Long time, no blog. Again.

I've done a few things since I last wrote in here. My Lisp hacking is going better. I've read more of Peter Seibel which is funny and engaging and at the same time very thorough and informative. I've also read some of Paul Graham's introductory book which is very succint and seems to betray a dislike of object oriented programming. Quite handy though, it's a lot smaller than Seibel and includes a useful reference. I also have CLtL2 installed from the Debian package and browsable with w3m mode in Emacs. I find this nicer to use than the Hyperspec. Talking of which, I found the free personal edition of LispWorks and had a quick play with it. To be honest, though, I think I prefer working with SLIME and Emacs.

We went on the Poppy Line of the North Norfolk Railway last weekend. The weather was lovely. See flickr photos. Reminds me of being a kid!

I went to a training event at Aston University in April on the National Grid Service (nothing to do with electricity). It's basically a grid computing service for UK e-Science (technologies for enabling research, also called "cyberinfrastructure" in the US and e-Research in Australia). It all seems quite clever. They use the Globus toolkit which has a dreadful Web portal for job management and a better command line interface. Using a Web interface for managing jobs I found pretty counterintuitive, so the command line interface is definitely my preference. The selection of applications installed on the NGS is heavily biased to science research (bioinformatics, chemical modelling, etc.) though they do have the Weka text mining framework installed and seemed quite amenable to users installing other software.

At the event I met someone called David Woolls who works in computational forensic linguistics. I've invited him to come and give a seminar at Goldsmiths and to show us how his software may be of use in our music literature analysis for Purcell Plus.

I haven't posted since the London snow which turned out to be a grand excuse for everyone to have a day off. I struggled back to London from Norwich, going on the DLR for first time. For some reason, while various tube lines were closed, the DLR seemed to work. Anyway, I got to College to find it was closed which was pretty annoying. There are some pictures on flickr.

I started using Twitter, mainly because there was an Emacs mode. I'm not really sure why it's a good idea. And also, someone is sitting on my nick.

I gave up on KDE after trying KDE 4. 4.0 wasn't ready for general use. 4.1 fixed loads of bugs and 4.2 was quite a bit more complete. But it's scarily reminiscent of Vista. So now I'm doing everything (mail, irc, jabber) except Web browsing in Emacs, using Fluxbox for handy keybindings and putting Emacs in a workspace, and using Conkeror for Web browsing. It means I don't have to deal with the fiddly little mouse on my Acer Aspire One. I've learned enough emacs lisp to be able to write a few handy extensions for things like file association, presence management (irc and jabber), and tinyurl and keep meaning to try using emacs as a scripting language

So that's me for now. Maybe I'll blog more about my Lisp learning and Emacs hacks.

Just a quick note to say: I've just passed my viva! So I'll be MMus from 15th July (when it's conferred).

It was really good so sit down and talk about what I'd been researching for two years with two people who actually knew what I was talking about and found it interesting!

set-fill-column no longer requires a prefix argument. So where I used to have to do C-u 100 RET C-x f I can now just do C-x f RET 100 RET which is much more semantically pleasing.

Tab completion now fills in blanks where it can unambiguously. So for example, C-x C-f nesc-in TAB will complete to nesc-workshop-invitation.txt.

A slightly weird one is that, for wrapped lines, next-line (C-n or pressing the down arrow) instead of moving the cursor to the next line of the buffer (as it used to) now moves the cursor to the next line on the screen, even if it's within the same line of the buffer. Similarly for previous-line. The result is that Emacs now behaves a bit more like GUI text editors (such as Kate and Gedit). It's quite easy to get used to but was a bit of shock when I first noticed it.

Anti-aliased fonts in Gtk Emacs are very lovely indeed

National Express have increased their rail fares from next week.

I regularly travel from London to Norwich for the weekend and my selection of single advanced fares with a Young Persons' Rail Card include: Mythical (�3.95), Cheap and Cheerful (�5.95); Acceptable (�7.90); Annoying (�10.55); Emergency Only (�13.85); and then various Absolute No No prices.

So I bought my tickets for this weekend just now and on Friday Cheap and Cheerful, and Acceptable were quite easy to get which is great. But on Monday they've all vanished and been replaced by a whole load of completely alien prices. It's like when you get into your office and someone has moved all the stuff around on your desk. It's slightly disconcerting.

I'm guessing this is in relation to simplification of ticket naming schemes. In fact, National Express already operated the Anytime, Off-peak, Advanced system so I don't understand why they're changing their ticket prices.

While I'm on the subject, I wish they'd sort out their quiet coach policy. They always say "please avoid using mobile phones, laptops and other audio equipment". In fact, it's people talking which is most annoying in a quiet coach. First Great Western say "please keep noise levels to a minimum" which is much more straightforward.

</rant>

So Harriet Harman said yesterday, "The public who elected us, and whom we represent, expect us to sort it out. As Leader of the House, I want to say--on behalf, I hope, of the whole House--'We get it. We're going to sort it out.' That goes for all of us. We are aware of the scale of anger and outrage. We all believe that the issue needs to be sorted out, and we will work together urgently to do exactly that."

This seems to be quite a common theme in this whole business; MPs saying, we must sort this out because the public don't like it. But I would've thought that they should sort it out because it's wrong, not because of how it influences our opinion of them. It's like they're saying, now that we've been caught we need to stop doing it. Would they still be making dishonest expenses claims now if the Telegraph hadn't published details of it, if the public weren't aware? It seems to me that it's only because of the matter coming to light that they've chosen to try to stamp it out.

For real, we've just seen a concert given by Philip Glass at the Theatre Royal in Norwich! It's important to understand the sheer enormity of this. Glass is one of the most well known American composers of the twentieth century and rarely plays outside the States. He played some of the Metamorphosis, Mad Rush, some of his new Etudes, a few bits of film music and Opening from Glassworks. The audience absolutely loved it. He played two encores and received two standing ovations.

I appreciate his status as a key figure in what critics call minimalism, but I have one overriding impression from this concert: Glass just isn't a performer. He has very little sense of his own stage presence. He rushes and fumbles over notes; there were quite a few very obvious mistakes in his performance tonight. Most of his speed changes didn't make musical sense and were clearly just as a result of his pieces being too hard for him to stay in time. (And these pieces really are hard.)

I still thoroughly enjoyed it, though. Now I can add Philip Glass to my list of towering figures of music I've seen in real life. I think the only other person on that list at the moment is Pierre Boulez, but it's a good start.

I'm just doing my thesis corrections and, apart from a few general typos, one thing my external examiner picked up on was what he described as "inconsistency" in my quoting style; sometimes my punctuation was outside the quote marks and sometimes inside.

Now this is a little system that I thought I had invented and which I knew would probably get me into trouble. It is that, when the punctuation (usually full stops) is from the text I'm quoting it goes inside the quote marks---it's being quoted---and when it's my punctuation it goes outside the quote marks. The stylistically normal method is to put all punctuation inside the quote marks.

I was about to give in and re-format all my quotations to use the conventional style. However, I then found an interesting article by Larry Trask from 1997, then in Informatics at Sussex. He describes exactly the system I've used and describes it as the "logical view", opposing it to the "conventional view", and giving some good examples. He also cites an article by Geoffrey Pullum called Punctuation and human freedom (Natural Language and Linguistic Theory 2:4, Nov. 1984, pp. 419--425) in which Pullum describes his use of the logical view and his annoyance at copy editors insisting on re-shuffling his carefully quoted punctuation.

As a result my logical punctuation stays. That is how it will appear in the final text deposited in the library. It's a small victory.