At the risk of educating spammers, I filter any mail with the following words in the subject header to a folder called potential junk. I then scan once a week to see if I have an legit e-mails (order by e-mail address). My hotmail account gets roughly 20 spams a day. The combination of my filter and hotmail's junk filter usually reduces my inbox to one junkmail a day.
Credit
Money
Income
Debt
free
casino
invest
adult
win
join
xxx
earn
sex
golf
naughty
girls
hot
cock
penis
cost
Enhance
$
!
I've never understood this and maybe somebody can engage. Isn't it really just a matter of how you implement?
I am assuming that most randomizers use chaotic formulas which means we don't need to worry about predictable patterns (In the topological sense there are patterns, but you can't determine what the next number will be without running through the equations yourself)
Of course if you continually seed the pattern with the same number, your starting value will always be the same.
My perspective:
I use python whrandom as an example
x=whrandom.whrandom(1,1,1)
int(x.random*100)
>1
int(x.random*100)
>89
now set y-whrandom.wharandom(1,1,1) and you get the same values for the same steps.
If I can predict how you seed, I've got your sequence. I can also chart out the number the 16 million start combinations and the first couple of thousand iterations to perform statistical analysis on what is more likely to pop up.
But:
But what about this.
RandomVal=''
x=whrandom.whrandom()
while len(RandomVal) 10:
y=whrandon.whrandom()
selectChoice=int(y.random()*100)
# 44 arbitrarily chosen
if selectChoice == 44:
# you could replace int(..) with a function that gives you alpha numerics
RandomVal=int(x*10)
You now run along one sequence line, but the start and stop points are based on multiple datetime seedings. Obviously this could be made more complex making statistical prediction as well as seed guessing a moot point.
Basically randomize your radomize selector which selects the next randomize selector, etc... This methodology, while slow, eventually gets affected by processor temperature, other processes, etc...
Go easy on my, I'm just an amateur.
Note whrandom.whrandom() without seed numbers uses a time based seed.
Slightly off topic, but still a good point.
I was the victim of that sort of scam with a ski-do rental operation. They tell you if you exceed 6 knots in the marina, you forfeit your $150 deposit. Of course the whole group attempts to stay at 6 knots, not having an accurate way of measuring. When we came back in, the entire group, including folks that didn't know each other, were fined for exceeding the speed limit. How did they know? We were creating a wake. Why didn't they tell us? Because they wanted an extra $150 over the $50 rental.
Those of us who had our deposits on credit card, immediately called our companies to dispute the charge. Those who gave them cash had an uphill battle. In the end they dropped the "fine" because the credit card companies took our side. As a side effect those people who gave them cash got their money back as well.
Think about it. If I were to write a corpate e-mail and try to spy, using this method, there is a decent chance that I would get caught. Would it be worth risking my job? On the other hand, spammers would love to use this method, but who forwards spam?
Personally, if I caught somebody setting up a URL to capture javascript spying, I would fire a script that hit that URL a hundred times a minute. I would have the script grab random stuff off the web for content:-)
Patents are a necessary way to protect those who invest a lot of time innovating otherwise, where is the motive if you know someone is going to steal your work. The trick is to differentiate between what most people consider real patents vs frivolous patents. A new non-obvious "fool proof" encryption scheme would seem worthy of a patent to me. On the other hand, trying to patent hyperlinks, etc... is a waste of everybody's time. Ohhh, I think I'll patent the concept of "Table of contents", then all book publishers can pay me BIG money.
Re:Use encryption needlessly, constantly! [MUCH MO
on
The Encryption Wars
·
· Score: 1
I am assuming this concept is more on principle than reality. If you are trying to protect your privacy, you're better off locking your computer, not writing down passwords, using hard to hack passwords, and looking for clues as to whether your sysadmin is crooked.
If you are really worried about all your e-mails being scanned, a much more realistic (and easy) way to overload the scanners is to include alert words in your signature.
- Kill the president, guns, rockets bombs, blowup the FBI building, access to explosives, purchase high quality cocaine, etc...
Wonder what trouble I'm getting myself into.
It means I no longer have to be confused in Java. Now I can be confused in my favorite language.
I think the big deal is DTD validation is now available.
