I was trying to login with a few different mail clients, and it gave me various errors, except for Squirrelmail (Webmail) which worked fine. Logins from pretty much any machine on my network failed, unless it was through webmail.

The error message I was getting in the logfiles was:

Jun 19 23:28:55 behemoth dovecot: imap-login: Aborted login: rip=192.168.50.100, lip=192.168.50.1

That's it. No matter how much debugging I turned on, that's all it would say.

I eventually discovered the problem was that the server was refusing to even speak plaintext UNLESS the process talking to it was local (like Squirrelmail, for example). So the remote clients got immediately disconnected the very moment dovecot realized they weren't asking for an SSL/TLS connection.

You have two remedies available to you. On the client side, you can switch everyone to use TLS or SSL. This is obviously preferable, since it ensures that passwords are never sent in the clear.

However, if you are only using your imap locally or through encrypted tunnels, it's probably safe to use the server-side fix:

In /etc/dovecot/dovecot.conf:
disable_plaintext_auth=no

Then restart dovecot, and all should be well.

If, while setting up a computer on a domain, you happen to end up receiving this error, I wanted to let you know what the problem was (in my case).

In short, the computer's hostname was set to something invalid.

I am using a Samba primary domain controller which also provides DNS/WINS resolution, and the way I had it set up, the computer receiving this error message was apparently not able to resolve it's own name using the domain controller's DNS/WINS. This is what caused the problem.

To fix it, I removed the computer from the domain, changed it's name to something that would resolve correctly using DNS (just to be safe) and reconnected to the domain. 3 reboots later, and the error was gone!

Hopefully that's helpful to you. There's not a lot of information on this message on the net. Good luck fixing it!

I cooked up a rather simple orbital simulator because I was bored. Then I plugged in some numbers from Wikipedia's entries on the various celestial bodies and came up with a minimally accurate model of the inner solar system.

Once I was done that, which I thought was pretty neat, I decided to see what would happen if I replaced the asteroid belt with a ~0.5 solar mass star. Mars did not survive the experiment. However, the system then settled into a fairly stable system despite Earth's orbit becoming significantly perturbed.

Anyhow, if you'd like to try out my simplistic orbital simulator, you can get version 0.2.0-stable or if you're feeling adventurous, the latest development version (identical, at the time of this posting, but probably not for long)

Installation

You must first install Python if you're running Windows (either Official or ActiveState distribution). If you're not running Windows, you can skip this step as you probably already have Python installed.

Now you need to install PyGame.

Once you've completed these two relatively painless installations, you can simply download and double-click the .py file linked above.

Controls

Key	Action
P	Play / Pause
T	Toggle "Trails"
C	Clear Trails
+	Zoom in
-	Zoom out
S	Center on Sun
E	Center on Earth
N	Center on next object

Caveats

• Due to laziness, everything is rendered and calculated in a 2D plane. Realistically, the planets in the solar system (except Pluto) are nearly on a plane anyway, so this doesn't make a tremendous amount of difference, however, it will likely make modelling anything other than the solar system problematic.
• Accuracy is probably not extreme:
  • The software uses calculations from Newton's Universal Law of Gravitation which is known to be inexact, it makes no effort to include any of the effects of relativity.
  • Little effort is made to coerce numbers into the range of maximum accuracy for floating point calculations, and standard 8-byte python/IEEE floats are used which are notoriously inaccurate for calculations on this scale.
  • The gravitational calculations are based on a point-source model. All objects in the system are zero-volume point sources of gravitation. I have not even used the oblate ellipsoids they are often approximated as.
  • Moons are not included, because they are too close to their planets to be visible without extensive zooming and they have limited effect on solar orbits.
• All smaller objects, rings, asteroid belts, etc have been callously disregarded.
• The vector math can become buggy in two situations: When the direction is nearly 0 degrees or nearly 180 degrees; or when very very large accelerations come into play, for example when two objects are about to collide (but cannot, because they are points and have zero volume)

Enjoy!