter on 98, worked fine on 2000 with no other changes.
In this case, the hardware upgrade that makes sense would have been a router. They are cheap enough that they are worthwhile for the firewall functions even on a single-user broadband connection, and they make setting up a new computer that much easier.
Asia gets blocked by some providers not because of the amount of spam, it's because of the proportion of spam to useful mail. I don't know anyone in Asia, so unfortunalty it makes sense for me to use a provider that blocks them.
a nyway.com/ , based on the same investigative techniques that gave us the compelling evedence of Iraqi WMD.
The other problem is that we don't really have laws that give the government the power to just shut off the top spammers, and I don't know if I want them to have that power--Their list is too likely to include sites like http://www.johnkerryisadouchebagbutimvotingforhim
I think it costs the mobile company a bit to keep a number active, whether minutes are used or not. I don't have a problem with something that is effectively a minimum quarterly charge, especially since the minutes don't expire if you keep the phone active.
It's only in the last 2 or so years that prepaid services here (Midwest US) were remotely competitive with contract plans. Before that they were 3-6 times more expensive per minute compared to a 100 minute per month contract, and the prepaid required you to purchase the phone while the contract included it in the price. Now we've got a couple of decent low-use plans, and one unlimited local calls, local coverage only for about the same as a land line. Since that one counts my work as a local call, where the landline was long distance, it winds up cheaper.
It depends on which consumers you're talking about. Millions of people bought 110 and disk film cameras, with picture quality about equal to the current $20 toy digital cameras. The equivilant holds true for music. A bigger issue is user interface. Many non-geeks can't deal well with multi-function devices. My mom will have a hard time switching funtions on a camera/mp3/phone/pda--for her, seperate physical devices are nearly essential.
Some of the prepaids are now reasonable as a payphone replacement. Since the customer pays for the phone, they keep cheap models avaialble--I've seen Virgin Mobile phones for $40 with no fine print. My wife got a bundle that was $160 for a small, simple flip-phone and $120 airtime, enough for the first year at the rate she uses it. Plus if one of these phones breaks, you can replace it with another $40 phone instead of paying an artificial "full price".
Good lord, indeed. I work where they build (or have built, in the case of Olds) versions of all those names. Buick offers thicker glass, and an engine/bodystyle combination that's available individually on the other models, but can only be combined with Buick. That's the complete list of substantial differences. Same chassis, same engines, same transmission, same assembly line, same people putting it together.
If I understand the article, you're stuck with their phones, their access points(!) their PBX and their service. While requiring certain phones and PBX makes sense, I don't see the point of requiring all-new 802.11 access points especially, and that makes me wonder if some artificial restrictions have been designed in to force lock-in when (not if) competitiors come up with similar systems.
Proably one of the biggest reasons Americans don't want Diesel passenger cars is the first ones most of us saw in quantity were the ones GM made. Mostly large cars, converted gasoline engines, horrible reliability, and while the mileage was better than the same displacement gasoline engine, the performance was far lower, so there wasn't a real advantage over a smaller, more economical gasoline engine. At least one used-car price guide had a note "If diesel, deduct 50%".
No, it's more like telling your neighbours that you're going to have a party and to contact you first if it gets too loud, they do, and then you get upset that when you don't stop, they call the police anyway.
Multiple valves, high RPM and turbocharging are all methods to increase horsepower without increasing engine size, or in this case methods to shrink engine size and increase economy without reducing the horsepower. 8 valve engines are more for economy of initial purchase price, rather than extended use.
Smaller pistons can reach high RPM with less stress than larger/heavier ones. The durability-limiting factor isn't rotary speed, it's the force generated by the piston and valves changing directions.
Even worse, does Microsoft manufacture every single of a million lines of an OS themselves? Hell no! Certainly not. There's an abundence of supplier supplying software for an OS, from the networking stack to the defragmenter in the control panel.
my math was off. but my point remains the same. be it 50 million or 50 thousand out of a billion, it is still worth it for spammers to track bounces. i'm not even saying all do. I'm just sayingt that known good addresses are worth far more to profitable spammers and so good addresses are worth digging for using a variety of techniques.
Spammers need lots of good addresses, but in general it's easier to get more good addresses than it is to weed out the bad ones from a list.
harvesting? are you now contradicting yourself? if it's not important for spammmers to have "known-good" addresses(which is my point), why bother with harvesting?
Spammers need good addresses, not known-good. I'll bet that if you offered a spammer the choice of 10,000 known-good addresses, or a million addresses with 15,000 of them good, they'd take the million.
harvesting is a generic term for obtaining known good addresses.
It's a term for obtaining likely-good addresses. Many harvested addresses are bogus, just not as many as randomly-generated addresses.
you can harvest from websites, newsgroups, bounces, fake opt-outs , whatever...
the point of harvesting is to have a db of known-good addresses that you can blast for as long as possible.
well my 1 in 20 was "conservative" to say the least. i wouldn't doubt a ratio way in excess of 1 in 100 plausible.
think about a dictionary attack. how many attempts, given the domain name, would it take to find trichardson@stuff.com ?
arichardson brichardson crichardson drichardson arichards brichards crichards drichards annrichards abrichards ann_richards...to infinity
A simple dictionary attack will get a high hit rate against aol.com, msn.com, any of the top ISP's. For smaller domains with fewer valid addresses, use the top 500,000 most popular usernames and combine these with each domain in turn. This will be far better than random.
i think 1 in 100 is way too optimistic.
we're talking about people who continue even when they get less then a 5% response to their spam. if i was a spammer i'd definitely want to minimize the time & effort for that 5% response...and one signficant way is to make sure your spam is going to a live person.
Your math is off again. Direct mail would be happy with a 5% response rate, the best estimates I've seen of spammer response rates are below 0.05%.
Spammers need to get out to as many total suckers as possible within the limits of their time and bandwidth. While decreasing the number of undeliverables is a good thing, a spammer would reach the point of diminishing returns long before doing all that would be necessary to pay attention to bounces.
Either your math is off, or my math is off. My math says 1/20th of 1 billion (us) is 50 million.
I don't think spamming makes sense, blindly or semi-blindly, but I think it's likely that spammers can generate addresses via harvesting and semi-intellegent dictionary attacks with more than 1 in 20 valid addresses. If it were hard to get the ratio under 1 in 100 valid addresses I might agree that tracking bounces is a good idea.
Another thing to consider--Due to IP based filters, bounces will vary based on the sending IP address.