Slashdot Mirror
At the risk of educating spammers, I filter any mail with the following words in the subject header to a folder called potential junk. I then scan once a week to see if I have an legit e-mails (order by e-mail address). My hotmail account gets roughly 20 spams a day. The combination of my filter and hotmail's junk filter usually reduces my inbox to one junkmail a day.
Credit
Money
Income
Debt
free
casino
invest
adult
win
join
xxx
earn
sex
golf
naughty
girls
hot
cock
penis
cost
Enhance
$
!
I've never understood this and maybe somebody can engage. Isn't it really just a matter of how you implement?
I am assuming that most randomizers use chaotic formulas which means we don't need to worry about predictable patterns (In the topological sense there are patterns, but you can't determine what the next number will be without running through the equations yourself)
Of course if you continually seed the pattern with the same number, your starting value will always be the same.
My perspective:
I use python whrandom as an example
x=whrandom.whrandom(1,1,1)
int(x.random*100)
>1
int(x.random*100)
>89
now set y-whrandom.wharandom(1,1,1) and you get the same values for the same steps.
If I can predict how you seed, I've got your sequence. I can also chart out the number the 16 million start combinations and the first couple of thousand iterations to perform statistical analysis on what is more likely to pop up.
But:
But what about this.
RandomVal=''
x=whrandom.whrandom()
while len(RandomVal) 10:
y=whrandon.whrandom()
selectChoice=int(y.random()*100)
# 44 arbitrarily chosen
if selectChoice == 44:
# you could replace int(..) with a function that gives you alpha numerics
RandomVal=int(x*10)
You now run along one sequence line, but the start and stop points are based on multiple datetime seedings. Obviously this could be made more complex making statistical prediction as well as seed guessing a moot point.
Basically randomize your radomize selector which selects the next randomize selector, etc... This methodology, while slow, eventually gets affected by processor temperature, other processes, etc...
Go easy on my, I'm just an amateur.
Note whrandom.whrandom() without seed numbers uses a time based seed.
Slightly off topic, but still a good point. I was the victim of that sort of scam with a ski-do rental operation. They tell you if you exceed 6 knots in the marina, you forfeit your $150 deposit. Of course the whole group attempts to stay at 6 knots, not having an accurate way of measuring. When we came back in, the entire group, including folks that didn't know each other, were fined for exceeding the speed limit. How did they know? We were creating a wake. Why didn't they tell us? Because they wanted an extra $150 over the $50 rental. Those of us who had our deposits on credit card, immediately called our companies to dispute the charge. Those who gave them cash had an uphill battle. In the end they dropped the "fine" because the credit card companies took our side. As a side effect those people who gave them cash got their money back as well.
My last job, I worked with folks who had their MBA's and when we ran out of toilet paper, their MBA's came in compltely handy.
Or the daily show for that matter. What do they do? Tell real news?
Think about it. If I were to write a corpate e-mail and try to spy, using this method, there is a decent chance that I would get caught. Would it be worth risking my job? On the other hand, spammers would love to use this method, but who forwards spam? Personally, if I caught somebody setting up a URL to capture javascript spying, I would fire a script that hit that URL a hundred times a minute. I would have the script grab random stuff off the web for content :-)
Patents are a necessary way to protect those who invest a lot of time innovating otherwise, where is the motive if you know someone is going to steal your work. The trick is to differentiate between what most people consider real patents vs frivolous patents. A new non-obvious "fool proof" encryption scheme would seem worthy of a patent to me. On the other hand, trying to patent hyperlinks, etc... is a waste of everybody's time. Ohhh, I think I'll patent the concept of "Table of contents", then all book publishers can pay me BIG money.
I am assuming this concept is more on principle than reality. If you are trying to protect your privacy, you're better off locking your computer, not writing down passwords, using hard to hack passwords, and looking for clues as to whether your sysadmin is crooked. If you are really worried about all your e-mails being scanned, a much more realistic (and easy) way to overload the scanners is to include alert words in your signature. - Kill the president, guns, rockets bombs, blowup the FBI building, access to explosives, purchase high quality cocaine, etc... Wonder what trouble I'm getting myself into.
It means I no longer have to be confused in Java. Now I can be confused in my favorite language. I think the big deal is DTD validation is now available.