Older blog entries for jfleck (starting at number 275)

Shane Mueller has done a clever bit of usability analysis for Nautilus and GNOME. Shane mantains the Nautilus File Manager Scripts project, a collection of scripts folks have written to add functionality to Nautilus. Shane reasons, properly, that the various itches being scratched by Nautilus scripts could reasonably viewed as gaps in Nautilus/GNOME functionality, and he has provided download numbers so we can see what people are most often trying to add. Archiving and compression seems to be the big winner (loser?). He is quick to admit that it's a crude measure, and I would add that the sample is biased toward geeks who are into the sort of tweakly customization a Nautilus script implies. But it's nevertheless a clever and valuable insight into how people are using our software.

An example of the correctness of raph's observation about the vibrancy of the independent media on the web: Steve Aftergood's Secrecy News. A counter-example (linked from Aftergood's newsletter): David Corn in The Nation. Old paper-based independent media that happens to be easy to find for free now on the Web.

Spent a good chunk of the afternoon just reading (I love my job) a couple of fascinating pieces.

One was the latest from the ubiquitious Malcolm Gladwell. Jeff Jones, a colleague who hangs around with cops, showed it to me at lunch, and it was too sweet to pass up. It's called The Naked Face, an extraordinary look at the way our faces are wired to communicate.

Reading the second is from another amazingly diverse and ubiquitous writer, Jared Diamond. It's a piece in the 8 Aug. Nature about the role of evolution in domestication of crops and livestock and the unanswered questions about when and how the shift from hunter-gatherer to agricultural cultures was made. A fun quote:

Wild wheats and barley bear their seeds on top of a stalk that spontaneously shatters, dropping the seeds to the ground where they can germinate (but where they also become difficult for humans to gather). An occasional single-gene mutation that prevents shattering is lethal in the wild (because the seeds fail to drop), but conveniently concentrates the seeds for human gatherers. Once people started harvesting those wild cereal seeds, bringing them back to camp, accidentally spilling some, and eventually planting others, seeds with a non-shattering mutation became unconsciously selected for rather than against.

Or this, on the co-evolution of humans to their newfound agricultural treats:

The evolution of allozymes of alcohol metabolism (permitted) consumption of large quantities of nutritionally important beer in western Eurasia...

Hmmm. I think "Nutritionally Important Beer" would be a great name for a band. (A link to the piece, but it's a paid subscription thing. If you're at a university with a subscription, though, check it out.)

A testament to Diamond's enduring popularity - all 11 copies of his "Guns, Germs and Steel" are currently checked out of my local library.

rillian makes a good point:

To be fair, it's not that black and white. Another journalist can actually go look for sources to confirm or deny anonymous quotes. As you imply it slows things down, but to call the situation a 'swamp with no way out' makes sense only if you do all your research on the internet.

I am, in fact, guilty of rhetorical excess. There is indeed a way out of the swamp. In the case where the anonymous source's allegation is true, other anonymous sources will say the same thing, and the story will converge on a truth that will be helpful to the reader. But if the sources remain anonymous, how is the reader to know whether you're really converging on truth, or simply hearing again and again from the same source? And when you have a named source, readers have an opportunity to evaluate for themselves the motivation of the speaker. "He's a Democrat! Of course he would say such things about Republicans!." And what of the case where anonymous leakers disagree?

The case of Wen Ho Lee, the Taiwanese-American who was implicitly accused by anonymous leaks of passing nuclear weapons secrets to China, is a case in point. The main accuser, given voice primarily by the New York Times remained anonymous through months of news media leaks, painting a picture of a dastardly Lee and his Chinese cohorts. The Washington Post, the Los Angeles Times had its leakers who told a somewhat different story. We now know, in retrospect, that the "somewhat different story" told by the Post, the Times (and my colleague Ian Hoffman (great book - spend money on it) was correct, that whatever Lee's transgressions may have been, spying for China wasn't one of them. We now know that the Times, which largely set the national agenda for the story, was depending on a small cabal within the intelligence community that held extreme views about China, and were quite possibly racist in their scapegoating. The Times and its sources whipped up a political frenzy, a frenzy that we now know had little basis in fact. A reliance on named rather than unnamed sources would have allowed the truth to be sorted out much more quickly, and a great deal of damage to individuals and institutions could have been avoided.

But the place it is most corrosive is not in the extreme cases, like Steven Hatfill or Wen Ho Lee, but the day-to-day business of government. It is not uncommon today for a senior government official to hold a formal briefing for reporters on some policy initiative under condition of anonymity, to be quoted in the newspaper the next day as "a senior government official". It is not uncommon for a reporter to call a federal agency asking for an explanation of the agency's position on a matter of public policy and be told that the only way the agency will speak in detail is under condition of anonymity. That allows great room for backsliding and mind-changing, and eliminates any accountability.

I have no problem with the use of anonymous sources shedding light on malfeasance within the government agencies for which they work. My problem is with anonymous sources using the shield simply to avoid accountability for their words. I also have a problem with reporters willing to grant them that shield.

raph: I believe the cases in which anonymous sources should be permitted are a very small percentage of the cases in which they're now used. They have become a corrosive habit in the national mainstream media in this country, a cozy collusion between source and journalist that should not be tolerated.

