What's being acquired as evidence is very wide, and the NSA is famous for both large data storage and building a database of interpersonal connections. Regardless if the particular reason this information is being gathered, I'm working under the assumption that they're going to be using the information in whatever way they can, rather than for the original reason they're taking the data.

I find it really concerning that a secret court can order such wide data transfer to the NSA, and also order that the order be kept secret.

I think you've got the right idea.

Performance reviews are a good feedback mechanism for both the worker and the manager. It sounds to me like this particular programmer is feeling disconnected and as such doesn't seem to be taking account of the reporcussions for the way in which he's doing his work. Likewise this is bad for morale for everybody else, because they're having to clean up the mess. Nobody wins in this scenario -- the lone programmer is unhappy and the people he works with that want to work collaboratively are unhappy.

So the recommendation I have is to at least try to have a conversation with the programmer about these things and try to help him reconnect with his fellow employees. This is not a "blame session", it's an attempt to get the team reconnected so that they can work as a team. It sounds like the programmer isn't able to say "I don't understand this" and as such can't learn from his fellow employees because of perceived judgment or a sense of insecurity. This is common, and can lead to all kinds of bullying behavior -- but it's the underlying disconnection that's the biggest part of the problem, IMHO.

Hopefully this works. If not, hopefully the effort wil bring to light what needs to be done.

Agreed. Debian workstations work fairly well. [Occasionally I have to build a newer kernel for new hardware, but I can deal with that.] I like that things are stable during the release cycle.

And in Debian Gnome3 now has a dependency on NetworkManager.

Users of the Wicd networking manager should be aware of this, because NetworkManager conflicts with Wicd. Neither Wicd no NetworkManager work when they're both active, and at the moment there's no warning about this nor instructions on what to do about it. :-(

That's interesting, and certainly possible.

You're probably right. "Real" shielded structures (such as anechoic chambers to test antennas) usually have conductive "fingers" in the doors, filters on incoming power lines, screens covering the air ducts... etc. Dispersion, intersymbol interference/distortion, multipath interference... those are more likely.

I didn't want to get into all of that; sometimes I just want to give a more "simple" kind of answer. :-P

Instead of "passively blocking", I think you mean "shielding". As in a " Faraday cage". This doesn't hamper signals outside of the structure.

Asking Slashdot, you'll get vastly differing opinions but you'll likely get some interesting information which will likely be useful.
However... the only way to really know which distro is "right" for you is to try several and pick one.
So I would suggest geting out your favorite Virtualization software (VirtualBox, KVM, etc) and try several distros at once. Find out what you like.

I recently did this back in August 2012. Here are the distros I liked:

*Mint Debian 201204
Fedora 17 (18 has a yucky installer)
OpenSuSE 12.1
*Debian (Squeeze, Wheezy, Sid)
Arch (fastest package installs by far, no sound in a VirtualBox VM)
*Pear Linux 5 (looks just like a Mac, except a Pear in place of an Apple. Fun to play with.)
Slackware 13
Vector Linux7.0 (based on Slackware, has package management. Fun, snappy.)

Least recommended: Gentoo. Attempting a base install + KDE4 was a THREE FULL DAY compile, after which X refused to start. Very frustrating.
2nd least recommended: FreeBSD, only because there's no GUI by default, and I couldn't find instructions to install one.

Regardless of which you choose in the end, best of luck to you. :)

I started with Slackare in 1995, used it up to 1999 -- I found it easy to use, but it wasn't easy to upgrade because upgrading meant reinstalling because it didn't have package management at that time. In 1999 I went to Debian because it allowed upgrading in place. I've since tested Slackare 13 and found it easy to use. ;-) [I didn't test whatever pakage management it may or may not have had, though.]

I don't like the Unity GUI in Ubuntu either (it's awful), but Mint has very few developers and I've been hearing complaints about the timeliness of downloading updates due to the small number of servers in their infrastructure. So I guess I'm saying that I agree, but that there are some other things to consider.

The only time I've heard of this happening are for mail servers that also run mailing lists. I've been running my own server for 14 years and I haven't been blacklisted once. (Maybe I've just been lucky?)

