Note that I don’t have any beef with Skype per se – I use their service all the time.

Original.

The people who installed this door thought were open-minded people, and decided that a bit of Braille could be useful to the blind community. Unfortunately, they put it in the one place where it would be least useful:

That’s right, they put it on the sign next to the door. It’s hard enough to blind people to navigate around and find doors as it is, and finding the correct door is frankly beyond my comprehension. But if I were in that situation, I’d guess that any indication of what lay beyond the door would be on the door itself, either in the middle or somewhere in the handle (or push-plate in this case). In fact, I’m pretty sure that’s what the good people at StockSigns were expecting too. What neither I nor (I hope) StockSigns would expect of blind people is to flail around the wall near the door in the hope of finding some information that’s of use to them.

In another part of the same building, there’s a cafeteria. There are four or five signs beside the door of this nature, pointing the direction to various other campus facilities. Not only are the signs equally hard to find, but someone had the brilliant ruse of putting the condiment stand in front of them, thereby making absolutely certain the Braille is completely useless.

Good work from the university on thinking of getting signs to help the blind. Complete failure in using them as such.

As always, you can find it on the Linux Icons page.

A new book is about to be published with the title Homeopathy in Healthcare: Effectiveness, Appropriateness, Safety, Costs. From the title alone it’s hard to say whether it’s a load of rubbish (i.e. pro-homeopathy), or scientifically rigorous. Luckily, Amazon have provided the blurb from the back cover:

This volume includes the full Health Technology Assessment (HTA) report on effectiveness, appropriateness, safety and costs of homoeopathy in health care. The report was commissioned by the Swiss health authorities to inform decision-making on the further inclusion of homoeopathy in the list of services covered by statutory health insurance. Other studies carried out as part of the Swiss Complementary Medicine Evaluation Programme (PEK) caused a massive stir due to their schematic and exclusively quantitative (negative-)outcomes for homoeopathy. The present report, in contrast, offers a differentiated evaluation of the practice of homoeopathy in health care. It confirms homoeopathy as a valuable addition to the conventional medical landscape – a status it has been holding for a long time in practical health care.

There’s a few choice parts of the above text that tell us everything we need to know about this book. First, it claims to “[confirm] homoeopathy as a valuable addition to the conventional medical landscape”, which is at odds with study after study showing that homeopathy is no more effective a treatment than placebo. Of course, homeopathists seem incredibly reluctant to take note of those studies, so it’s hardly surprising that they should be ignored once more.

We can also look at the raison d’être for the book to understand the motivations of its authors. Apparently, it presents the results of an assessment commissioned in response to previous studies that consistently gave “(negative-)outcomes for homeopathy”. In essence, many previous studies found homeopathy to be ineffective; the commissioners here weren’t happy with those results, so set up a study of their own. A study being instigated with this mindset can only lead to a strong suspicion that its instigators were far from impartial regarding of the outcome they desired.

It also pays to look at why the authors decided that the previous studies did not meet expectations. Apparently, those studies were “exclusively quantitative” in nature. How this can be perceived as a negative trait is a mystery. If you’re going to assess whether a particular treatment is effective, you treat a number of people with that treatment, another set of people with a placebo, and count how many people got better. That’s the only way you can tell if a treatment is effective – by counting patients. And guess what? That’s what quantitative means – counting results and comparing them. Apparently, this wasn’t good enough for the authors of this book – simply working out who got better and who didn’t and whether or not the treatment had any effect on those patients is no way to assess the quality of homeopathic remedies.

Unfortunately, the snippet from the cover doesn’t state what methods the authors used to reach their conclusion, but since they dislike quantitative experiments on general principle it’s pretty certain that their approach is unlikely to impress the more rigorous scientists in the healthcare profession. For the sake of the Swiss public, let’s just hope the people holding the reins of their healthcare feel the same way.

(Image taken from Hirotoshi’s DeviantArt page.)

Today there’s a couple of small updates to make the module play nicely with the very latest code from svn. In the last few weeks or months some of the multi-screen stuff has been improved, which actually makes eplanet work better on those systems. I love it when my code gets better without me having to do anything!

As always, full details and installation instructions are on the eplanet page.

The other day I updated the packages on my LTSP server. I did it the way I always do: update the server itself, then enter the chroot environment for the client image and update the packages there. During the update process, Synaptic told me that there were a bunch of packages that it could auto-remove since they were no longer required – this would save me some 133Mb of space in my client image. Since I always take any opportunity I can to reduce the client image size, I thought ‘yippee’ and consented to the package removal. Everything updated as normal, so I exited the chroot environment and proceeded to update the image:

sudo ltsp-update-image

The result? Nothing. No output, no error, and no updated image. I ran the command three more times just in case a miracle was due to me, but apparently not. After a bit of head-scratching, I headed over to the #ltsp IRC channel, and asked for advice. After explaining what I’d been doing, a nice person by the name of Gadi asked me to check whether or not I had either the ltsp-client or ltsp-client core packages installed in the chroot environment. It turns out that I hadn’t – Synaptic’s auto-remove had deleted them, which meant ltsp-update-image didn’t recognise an LTSP environment and sulked. (It would have been nice if it told me that, but never mind.) I re-installed both (which in turn reinstated half the packages that Synaptic said it could remove in the first place), and all was well – ltsp-update-image could do its thing.

The moral of the story is, don’t believe anything Synaptic tells you about auto-removing packages on an LTSP client image.

As a side note, somewhere in this mess my NBD server config got trashed – the /etc/nbd-server/config file had disappeared completely. No NBD config = no NBD server = unhappy LTSP clients. However, a bit of quick reading around some documentation meant that I could quickly reinstate it. For the sake of completeness I thought I’d post it here. I tailored the file to my specific setup, but since I accepted all the defaults when I first installed LTSP it should work for most people. In any case, it’s a good starting point.

[generic]
user = nbd
group = nbd

[export]
port = 2002
exportname = /opt/ltsp/images/i386.img

As server configs go, that’s a nice one.

The UEA’s computers were hacked sometime in 2009, and over 200,000 emails and other documents were obtained illegally. In late 2009, just before the UN conference in Copenhagen, a small portion of these emails were released onto the Web, causing no small measure of havoc.

In the wake of the release, the UEA and CRU were investigated under claims of misleading and fraudulent science. While some of their practices in sharing data and general transparency were found to be less than ideal, they were exonerated of all serious charges of misconduct in multiple investigations. Indeed, a completely new, independent temperature record was created to see how well CRU’s results held up, and there were more or less identical. In fact, the temperature increases seen in the new record were slightly higher than those published by CRU.

So what new details will this second release of stolen emails and documents release? In a word, none. Many people will use this to show that the people at CRU and their colleagues are still up to their same old tricks. But these documents were obtained at the same time as those from the original set in 2009, and cover the same time period. They involve the same people, doing the same work, in the same manner. Therefore it follows that the flaws in the working practices – those which have already been examined multiple times – will still be there to be seen.

So what can we say about the whole ClimateGate affair? From the first release, we know that during 1997-2009 CRU produced a robust and reliable global temperature record. There were a few less-than-ideal working practices in play, but nothing that came even close to requiring serious censure. And guess what? These new emails show exactly the same thing.

Yes folks, it’s a repeat – it’s the climate skeptics’ equivalent of Channel Dave. And since the person holding the stolen emails still has plenty in hand, we can look forward to many more repeats to come.