Mikeweb


mike@crux:~$ host emergency.utexas.edu
emergency.utexas.edu is an alias for metro.gatech.edu.
metro.gatech.edu has address 130.207.244.30
emergency.utexas.edu is an alias for metro.gatech.edu.
emergency.utexas.edu is an alias for metro.gatech.edu.


How very odd. Thanks to MrBeige for pointing this out.

This is more for my information than anything else…

To make it so a specific program (say, s2ram) doesn’t require a password when use with sudo, do this:


$ sudo visudo


Go to the end of file, add


(username) ALL=NOPASSWD: /usr/sbin/program


I find myself using the word “obvious” far too much. If something truly is obvious, then why does the fact that it’s obvious need to be pointed out? Furthermore, why does the fact itself need to be pointed out in the first place if it’s truly obvious? Shouldn’t everyone just see it without much thought?

The word is especially misused in physics and mathematics. Many times, it is stated that some mathematical property is “obvious,” when if you were to mention it to a layperson, he would not understand what was just said. Stating that even the most trivial concept is “obvious” can be belittling if it is not understood immediately, even if it is just a comment in passing about something relatively unimportant.

The only legitimate context that I can think of in which it is acceptable to use the word obvious is when stating that a particular fact is “not obvious.” It happens all the time in math and physics that a result that on the surface seems intuitive (whatever that means) is in fact considerably more involved to rigorously prove. It is far easier to state that a particular fact is “not obvious” as opposed to “it takes considerable mathematical insight and finesse in order to show that this is the case.”

So please, stop saying that things are obvious. If you catch me saying it, bring it to my attention.

So I’ve been a grad student at Georgia Tech for officially two weeks now. The course load for the first semester is pretty heavy: Three “real” classes, plus math methods. (Math methods is a pass/fail course without any tests, and homeworks regularly take under half an hour to complete, so it is not incredibly stressful to be taking it.) Those “real” classes are: Quantum mechanics, classical mechanics, and electromagnetism. The first and last are two-semester courses, while classical mechanics is just one semester (next semester it will be replaced by statistical mechanics).

At some point, I’ll get around to posting more pictures of Atlanta and the Georgia Tech campus, but your guess is as good as mine as to when I’ll actually do that. Until then…

Over on Planet Debian, Biella posted a programmer hierarchy diagram.

Absolutely hilarious. My favorite part is “People who refuse to use the word Ajax.” Oh, and Lisp programmers thinking they’re better than everyone else.

Yeah, you heard that right.

Link.

I really don’t know what to make of this, but when I saw it, I couldn’t help but laugh. What’s even funnier is that there’s a thread on one of the Ubuntu mailing lists that calls it “offensive.”

I started experimenting with Markdown yesterday. Basically, it’s a simplified syntax that can be easily translated into XHTML-compliant markup. It’s nice in that it allows you to keep your source files incredibly easy to read. While HTML isn’t really that bad to look at, it is still nice to be able to do something like `code` instead of the normal <code> tags. The only problem is that the Markdown script won’t correctly process things like <title> tags and other things such as meta tags and so forth.

So, I made a little Python script that will take the Markdown text file and convert it into a fully compliant XHTML file along with appropriate title and metadata. It is available here. For an example of before and after, look at the source of the mdscript.html file.

For some time now, I have been somewhat unsatisfied with Firefox. The main reason for this was really memory management issues — it always seemed as if Firefox was taking up way more memory than it should need*. It also seemed that the Windows version was often somewhat better performing than the Linux version, which to this date is the only piece of software that runs natively on both OSes that I have been able to say that for.

I had tried both Galeon and Epiphany in the past, but was not incredibly thrilled by either. In Galeon’s case, when I had originally tried it (many, many years ago), it seemed to also be plauged somewhat with memory issues. In Epiphany’s case, when I first tried it several months to a year or so ago, it just didn’t feel right to me… There were several things that were not configured to my liking, and the incredibly sparse preferences dialog didn’t make it any easier to adjust it to the way I wanted.

