I'd always wondered if someone's decryption passphrase was a short description of how he committed a crime like:

"I buried the murder weapon at gps xxx,yyy"

"I had premeditation in committing my illegal acts"

Then it would definitely be self-incriminating to reveal his passphrase, and if forced to, he could get the conviction overturned on appeal.

I'm afraid that your information is out of date, with the progression of Moore's Law, even ROT-39 now fails to today's botnet based distributed key cracking apps, i recommend using 128-bit ROT keys (or even-256 bit ROT keys for super secure data) despite the performance impact requirements.

Using just ROT-128bit (aka ROT-3.40282367e38 or ROT-340,282,367,000,000,000,000,000,000,000,000,000,000) should keep your data secure for years.

I might add the Swartz was charged with 13 felonies, with a maximum sentence of 65 years in felony lockup, effectively life in prison. Murder, even multiple murders, has no more harsh a punishment (except in death penalty states).

You're point is valid, but it's, at best,a Type D "crime" being punished as a type A "the most harsh society can inflict" and might not even be a good civil suit for mild contract violation.

At the risk of admitting to being "whoooshed", i have to wonder if this is sarcasm?

It's wrong in so many ways, i can't even being to start.

Well, it was mainly meant as a tongue-in-cheek dig at the folks in Redmund.

However, while it's not like I've gone to trouble of checking it, it's my understanding that modern password guessing dictionaries are incredibly extensive and have lengthy sections of common key combinations such as single letter repetitions of all acceptable lengths, numeric sequences, and keyboard patterns like qwerty, extended qwerty (qwertyuiop[]\asdfghjkl;'z), as well as many more folks have been dreaming up for decades now.

Of course the webpage is just a local javascript for simple complexity checking, but it's important to remember that it's not really a good simulation of a password's unguessability.

Some "best" password tested results from the Microsoft site:
  - aaaaaaaaaaaaaaaaaaaaaaaaaaaaa
  - 123456789012345678901234567890123456789012345
  - qwertyuiop[]\asdfghjkl;'z
  - `1234567890-=qwertyuiop[

I'm thinking they don't do dictionary attacks here...

As a former broadband network architect (Telocity/Directv DSL) my understanding of modern Cable and DSL provider use of the term "modem" is that Modem implies a tunneled link between the CPE (Customer Premise Equipment) and a box at the ISP.

So here at my home using Comcast, we have a CPE Cable Ethernet/Cable COAX proxy that gets a outward facing DHCP IP from a server at the local Comcast ISP. This IP is part of a private non-Internet network. The CPE/Proxy then connects to a server at Comcast's data center using a PPTP tunnel, authenticating (customer-side) using keys stored in the CPE and (server-side) with a database of current active paid customers.

If authentication completes successfully, then the CPE Device links the PPTP tunnel to my local network and to the real internet on the far side of the multi-interface ISP Server.

so it's more like:
_______________{------- PPTP tunnel ------}
PC-- local net -- CPE -- ISP Private net -- ISP Server -- Internet

This is much more like a real telephony modem than a router or network type converting Gateway.

The reason they use this is of course now all they need to do to cut you off is change the ISP Database if you don't pay your bill.

If the real Internet was run out to your house, you could just run your own CPE and get free service until they sent a Tech to unwire your link (or have smarter more expensive hubs that can remotely cut-off links). That's too expensive to be cost effective, so they use the "modem" model to control costs. They usually make you pay for the CPE too so it's really a net gain.

Interestingly, this is why rebooting your CPE "modem" often fixes broken Cable Internet connectivity. It forces it to connect out to a new PPTP server and if the ISP PPTP server being broken/overloaded is the issue, then you have a whole new one and it tends to work better.

No, instead I sniff somebody else's pickles.

Who needs iVillage. Speaking from recent experience: Toddlers don't do what they're told because either they weren't listening or you weren't telling them something they wanted to do.

The solution is simple and goes something like:
"Billy... Hey Billy...Billy! BILLY! Now eat all the cookies and spill the milk everywhere. Good Boy! That's it!".

Follow that model and you'll have no trouble.

My only problem is figuring out how to reply to this kind of post by hitting the "Parent" button.

I'm sure the design targets memory intensive applications like VMware ESX virtualization hosting servers. (You could also use Solaris Zones or Xen Server)

With 8-16GB of ram statically assigned to each Guest VM (Virtual Machine), 128GB only covers 7 to 15 hosted Servers (less ESX memory overhead)

If you're doing VDI (Virtualized Desktops with Vista), that's only up to 31 VM PCs per blade.

Storage is commonly not an issue/botteneck since a SAN is often used (It works even with VMotion).