MY new encryption key

The key’s randomart image is:

+--[ RSA 2048]----+
|        .        |
|       o         |
|        o        |
|     o o         |
|    . o S        |
|  ..   =         |
| +....  o        |
|=+o.oo..         |
|Xoo.Eo.          |
+-----------------+

At revision 666

I use subversion (SVN) to handle revision control on important documents and data as well as to keep a sync’d backup of all my important stuff on my office computer. Every time you commit a change to the repository, one per working day usually, a new revision is created. I committed changes on a paper I’m working on today and in doing so reached revision 666.

Super (Ultra)-powered links

• New York Times Article Skimmer: http://prototype.nytimes.com/gst/articleSkimmer/, a prototype article reader provided by the Times itself. This is the best way to read it online IMO
• Readability: Go here to retrieve your customized bookmarklet: http://lab.arc90.com/experiments/readability/, it’s a proxy that filters out all the crap (adds, graphics, pulsating flash applications, …) on a webpage you’re trying to read and makes the main text actually readable.
• Note in Reader: Link: Click on your shared items in Google Reader and look on the left hand sidebar for the bookmarklet you can drag to your bookmark bar. surf to a page you think is cool, click the bookmarklet, and the page gets shared along with your shared Google reader items, with a note and comments.

Sage and group cohomology

I’ve been thinking about computing group cohomology by computer
recenty, I thought I’d point out some features of Sage (the leading
free and open-source alternative to Magma, Maple, Mathematics, etc..).

Specifically, with the 4.1.1 release there was added a package for
computing p-group cohomlogy. Below is the link and description.

Also, if you don’t know about Sage, you can download and install it
here: http://www.sagemath.org/

One nice side-effect of installing Sage is that you get a whole bunch
of other useful mathematics programs pre-compiled, configured, and
installed for you, like GAP (for doing group-theory, symbolic linear algebra, and much more).

————————

http://sage.math.washington.edu/home/SimonKing/Cohomology/

New optional package p_group_cohomology version 1.0.2 (Simon A. King,
David J. Green) #6491 — The package p_group_cohomology can compute the
cohomology ring of a group with coefficients in a finite field of
order p. Its features include:

* Compute the cohomology ring with coefficients in
for any finite p-group, in terms of a minimal generating set and a
minimal set of algebraic relations. We use Benson’s criterion to prove
the completeness of the ring structure.
* Compute depth, dimension, Poincare series and a-invariants of
the cohomology rings.
* Compute the nil radical.
* Construct induced homomorphisms.
* The package includes a list of cohomology rings for all groups
of order 64.
* With the package, the cohomology for all groups of order 128 and
for the Sylow 2-subgroup of the third Conway group (order 1024) was
computed for the first time. The result of these and many other
computations (e.g., all but 6 groups of order 243) is accessible in a
repository on sage.math.

Facebook, no facebook

I gave in and started a new facebook account, under cover. Now I’ve deleted it again because I realized it’s stupid to be on facebook as your dog. It’s too bad that facebook is such a good organizational tool especially when one has so many friends who only check facebook and never check their email.

What alternatives are there? Start a listserve / google or yahoo group? A meetup group? Mass text messages several times a week? That seems expensive. I also feel a little guilty for being the only one not on facebook, so why should everyone adapt to my preference?

Graphical diff — FileMerge

There is a great introduction to the Mac xcode utility “FileMerge” here:

http://www.macdevcenter.com/pub/a/mac/2006/05/02/filemerge.html?page=1

Basically, its a graphical interface to diff, allowing you to compare two similar files for differences and merge or ignore them.

Bike-cam 3000 and my video editing woes

I built a handlebar mount for my digital camera the other day. The idea was to make some time-lapse movies of long rides and/or of community rides like the Critical Mass and the Courteous Mass (yes, two different rides on two different Fridays every month).

My first attempt at a camera mount sucked pretty bad. It was one of those Gorilla-pods wrapped around my seat post and seat stay so that the camera pointed backwards. What ended up happening though is that the bumping around on the road caused the camera to slowly point downwards toward my rear wheel and as a result I have hundreds of stills of my back wheel and some pavement. This is not very interesting.

My second attempt involved modifying a reflector bracket. I saw someone doing this in a HOWTO on Instructables. It turned out that somehow I had all the parts I needed at home, including the 1/4-20 bolt scavenged from an unused tripod, so it was easy to put together.

The new mount works great, except for the shaking. When I’m on the road every little bump in the pavement translates into a big bump on camera. So this is where my video editing woe comes in. I thought that it would be easy to run the video clip through some software to “stabilize” the image. IT turns out that this isn’t as easy as I’d like, or as free as I’d like.

First, in iMovie ‘08 which came with my mac there is no such image stabilization feature (nor is there a feature to speed up or slow down a clip, which is ridiculous). Also, Apple, in all their wisdom, decided to remove the capability to run 3rd party plugins. Why on earth would they do that?

So after downloading iMovie HD (came with iLife ‘06, outdated indeed) from the Apple website (for free, which is nice) and loading the “piStabilizer” plugin (which is not free, but you can use the demo version if you don’t mind a nice big “DEMO” watermark on your footage) I got my footage stabilized and sped up (because iMovie HD **DOES** have the slow-down/speed-up video effect that iMovie ‘08 lacks).

You can see it below (again, with a huge DEMO watermark which sucks).

Link to YouTube

More bandwidth experiments

Update to this post. Another common internet activity and the bandwidth it consumes:

• Watching Star Wars in a terminal
  ~ $ telnet towel.blinkenlights.nl
  Ave. 2.85 KB/sec

Bandwidth experiment

Since my internet has been slow lately, I decided to do a little experiment and see how much bandwidth is used while doing various internet activities.

Here’s a plot over 10 minutes of three things (in this order):

1. Listening to music on somebody’s myspace page
2. Watching a (MP4) video streamed by castfire
3. Watching a YouTube video

Float placement in LaTeX

I’ve been plagued by misplaced figures in a paper I’m writing, but fortunately I just figured out how to solve it and I want to write it down (here) so I don’t forget.

The problem involves a LaTeX document with sections and subsections. In one of the subsections I want to place a LARGE table (an instance of \begin{sidewaystable} … \end{sidewaystable}). However, no matter what parameters I give to the figure environment, LaTeX always decides to push the table’s appearance until sometime after the start of the next section (sometimes even to the very end of the document, right in the middle of the bibliography in fact!)

So basically I want to tell LaTeX that, no matter what, you should place the figure before the start of the next section. There should be a command to do this right?

\clearpage

This tells LaTeX to place all outstanding floats, on their own page if necessary, before continuing to the next page. So putting this after the table in question and before the start of the next section works perfectly.

Actually, now that I think about it, I’m not sure I know why this works in my document because the if I remove the table, the beginning of the next subsection starts in the middle of a page. So in principle, \clearpage could have the effect of putting the table on the first page after the start of said subsection.

Oh well..

(update — Thanks to Piet van Oostrum)