One can fix an engine or even put it together the other designs it.

I think that, in this case, it is more like someone trained to change your oil at one of those 5 minute places.

Someone working there CAN move on to bigger things, but it won't be because that training taught them how.

However, I still say it's correct that even on the basis of a small sample, you can rule out claims about the background population.

You can say that but you are wrong.

With a small, non-random sample you cannot say ANYTHING about anything.

You reach in, grab a ball at random and pull it out, and see that it's red.

Random is not the same as non-random.

A small sample size that is random is NOT THE SAME as a small sample size that is non-random.

It's trivially true that "any small sample is going to have some non-random attributes", but that doesn't mean the sample itself isn't random, ...

Again, your sample was not random.

No matter how many times you try to imply/claim that it was random, it was not random.

Some Slashdot commenters have shown that they need an article about basic statistics, more specifically what can be inferred even from a small sample.

There are lots of people out there (and here) who do not understand basic statistics. Bennett Haselton is one of them.

The FIRST problem is not the small sample set. It is that the small sample set is "some people on Amazon's Mechanical Turk who are willing to take a survey for $X". His sample set is flawed.

And his home-written "survey" is also flawed.

So his math is meaningless. Garbage-in, Garbage-out.

In order to deal with the flaw in his sample set he'd have to have a much larger sample set. OR a properly selected sample set.

THEN he'd need his "survey" re-written.

And only then could he try his hand at the math. He hasn't even explained what his margin of error is or which method he used to calculate it. BECAUSE HE DOES NOT UNDERSTAND STATISTICS.

Anyone who believes that buying private links into a providers network is the same as your traffic getting paid priority knows jack shit about network ops.

The Fallacy of Equivocation.

You've substituted the more specific network-jargon "priority" for the usage of "priority".

Once Netflix PAID Comcast then Comcast gave Netflix PRIORITY access to the Comcast network. The PRIORITY access means bypassing the choke point that Netflix was previously restricted to.

No one is saying that Comcast changed the QoS or priority of individual Netflix packets. But that is what you are denying.

Now I'm sure a bunch of people (who are not network engineers) are going to argue over the wording and philosophy as to whether or not buying paid links into a providers network constitutes priority or not. It's not.

Again, you are substituting a more specific network-jargon usage of "priority" that no one other than you is using.

And you are denying something that no one else is claiming.

That is the Fallacy of Equivocation.

The only difference that buying direct links in meant was that they got to skip the congestion in the peering points.

Which is what everyone, except you, is saying.

Once Netflix paid Comcast, Comcast users could suddenly get better access to Netflix.

But Comcast refused to do anything to address that congestion UNTIL NETFLIX PAID THEM.

It is the only machine on the network that uses that IP.

ON A WIRED WORKSTATION ON THAT NETWORK, go to http://www.whatismyip.com/ and see if the IP address it reports ends in .157.

ON A WIRELESS DEVICE ON THAT NETWORK, do the same.

This will tell you whether a machine on your network may be sending spam from the same address as your email server.

The traffic coming from my server is so ridiculously small, that I was shocked to begin getting messages like these from those email providers.

Not your server.

Your network.

Monitor the traffic going into or out-of your cable modem to see what is happening on outbound port 25 for that IP address. Do this for 24 hours.

Move your mail server to a different IP address if that is possible. You have 5 addresses, right?

The rejection messages are saying that YAHOO and HOTMAIL are seeing too many messages from your specific IP address.

GMAIL is accepting the messages but flagging them as spam.

It is extremely unlikely that three competing services are all using the same SMTP-blacklist (that they refuse to identify) to reject messages.

So, in other words, both of these messages are crap and not accurate.

They are similar messages from two different services. It is very unlikely that they are both claiming the same problem ... incorrectly.

You've had those IP addresses for 2 years without problems so it probably is not a pre-existing issue with the IP addresses.

Do you have a firewall that you can configure to monitor outbound port 25 attempts from your network? Or do you know how to use a sniffer such as Wireshark to do so?

Or can you move your email server to one of the other IP addresses you have? And see if it is still blocked?

Right now it is looking like the problem is on your network. Not Comcast and not GMAIL or YAHOO or HOTMAIL. I might be wrong. But if it were me, I'd test my network first. Otherwise, even if you do get through to YAHOO or HOTMAIL they'll look at the logs and say the same thing.

