You know what though, you know what the immigrants who were welcomed by the Statue of Liberty did? They stopped at Ellis Island to register and apply for residency.
And the ancestors of the people who were manning the registration desks at Ellis Island had distributed infected blankets to American Indians, amongst other things. The new ones merely take away jobs and distribute money more efficiently.
I suggest that you go for as broad a spectrum of skills as possible. Any good administrator should be able to write code, debug applications, set up a network...
These aren't separate skills, they are part of a spectrum. If you do things right, don't worry about the first job you get. I suggest trying for a small shop where you get to be the jack of all trades. Take it for the first couple of years. Then with _that_ experience, go for a bigger company.
While you are at it, join your local LUG (or other IT group(s)). If you answer questions regularly and politely, you will have a history on that list. Also, you might be able to find work on that list itself.
That it/is/ going to damage reiserfs is beyond any doubt, no matter whether he's proven innocent, not proven guilty, or proven guilty. The name is tainted, and a business executive will not likely touch anything related to that person, no matter whether it gets taken over and run by other people or not.
That said, I think the best approach is the one used by SpamAssassin, of associating different scores to each of the tests that can be used to distinguish spam from ham. The scores are found using neural networks. While this has the minor disadvantage of requiring the message to be transported to your server (instead of being able to reject it as part of the SMTP transaction), it should have the highest reliability as far as picking spam from ham does. Usually I drop messages with a really high score, store messages with a high score in a separate per-user "Spam" folder and store the rest in the users' regular inbox.
First off, filtering on the content is wrong. Spam is about CONSENT, not CONTENT. The only possible value you can give for CONSENT is a reputation based on the initial components of the SMTP transaction (the envelope). Client IP, HELO, sender domain, sender address, recipient domain, recipient address. Given that the recipient address (and domain) aren't useful in doing anything other than blocking open relay probes, the only trustworthy way to check for consent is to use the reputation of the client IP.
If you think that using SA is right, you haven't understood the spam problem.
Humans actually have more false positives than software. Finding the one false opsitive in a full spam folder is more difficult.
We don't DROP email. We reject inline in the SMTP transaction. There is no data phase for most SMTP tranactions which we reject . We don't lose mail. Filtering can, and does, lose email.
The traditional use of a DNSBL was in BGP, and you were expected to block all traffic from that network. Currently that is restricted to SMTP only. Think of it like this "if your IP is in the sbl-xbl, we don't want your email at all" is a 100% spam scoring reputation check.
Spamhaus method of fighting spam dont stops 3/4 of the spam of the world. Probably graylists, bayesian analisys, and other methods stops far more.
You obviously don't run a mail server with > 1 user. The sbl-xbl list stops ~ 80% of our spam. That's for a small email service provider, defending only about 75 million email addresses.
Bayesian doesn't stop spam. It just flags stuff as possible spam. Humans are worse filters than any software. If you have to look for false positives in a spam folder, don't even bother to filter stuff. That is just a waste of CPU cycles.
On the smaller servers I run, recipient validation handles ~ 50% of the spam, the sbl-xbl stops ~ 80% of the rest, dynamic IP blocks and hostname checks stop the remaining.
The law is being made a lot more stringent, and every person whose personal data has been compromised can get compensation upto 5 crore INR (50 million INR) as civil damages, as well as criminal action leading to fines and/or imprisonment. Under Indian law, any affected individual can bring a criminal lawsuit, without having to wait for the government to intervene.
Why not tariff goods from China? Are you in favor of buying goods made by unpaid slaves in Chinese prisons?
Please do. I want to see electronics become cheaper here (I don't live in the USA).
The US is going to have to deal with the fact that access to Indian and Chinese markets requires that they also buy from there. And right now, people are cheaper in India and China than in the US.
Labour is cheap, technology is expensive. Call centre jobs are highly desirable in India. They offer a chance at a lifestyle which was previously enjoyed only by people in senior management. The alternatives to the wages you term as slave labour are worse (A call centre job for 200 USD/mth against a regular, blue collar job for 60 USD/mth, if you can get it, there _is_ a lot of competition. If you are one in a million, there are over a thousand people like you.).
What I would really like to see is compensatory tariffs imposed by governments globally to compensate for subsidies (including fixed exchange rates). It would work something like this (random numbers): 1 Yuan should be close to 2 USD now. However, China is holding the rate at 4 Yuan = 1 USD. Hence, there will be a tariff of 200%.
Slashdot Mirror
It can bounce through a bunch of relays anyway.
MUA -> Local server -> central corporate server -> outbound gateway -> MX as firewall -> delivery server.
Just that most of these are either internal, or carefully control who can use them for relaying.
The only time a reboot should be necessary is when the kernel is updated.
You know what though, you know what the immigrants who were welcomed by the Statue of Liberty did? They stopped at Ellis Island to register and apply for residency.
And the ancestors of the people who were manning the registration desks at Ellis Island had distributed infected blankets to American Indians, amongst other things. The new ones merely take away jobs and distribute money more efficiently.
Which company? I am seriously considering blocking blowback (including Out of Office replies) at work.
I suggest that you go for as broad a spectrum of skills as possible. Any good administrator should be able to write code, debug applications, set up a network...
These aren't separate skills, they are part of a spectrum. If you do things right, don't worry about the first job you get. I suggest trying for a small shop where you get to be the jack of all trades. Take it for the first couple of years. Then with _that_ experience, go for a bigger company.