Yes... it's got a lot of advantages. Dealing with your own email server is rather interesting, it does take time to "get it right" (that takes a couple of years of tweaking) but after that it generally requires very little maintenance. And it's nice to have a server online to store and distribute files, have a personal blog that won't go away, etc.

If I remember correctly (it's been a long time) the user I was dealing with was deleting messages he had dealt with, but got into the habit of referring back to the messages when necessary in the Trash folder. I suggested moving these sort of messages to another locally-stored folder, but the user refused saying "I'm used to working out of my Trash now."

Logical. I considered trying that, but never did.

That reminds me... another thing that was going on back then was that a small set of users refused to delete any messages off the server, and at the time the server used MBOX storage and ext3 had a 2 GB filesize limit. One particular Sales guy got so much email that he would hit the 2 GB storage limit about every 3 months. He'd then call me up and tell me he wasn't getting his mail, I'd find his mailbox full, and I had to use 'mutt' to go delete old messages. I kept requesting to set his mail client to delete messages older than 3 months, but he kept refusing to allow me to do that. Instead he kept calling me every three months to fix his mailbox.

The next attempt at a fix was to make a script to watch mailbox sizes and warn the users of those mailboxes that were getting full to do something about it or to call me. I didn't expect that to work -- and it didn't -- but it was the next thing to try.

The final solution was to modify the "watch mailbox sizes" script to first warn the user about their mailbox size at a "soft" threshold, and let them know that when the mailbox size reaches a "hard" threshold that the system will automatically delete old messages out of their mailbox. I wrote another script that did just that, which (if I remember correctly) deleted messages that were older than 3 months. And with that, I never got another call about users reaching their mailbox size limits.

Right. Or they build up bad habits of using the system in a way that it wasn't designed for -- just like your story about the garbage bins used to store backup tapes. (Good story, BTW -- thanks.)

Ah. Good point. True.
I was referring to the destination MTA (your server) not generating a bounce, and in that equation I hadn't thought about the possibility of the sending MTA sending a bounce.

Thanks for the correction. :)

/dev/null isn't, but sadly the "Trash" folder is.

A few years ago I was working as an email administrator and got a call to someone's desk that was having a problem with their mail client because some of the folders were too full. One of them was Trash, so I was about to erase messages from the folder when the user paniced; "wait, that's important!"

For whatever reason, they were using the Trash folder for "real work"
(Sigh.)

That's true.

How bad this is depends on the system -- in this case it sounds like Shaw was doing "accept, then drop" which is the worst case because no one is notified of the failure. If however Shaw rejected spam rather than accept it, the sending mail system would notify the sender that the message was not delivered. It should be noted that this latter solution also does not cause backscatter because it doesn't generate a bounce. [For a bounce to occur, the message first needs to be accepted, but then for some reason cannot be delivered.]

You don't even need to be on the Evil Twin network to do this; at least some Torrent clients have a setting for what IP address to report that you're using. Some time ago I remember researchers from one of the big-name universities (MIT?) doing this and reporting the IP address of local network printers, and successfully received DMCA notices sent to the university reporting Torrent use by the IP address of the printer.

The other thing I'll mention is that doing a "tit-for-tat" hack on the Evil Twin network is probably the wrong thing to do.

For others suggesting the use of a directional antenna -- this is trickier than it sounds. Most directional antennas (like yagis) have a narrow horizonatal beamwidth, but a very wide virtical beamwidth (when mounted for the horizonatal beamwidth -- vice-versa when mounted for the virtical beamwidth). A dish might be more pinpoint, depending on the type of feedhorn used. Likewise there's no limit to the "depth" that the signal travels -- so for instance if you think that a particular appartment building contains a transmitter you're trying to locate, it's not simple to pin down which appartment has the transmitter. This is furthermore exaserbated by reflective surfaces, including typical corrogated metal floors used in building construction. Typical direction-finding requires a set of adjustible attenuators to further localize the originating strong signnal from the weaker reflected signals. As tough as this can be in an urban environment, "DFing" (direction finding) using a directional antenna is still cheaper and more reliable than using a Doppler Scanner.