aghitza

This blog is moving…

aghitza — Mon, 15 Feb 2010 10:08:30 +0000

… to the following address:

Fix for annoying repository problem in Archlinux

aghitza — Fri, 22 Jan 2010 21:13:33 +0000

I have been periodically getting into a situation where my Archlinux installation would not see some repositories, notably extra and community. I would try pacman -Syu and it would pretend that everything is just fine and up-to-date, but searching for a package I know is in extra would return no hits. I have no idea what is causing this problem, but I finally managed to find a way to get Archlinux to see these repositories again. This is as good a place as any to document this, and it might help out somebody else having the same problem.

It’s pretty simple once you know where to look: as root, go to /var/lib/pacman/sync. You should see directories for the various repositories you get packages from. If you have the problem described above, then the corresponding directories will be empty, except for the hidden file .lastupdate. Just delete this file (in my case, from both extra/ and community/) and the next pacman -Syu will make everything right again.

Of course, I’d appreciate it if someone would tell me exactly what causes these problems in the first place, and how to avoid them instead of fixing them when they occur. Some of the trouble might be related to my (mis-)using various wrappers for pacman, such as yaourt or powerpill or bauerbill (but I don’t want to blame this on them since I honestly don’t know what’s wrong).

Tagged: archlinux

Lessig on copyright wars, and science

aghitza — Mon, 18 Jan 2010 12:43:59 +0000

Lawrence Lessig was awarded an honorary doctorate by the University of Amsterdam, on the University’s 378th birthday. I have just watched his acceptance speech, which is a very compelling description of the ways in which copyright law is failing to adapt to the challenges and opportunities of the digital age. Blip.tv makes available the video of the speech itself, as well as the slides with synchronised voice. These 52 minutes were definitely well-spent.

Python script for getting the number of new messages in an mbox mail folder

aghitza — Sat, 24 Oct 2009 05:17:52 +0000

I have been tweaking my email setup and got to the point where I needed a script that would tell me the number of new messages in an mbox-style mail folder. I have recently switched to using mutt as a mail reader, and mutt marks read messages by inserting the header “Status: RO”. Therefore all I had to do was to check, for each message, whether this header was present or not. Fairly short and simple; here’s the Python code:

#!/usr/bin/env python
"""
This counts the number of new (unread) messages in an mbox-style
mailbox.
"""

import mailbox
import sys

def main ():
    mb_name = sys.argv[1]
    mb = mailbox.PortableUnixMailbox(file(mb_name, 'r'))
    msg = mb.next()
    new_count = 0
    while msg is not None:
        status = msg.getaddr('Status')
        if status[0] is None:
            new_count += 1
        msg = mb.next()
    print new_count

if __name__ == '__main__':
    main ()

Tagged: python

Terry Tao in Melbourne: Mathematical research and the Internet

aghitza — Mon, 31 Aug 2009 13:35:23 +0000

This evening I attended a public lecture by Terry Tao, who is currently on a mathematical tour of Australia. He spoke about how the rapid development of computers and the Internet is changing the way mathematical research is done. It is an interesting topic, and it was a good talk. I won’t go into a long description of the talk itself, since Terry made the slides available on his blog. The lecture was also videotaped, so it is possible that the video will eventually be available from the Faculty of Science or the Department of Mathematics and Statistics. (I will post a link to it here when/if it becomes available.)

There was time for two questions at the end of the talk, and both of them were about common fears: “Since preprint servers and blogs are not peer-reviewed, and therefore less trustworthy, is it not dangerous to ditch the traditional publication model in favour of the online tools?” and “If the internet is becoming such a great medium for learning, collaborating, and disseminating knowledge, is there still a place for universities as physical entities?” Terry’s replies to these questions were similar: he is not advocating that modern technology and approaches replace the established methods completely. Peer-review and formal publication have an important role to play in the mathematical research ecosystem, and rapid-dissemination tools such as wikis and blogs are meant to complement rather than replace them. Similarly, there is great value to being physically present at a lecture or seminar, and to spending a week locked in your office with your collaborator.