While you are at it, join your local LUG (or other IT group(s)). If you answer questions regularly and politely, you will have a history on that list. Also, you might be able to find work on that list itself.
So if someone puts: *Wrote Perl6 interpreter in Haskell*, that wouldn't be considered?
I Think You Mean, Bangalore or Hyderabad, surely?
Political correctness is another matter, and not really relevant to this discussion.
I believe they call it newsspeak. Doubleplusungood.
That it /is/ going to damage reiserfs is beyond any doubt, no matter whether he's proven innocent, not proven guilty, or proven guilty. The name is tainted, and a business executive will not likely touch anything related to that person, no matter whether it gets taken over and run by other people or not.
You mean, like it hurt Microsoft?
They would obviously start by putting up the marketing dweebs and lawyers against the wall.
That said, I think the best approach is the one used by SpamAssassin, of associating different scores to each of the tests that can be used to distinguish spam from ham. The scores are found using neural networks. While this has the minor disadvantage of requiring the message to be transported to your server (instead of being able to reject it as part of the SMTP transaction), it should have the highest reliability as far as picking spam from ham does. Usually I drop messages with a really high score, store messages with a high score in a separate per-user "Spam" folder and store the rest in the users' regular inbox.
First off, filtering on the content is wrong. Spam is about CONSENT, not CONTENT. The only possible value you can give for CONSENT is a reputation based on the initial components of the SMTP transaction (the envelope). Client IP, HELO, sender domain, sender address, recipient domain, recipient address. Given that the recipient address (and domain) aren't useful in doing anything other than blocking open relay probes, the only trustworthy way to check for consent is to use the reputation of the client IP.
If you think that using SA is right, you haven't understood the spam problem.
Humans actually have more false positives than software. Finding the one false opsitive in a full spam folder is more difficult.
We don't DROP email. We reject inline in the SMTP transaction. There is no data phase for most SMTP tranactions which we reject . We don't lose mail. Filtering can, and does, lose email.
The traditional use of a DNSBL was in BGP, and you were expected to block all traffic from that network. Currently that is restricted to SMTP only. Think of it like this "if your IP is in the sbl-xbl, we don't want your email at all" is a 100% spam scoring reputation check.
Or Google buys out members of the *AA?
Screw SRTP. IPv6 with mandatory crypto (AH and ESP) FTW.
That's how I know this estimate is shit. RBLs do not stop much spam. I know because according to them, I work for a spammer, and it hasn't stopped us.
Which particular spammer?
Gmail is a pretty small player in the mail field.
*Small company == small in number of employees.
92K messages in a maillog file? Over what time period? Is that a toy server?
My current estimates say that $ORK is blocking ~ 400 to 500 million messages a day using DNSBLs, about 80% of which is the sbl-xbl.
Nah, AOL, Hotmail and Yahoo! are bigger than us. Otherwise, no one else. Like I said, small company.
Spamhaus method of fighting spam dont stops 3/4 of the spam of the world. Probably graylists, bayesian analisys, and other methods stops far more.
You obviously don't run a mail server with > 1 user. The sbl-xbl list stops ~ 80% of our spam. That's for a small email service provider, defending only about 75 million email addresses.
Bayesian doesn't stop spam. It just flags stuff as possible spam. Humans are worse filters than any software. If you have to look for false positives in a spam folder, don't even bother to filter stuff. That is just a waste of CPU cycles.
On the smaller servers I run, recipient validation handles ~ 50% of the spam, the sbl-xbl stops ~ 80% of the rest, dynamic IP blocks and hostname checks stop the remaining.
New Delhi has long refused to sign the Nuclear Non-Proliferation Treaty (NPT) and has aggressively pursued the development of nuclear weapons.
The Indian government has only been asking that the treaty be fair, and everyone disarm. An unfair balance of power is not acceptable.
Actually, 2.56 USD. 0x100 cents.
Perhaps adding the term "review" to the search would help? I personally like dpreview.com for camera information though.
The law is being made a lot more stringent, and every person whose personal data has been compromised can get compensation upto 5 crore INR (50 million INR) as civil damages, as well as criminal action leading to fines and/or imprisonment. Under Indian law, any affected individual can bring a criminal lawsuit, without having to wait for the government to intervene.
/ message/2848
http://tech.groups.yahoo.com/group/cyberlaw-india
So why not switch to real thin clients (preferably with X)? Much cheaper than running Windows, particularly as your user count goes up.
Why not tariff goods from China? Are you in favor of buying goods made by unpaid slaves in Chinese prisons?
Please do. I want to see electronics become cheaper here (I don't live in the USA).
The US is going to have to deal with the fact that access to Indian and Chinese markets requires that they also buy from there. And right now, people are cheaper in India and China than in the US.
Labour is cheap, technology is expensive. Call centre jobs are highly desirable in India. They offer a chance at a lifestyle which was previously enjoyed only by people in senior management. The alternatives to the wages you term as slave labour are worse (A call centre job for 200 USD/mth against a regular, blue collar job for 60 USD/mth, if you can get it, there _is_ a lot of competition. If you are one in a million, there are over a thousand people like you.).
What I would really like to see is compensatory tariffs imposed by governments globally to compensate for subsidies (including fixed exchange rates). It would work something like this (random numbers):
1 Yuan should be close to 2 USD now. However, China is holding the rate at 4 Yuan = 1 USD. Hence, there will be a tariff of 200%.
Ford changes the bubblegum to a secret formula, and starts its own chain of repair shops. The third party objects to that combination.