I recently had to replace a cracked top screen of a Nintendo DS - the replacement part was less than 10% of the cost of buying a new one. I had read very briefly that it was "a little fiddly", but not too tricky, so I just ordered the part. Only after it had arrived and with screwdriver in hand that I realised the scope of that fiddlyness.

[Image of DS in many parts]

I followed the excellent thirty minute guide here: http://www.youtube.com/watch?v=5a1AIWYXPmg and it went almost without a hitch. Just two things to add.

Another dense post - I would apologise, but if you don't know what the title means, you wouldn't be interested in what the post says. When persuading Matplotlib to render its text using TeX on Ubuntu 10.4 (useTex=true), you may find that:

the base texlive install does not ship with the type1cm package. You may need to install some of the extra packages to get all the goodies that come bundled with other latex distributions.

Latest installment of the series "this took longer than it should have to sort out", in the hope it may be useful to someone else.

My computer has two E5520 Intel processors (i.e. eight cores), and all of my development happens on a number of VirtualBox machines. Today, I tried assigning more cores and memory to a Windows 7 guest, and it failed to start up with error VERR_VMX_MSR_LOCKED_OR_DISABLED - the log gives a little more information:

00:00:01.418 HWACCM: No VT-x or AMD-V CPU extension found. Reason VERR_VMX_MSR_LOCKED_OR_DISABLED

Making a good "screencast" recording from an Ubuntu 9.10 desktop is simple. I found editing that video (in my case for uploading to YouTube) to be not so simple. Read on for how I did it.

Following a routine upgrade of a Trac site from version 0.10 to 0.11, I needed to change the database from sqlite to PostgreSQL, because the former (at a little over 1GB) had been giving the dreaded "database is locked" errors. This meant that continuous builds, handled by the Trac plugin Bitten, were often failing to complete. This post describes how I fixed the couple of problems I encountered.

So I wrote some software that outputs test results in a simple xml format, which can be converted to preferred formats (csv, brief and verbose html) with xslt. These results are calculated with every commit, and archived along with lots of other data from the build.