Steven's Weblog

I recently upgraded a machine to Ubuntu 10.04, which has been an overall positive experience. But, the one bad thing was that I could no longer launch links from my emails or RSS feeds (using Thunderbird and RSSOwl, respectively).

In searching for the answer, I used
strace -f >trace.txt 2>&1
to try to see what was going on. I found that somehow, something was trying to launch firefox-3.5, which no longer exists (3.5 having taken over as the default version now). Nothing in /etc/ was referring to this… Some searching on the web led me to learn about xdg-open, which calls gnome-open, which uses settings in GConf to figure out how to launch a URL. Launching gconf-editor and finding ‘/desktop/gnome/url-handlers/http’ and ‘/desktop/gnome/url-handlers/https’, I saw that the command for each was ‘firefox-3.5’. Changed those back to ‘firefox’ and everything was cool again.

Now that I see that, I think I used gconf-editor to set those the other way when I first started using firefox-3.5. So it’s sorta my fault, although it wouldn’t be a terrible thing if the install scripts looked for and repaired that sort of error.

Occasionally I get on a little anagram kick. Last time, I wrote a little program that let you input a starting text, then interactively create your anagram text. As you typed the anagram text, it would complain if you used a letter that wasn’t available, or suggest words from the dictionary that were available to you. I made the search pretty fast by putting each word in the dictionary into a sort of radix tree, where both internal nodes and leaves had lists of words that could be formed with the letters used thus far on that path.

This time around, I wanted a bit more of an automated approach. I haven’t seen any great results yet, but here are some that are at least sorta evocative:
“The Miss Rhode Island pageant” < => “The time and rags and polishes”
“federal constitution” < => “failure to discontent”
“that girl with sunbonnet eyes” < => “but only the greatness within”
“college teaching is almost a” < => “glance at the seismological”
“a cut over his left eyebrow” < => “over by the sluice of water”

The approach I took this time was to start with an English corpus (well, a couple, appended, totaling 3M words). For each ‘phrase’ (sequence of consecutive words, really) of between 3 and 6 words, I stick the phrase in a list in a hashmap, where the key is the sorted list of letters used in the phrase. After eating through the whole corpus (which bloats the Python hashmap to about 4GB of RAM), I look for any lists of length longer than one, sort and filter the results to be a little more useful, and spew out the results. Even though the input comes from a nominally grammatical source, there’s still plenty of dumb junk, so it still requires a human to pick out the few good answers, and maybe do some word reordering. And even after that, you end up with what you see above, so, ya know, why bother? Just cuz.

I’m ‘playing dead’, waiting for them to cut the rope off their bumper and peel away. This sort of outcome is really a drag, but I still like to see it out. I think it gives them a feeling of power. That’s close enough to success, for me. I hear my show described as ‘passive recruitment’. No. I mean, definitely, if someone wants to become a zombie, I’ll give ’em a bite, but what I’m really trying to do is convey the ineffable sense of freedom that we feel on this side. Or give just ’em a little thrill. Whatever.

I don’t like link shorteners. Let’s not use them.

I love bit rot. Well, OK, maybe I don’t love that it happens (then again, maybe I do). I guess I love ‘bit rot’. It seems to be a farcical analogy at first glance, but there’s some depth to it, and that somehow adds a certain sort of liveliness to software, in that it’s subject to decay.

For those who aren’t aware of the term, Wikipedia gives a decent definition, in the “Problems with software” section. I might say it this way: if the world around it changes but your code doesn’t, it’s probably rotting. That violates our deepest wish that software could be written once and left to run forever of its own accord, but it’s simply a fact of life, or at least a good way to accumulate billable hours.

When I get a time machine, I’m gonna go back to the 1600s. I’m going to gather together all the texts, blueprints, plans, code, etc. I need to get people started working on reproducing the following technologies:

• electric generators
• digital computers
• rockets
• satellites
• radio transmission
• LCD monitors
• video cameras

and, ya know, whatever other infrastructure is needed to effectively support those. When the inevitable question arises about motivation for pursuing these projects, I’ll explain to them that when this is all knit together, everyone in the known world will be able to sit in their homes and watch a bunch of strangers compete to lose the most weight.

Sometimes sheer boredom, not a virus, spurs the dead into action. Dave had been dead for about 8 months when he decided to pop out for a little walk and chat with friends. Of course, in that time, he’d changed, and his friends had changed. Had they been ready to accept him back into their lives, even his closest friends couldn’t get past their prejudices, and they reverted to that old zombie script, wherein they insisted on killing him again. Anyway, his fresh perspective led him to realize that death wasn’t really any more boring than his life had been.