Well, I decided to give Galeon another chance since David (a.k.a. MrGreen) has been using it for quite some time and is very pleased with it. To make a long story shorter, I came to like Galeon a lot more than I did in versions of old. The “smart bookmarks” feature is wonderful (though using Yubnub in Firefox’s search bar effectively serves the same purpose), and being able to add several nicknames for the same smart bookmark (e.g., g, google, and google all short for searching Google) is quite nice. However, I noticed that Galeon did not seem very apt at switching tabs quickly — there would be a slight delay (slight, but significant enough to be noticeable every time) between clicking on a tab and having that page redrawn. I also was pointed to the fact that Galeon is essentially finished in favor of Epiphany as far as development goes.

Given this, why not try Epiphany again? After all, I was happy with Galeon this time around, so maybe Epiphany also had improved. As it turns out, I rather like it this time around. There are still a couple of minor issues with it in terms of my not liking the sparse preferences dialog, but as a whole, it is quite elegant in its simplicity. I found an unofficial extensions page that fixed up two issues I had: Being able to use C-tab and C-shift-tab to navigate tabs instead of C-PageUp/PageDown, and being able to middle click on a tab to close it**. The only issue that I still have is that I can’t use keywords completely integrated with the bookmarks editor just yet. For now, I’m using this python extension to tide me over. It’s not perfect, but it works.

So, for now, I think I’ll play with Epiphany for a while. The one major advantage that Firefox has over it still are the available extensions, but I only use a small handful of them from day to day anyway, so that is not necessarily a showstopper.

* In the Firefox developers’ defense, these issues seemed to have largely been fixed in the 1.5.x series, and performance went up significantly under Linux.

** This is, of course, a very minor issue since each tab by default has the ‘X’ on it to close it. But, since I’m used to the functionality of Firefox, I felt it couldn’t hurt to use this extension until I’m weaned off from it.

I wrote previously how I have started using Sawfish as my window manager of choice. I also decided, at least as far as my laptop goes, to run Sawfish by itself, without the undue burden of also running Gnome. The main reason for this is speed; it just takes too long for an impatient person such as myself to start Gnome (at least the first time — GDM always seems to start up before a lot of other services, so there’s a lot of background work being done by the CPU when I try to login for the first time), though I also like the simplicity of a clean interface without two panels that take up a lot of space.

However, one thing that I really like about Gnome is the system tray in the panel. This is especially nice for programs such as Gaim, which I can hide so as to both not take up space and to not distract me (additionally, it’s nice to be able to just have to right click on the icon and select “Mute Sounds” from time to time rather than having to go all the way through the preferences dialog). I found a gDesklets applet that was just a system tray, but I had several issues with it, aside from having had problems in the past with gDesklets as a whole. So this morning I started to look for information about making my own standalone applet that would do exactly what I want without getting in the way. Low and behold, I find that it already exists.

This is (yet another reason) why I like open source: Every time I set out to make a small project to make things better, I find out it already exists, thus saving me time.

I guess the moral of the story is that no matter what you want to do, in the open source world, Someone Else wants it too. Quite often in my case, this proverbial Someone Else gets to it before me, which saves me the hassle of having to do it.

I recently setup a Planet feed aggregator so that I can easily keep up with the RSS/XML/whatever-enabled sites that I frequent every day. For a few weeks, I was using feed2imap, which delivers new updates directly to one’s IMAP mailboxes. However, this was a bit annoying, particularly if a blogpost had a lot of pictures (or even worse, if I’m trying to check out the updates to a webcomic), since I would have to manually click on each link that it generated in its place. I also found it just generally hard to actually be able to determine whether or not something was worth reading from the subject (oftentimes it’s easier to be able to skim through it first to determine if it’s interesting or not).

So, I setup Planet (at a location that you can probably guess if you’re sly enough — I just don’t want to get too much traffic to it, so I’m not directly stating the link here). This is the same system that Planet Debian, et al. use. Now I can read all of my feeds from anywhere (no need for an RSS aggregator that is local to one and only one location), I can skim through entries more easily, and there is no need to click on image links. On top of all that, it was incredibly easy to setup.

I recommend it to anyone with a webserver and can run cron jobs to update it every n minutes.