Let me try to formulate two other issues that occurred to me during and after the talk, and which indicate to me that the technological changes need to be followed by cultural changes in the way academia functions.

One issue is that of professional recognition. Writing a detailed Wikipedia entry on a specialised subject, or a blog post popularising recent developments — these are great contributions to the mathematical community, they take time and effort, and they often require expertise that few people have. These contributions should somehow qualify as scholarly work in the new age of mathematical research. Are they taken into account by promotion and tenure committees? It would be interesting to know more about this.

The other issue is not really about mathematical research, but it stems from the increasing availability of online mathematical material of good quality. How does one assign homework today, given that a Google search is more and more likely to provide solutions? This is by no means a new problem — it’s been around ever since it became possible to post solutions on the web. From a lecturer’s point of view, recommended “best practices” included putting a special directive for search engine to not index such files, or creating pdf files without searchable text, or using a closed course management system only available to the students in the particular class. It seems that all these measures are only delaying the inevitable. But what can be done? It is upsetting to not be able to assign any homework questions from, say, Hartshorne or Atiyah and Macdonald. (Or, for that matter, any question to which Wolfram Alpha might know the answer.)

Terry finished his talk by saying that there are many new gadgets and technologies that have a bearing on today’s mathematics; and that it is hard to predict such things, but that some variants of these are likely to survive beyond the lifespan of simple fads and become as ubiquitous as email and LaTeX are today. I agree, and would moreover argue that the change in methods and approaches to mathematical research needs to find its reflection in changes in the perceptions and culture of the subject.

Tagged: math, unimelb

new keyboard!

aghitza — Fri, 28 Aug 2009 04:11:07 +0000

After a long period of indecision, I finally gave in and ordered a new keyboard: it’s a Kinesis Advantage (not cheap by any stretch of the imagination, not to mention that the cost of shipping it to Melbourne via UPS was rather high). But it’s now here, so it’s time to get used to it and see if it’s worth its price.

So far I’m pretty happy with it. I’m still in the transition stage. If you’re wondering what in the world I need to transition to (after all, it’s just a keyboard, how hard can it be?), have a look at this:

The picture is not that great but it gives you the gist of things. Luckily for me, I touch-type so I’m having almost no problem typing text on this thing, as that part of the layout is just standard. Getting used to the new spots for Backspace, Enter, and Space also seems to go down fine (they’re meant to be typed with your thumbs). The most difficult at the moment is the Ctrl and Alt situation, and that’s a bit of a problem since I use these quite a bit in Emacs, for instance. They seem a bit out of the way…

The other problem is the Windows key. I haven’t used Windows in at least 10 years, but I’ve made the Windows key the main modifier for my window manager actions (I use wmii, more on that in another post, maybe). Combine this with the arrow keys being in really weird places, and my navigation around wmii is seriously messed up now.

My last gripe (for now): I really wish this keyboard had a built-in touchpad, like the Adesso keyboard I used previously (and now stolen by my wife). Having a small touchpad fit between the two small groups of keys at the bottom of the keyboard would be just great. At the moment I have a Cirque Smart Cat touchpad around, but it doesn’t fit in that space, and having it above those keys doesn’t seem optimal.

But all in all, I’m pretty happy with this buy. Typing feels good, I’ve already made us of the key remapping mechanism to make a few changes, and I’m sure I’ll think of something for putting the macro capabilities to good use. One thing that I haven’t had a chance to try extensively yet is writing LaTeX, or (say) Python code. We’ll see how that goes, since some of the symbols are in weird spots.

Tagged: gadgets

hello world

aghitza — Fri, 14 Aug 2009 03:05:43 +0000

Yet another blog. What can you expect to see here? Most likely it will be a mixture of mathematics at Melbourne and elsewhere, Sage stuff, and the occasional Linux and open source-related post.