Deferred: 421 4.7.0 [TS01] Messages from XXX.XXX.XXX.XXX temporarily deferred due to user complaints - 4.16.55.1; see

That seems to indicate that at least one of your recipients at YAHOO is actively flagging your messages as spam. Maybe they have incorrectly written a rule that is doing so.

Deferred: 421 4.7.1 [TS03] All messages from XXX.XXX.XXX.XXX will be permanently deferred; Retrying will NOT succeed.

... and ...

Deferred: 421 RP-001 (BAY004-MC5F24) Unfortunately, some messages from XXX.XXX.XXX.XXX weren't sent. Please try again. We have limits for how many messages can be sent per hour and per day.

And that one seems to be saying that your IP address is sending too many messages.

How many messages per day are you sending?

The code is what matters. Here's a site with a bit more info:
http://tools.ietf.org/html/rfc3463

If HOTMAIL is rejecting with one code but YAHOO is rejecting with a different code then there may be THREE issues for him to deal with.

And since he is running a server he will most likely be using port 25. Encryption may change that. But for initial testing purposes he should skip encryption for HOTMAIL and YAHOO until he can determine why his messages are being rejected.

He's having problems with 3 services.

1. GMAIL - messages accepted but marked as spam.

2. YAHOO - messages rejected (what do the logs say?)

3. HOTMAIL - messages rejected (what do the logs say?)

So the first step is to look at the logs and see if the rejection message has any information in it. Do the rejection messages at YAHOO and HOTMAIL have the same code?

The next step is to check with a service like http://www.dnsgoodies.com/ to make sure that Comcast has configured their side correctly. The reverse DNS should point to your domain. You DO have a domain, right?

The more information you have before you contact Comcast, the better. Because the first 2 levels won't know anything about anything. They will be reading off of a script.

Really. He's offended by a FACEBOOK posting.

So he decides to write his own "survey" or whatever. Except he knows NOTHING about writing them. Or how to conduct them.

And then he puts it up on Amazon's Mechanical Turk site. Further evidence that he knows NOTHING about conducting a survey.

Which leads him to "analyize" the crap "data" that he has "collected".

The only "News for Nerds" here is how badly this was done. Anyone who publishes is (that would be you, Timothy) is an idiot for doing so. If anyone else had conducted this at any other site it would have been mocked here.

Sure, there is a lot of less-than-competent admins out there, but a lot more of the problem is political rather than technical than most people realize.

Yes. I think it is because the political issues stem from status battles. If you can overrule IT then you have more status.

If you cannot overrule IT then you have less status than the nerds.

And YOUR status, today, is worth more than the risk of someone else's life, possibly, sometime in the nebulous future.

Particularly because you can still blame IT for not being able to deal with the situation. After all, isn't that their job. That's if they can even prove that it was your demands that caused a problem. Because all the other managers had the same demands.

Anyone fired from Home Depot? Target? Any of the others?

Fascism begins when the efficiency of the Government becomes more important than the Rights of the People.

There are a lot of people out there who would like the world to be a bit more orderly. Even if there is a bit less freedom. As long as they're still at the top.

Think of all the dictatorships and such that would love to be able to lock down the Internet like that. With the support of those Canadian politicians and police.

If education could have worked, it would have worked by now.

So much for AI in doing anything useful in protecting systems, and it's not the overall workforce that needs educating ... it's the fucking gate keepers -- IT and software/hardware manufacturers.

The problem is that even if the IT people are competent they have to be MORE competent than everyone who can attack them. Why does everything have to be connected to the Internet?

And they have to that competent with the software/hardware that they're using. How many times has the purchasing decision been made before you've even been aware of the issue?

Which leads to the issues that the software/hardware vendors have within their own companies. Ship today and we'll patch tomorrow. Got to get to market before the competition.

And that isn't considering the problems that "management" at the company you work for keeps introducing. I cannot tell you how many times some executive simply had to have admin access on his laptop which resulted in massive infections being brought onto the network.

Security is easy --- in theory.
But it depends upon hundreds or thousands of decisions being made correctly. By people who have no incentive to protect the security of the systems you support.