Just bought a Brother MFC-9320CW. It’s good so far. I use Ubuntu 9.10 for almost everything, and Brother actually supplies Linux drivers and source for both the printer and the scanner. The printer driver is just a PPD, but the scanner driver has some code-y components. But they provide a scanner .deb, even for 64-bit. In addition, the scanner is smart enough to be able to scan directly to a CIFS share, so if there are ever problems with the scanner driver, there’s still a decent way to scan.

The onboard web-based configuration software is pretty lackluster, but it does seem to do the job if you know what you’re doing and/or are curious enough (like, if you want to set up the CIFS parameters for scanning, don’t look under ‘Network Configuration’, and don’t try to find a ‘Scanner Settings’ like the ‘Fax Settings’ and ‘Printer Settings’… You have to find ‘Administrator Settings’ and the subpage ‘FTP/Network Scan Settings’ then ‘FTP/Network Scan Profile’.).

Since it’s all network-based and the onboard computer seems to be pretty smart, I can share it with all my computers easily and with Mac or Windows when I happen to be in those sort of modes.

The hardware is also good; the automatic document feeder is a big reason I got this model, and it seems to be robust and fast. Prints look very nice. I like the fact that it’s a toner system rather than an ink system, because I use my printer so seldom that inks would dry out on me between uses (though I haven’t used an inkjet in about 5 years, maybe they’re better now).

I’m not much for writing comprehensive reviews, but I just wanted to make a note for those who use Linux and were looking at this model. I’ll answer questions should any such person come around…

Have I mentioned before how much I love libraries? If there’s a single scene that summarizes me as a person, it’s me walking a couple miles to a library to read about some stuff I don’t need to know.

I love the insider’s view of things. Even if I don’t want to be an insider in a particular field, I really like to immerse myself in their views and language, when I have the time to, rather than sticking to the strictly outsider accounts (popularizations). So a library (especially a university library) is a perfect place for me: it’s mostly full of insider accounts of fields and topics within those fields. If I had to go out and spend $50-$100 per technical book, I wouldn’t be able to frequently dip my nose in a field for a few dozen or few hundred pages and then put it aside again.

I ran across an unexpectedly interesting book in the ISU library today. I did a search for “domain specific language” in the catalog. Since their computing collection is a little slow to catch up to the outside world, they didn’t have any of the several books on domain-specific computer languages published in the last few years. But they did have one on “Domain-specific English”, which turns out to be pretty cool for reasons that I won’t really go into too deeply here. Go find it at your library if you want to know more about that…

There’s a bit in here that was particularly relevant to me:

I remember one winter afternoon, mothers chirping in a corner of the playground as they waited for the bell to announce the schoolday’s end. Their talk focussed on homework – a subject of daily comments – but the conversation sounded more lively than usual. The topic of discussion, and of much stigmatizing, was the reading assignment for that day. Since my own child was in the class, I was aware that the purpose of the assignment had been to encourage children to read for global comprehension and to guess what unfamiliar words could mean. To my surprise, most mothers had missed this point entirely and were indignant. In their opinion, the assignment featured too many difficult words. Some complained about having to consult the dictionary to help clarify the precise meaning of the new words: wasn’t it too early to face such vocabulary? How surprising from an experienced teacher like our maestra!

(And yeah, I know it’s pretty meta to be quoting that in this context.)

I was lucky enough to have parents that would have been among the dissenters in that conversation. They never acted like or told me that some particular text might be ‘too hard’ for me, and they never would have recoiled in shock that some assignment had caused me to use a dictionary. I don’t know that they ever positively subscribed to a pedagogic theory that included the idea of reading for global comprehension and guessing, but the fact that they didn’t discourage it allowed me to develop my own unconscious theory of learning along those lines. If I had to guess, I’d say that children tend to want to explore the world according to those principles, and only discouraging them would stop that short.

Anyway, what was I saying? Oh, yeah: I had fun at the library today. That’s all.

This article from ‘MicroISV on a shoestring’ (which is a blog worth keeping up with, if you ask me) has an interesting insight. Patrick says that ad-supported sites are less likely to craft a good user experience, because if the site is better than the ads, the site doesn’t make money. Oversimplified, sure, but it does point out one of the forces at play in the overall dynamic.

I guess that’s a variation on something I’ve been thinking about, that I should get paid for the value I bring, rather than for the value that my advertisers bring. Or my ability to get people to look at little come-ons that give them the perception that there is some value elsewhere, to be more precise.

‘course, setting aside the absurd logic and perverse incentives, I’m glad that somehow ads occasionally work out such that people make good content and also make money from it.