I think of it as an epistemological problem, and liken it by a very close analogy to the standards of scientific publishing. The epistemological problem is this: I think it's important to tag everything we know with an attribute - how do I know this? It's the things we think we know that lack this attribute that get us into trouble. In scientific publishing, this is analogous to the question of reproducibility. Scientists publishing research results must explain their work in such a way that we can understand the details of how they achieved their result and, if we so choose, reproduce it. Thus we can identify how we know this "fact", and independent peer review can either confirm or refute it.

A journalist using a named source, or citing a public document that others can obtain and independently review, is doing this same thing, and the competitive crucible of journalism is generally swift and sure. But an anonymous source offers no such opportunity, leading us into an epistomological swamp with no way out, no way to tell how we really know the thing being alleged, no way to check it out.

That said, however, I don't expect much to change. It's a marketplace of ideas (a very commercial marketplace), and consumers seem to like the idea that they're getting the inside dope from Newsweek's "highly placed sources". (You see, Raph, I despair of the state of U.S. maintstream media, but for very different reasons.)

raph: Beward the "seemingly fair, accurate article" that is based entirely on anonymous sources. They inevitably have some axe to grind, and the shield of anonymity means they have no accountability for what they say. In the case of the Newsweek article you cite, their particular axe seems a desire to create the perception that investigators are really doing something, rather than sitting on their hands. This requires a scapegoat, and Hatfill may very well be their victim. This is what we saw in the case of Wen Ho Lee, when anonymous sources for months suggested he passed sensitive design details of the W88 nuclear warhead to China. We now know that was false, but the anonymous sources have nothing to answer for because they remain anonymous. Likewise Richard Jewell, the "suspect" in the Atlanta Olympics bombing case. I have no opinion about Hatfill's innoncence or guilt, I merely suggest caution at the potential for self-serving government "leaks". Sometimes the most seemingly "independent" media filled with breathless quotes from anomymous sources is really the least independent of government influence.

Spam sucks, but so do the solutions.

My ISP, Southwest Cyberport, (these guys rock if you happen to be in New Mexico and are inclined, like I am, do deal with a local ISP), installed SpamAssassin so customers could use it if, if they desire, to filter their mail. I've been experimenting with it in recent weeks, and it's been greatly reducing the dreck arriving in my inbox. But it comes at a price - sometimes you lose real mail. Which means I have to check the Spam File every few days just to make sure nothing important gets lost. So far, the only real mail to get shitcanned has been the html version of the Albuquerque Journal (my employer's) daily news email. For some reason, the text version gets through fine, but the html version has some crap at the top about our personal ads. At least that's my best guess about what's triggering SpamAssassin.

Today, though, it caught the latest draft of the GDP Handbook, which drake had sent me to put on Inkstain so people could read it. Thankfully I checked the spam file this afternoon and was able to retrieve it and post it. I'm not sure why SpamAssassin didn't like it - perhaps it was the bit about how writing user docs improves your sexual prowess.

It's a pretty funny solution to a pretty massive problem - how do you cost-effectively move around the terabytes of data being generated by the Sloan Digital Sky Survey? Jim Gray's very practical solution: TeraScale SneakerNet. Turns out it's cheaper to just ship a whole computer.

"Why do stupid people not know that they're stupid?"

Spent some time over the last few days wrestling with ScrollKeeper, the document "card catalog". We're using it for GNOME docs, and others are adopting it as well, including KDE and the Linux Documentation Project. Its aim is to help organize the welter of docs that get installed on a Unix box by getting everything registered in a central repository that includes metadata about each doc. It uses the Open Source Metadata Format (OMF), a derivative of the library world's Dublin Core. An arbitrary help browser can simply query ScrollKeeper and know about all registered documents, even those installed in weird-ass places.

I've just been using it blindly, without paying much intention to its internals, but began running into problems over the last couple of weeks. The problems are of my own making - working on GNOME2 development, I have several different versions of GNOME2 installed in different sandboxes, and ScrollKeeper was getting, for lack of a better word, "confused". I've not thoroughly sorted it out yet, but did make some useful progress in understanding SK's internals, which is, I think, something I need to know because we're depending on it so much.

Finished Sam Abt's Breakaway, about the 1984 Tour de France. A quick read picked up at the library by accident, while I was looking for books about the physiology of cycling. Abt is sort of the dean of English-language cycling writers, and the book is good in a number of ways. Perhaps the most important is that he doesn't slip into the trap of hero worship of these guys. He shows what great cyclists they are, but also what assholes they can be.

Libraries are a wonder, with the serendipity of the Dewey decimal system.

enigma: There's a group here (Albuquerque, New Mexico, USA) that is putting FreeBSD and Linux computers in a school, but more importantly volunteering to work with the kids to learn how to use them. It's a great project in part because they picked one of the poorest schools in town, working with kids who are bright but don't have the luxury of computers at home. They call themselves Meta-Learn.

raph: I don't disagree with any of what you're saying about structural versus presentational formats, except that I believe authors are best served by thinking primarily about content. Perhaps this is because I have no facility for graphical presentation, and prefer instead to focus on the aggregation of words into ideas. But I believe that the proper structural markup, combined with the proper toolkit, can handle the presentation as a back-end activity that frees the writer up to think about the words. I may be limited here in my own understanding of the issue because I've never used graphical presentation to carry semiotic content (am I using that word right?). If I was more graphically oriented in the way I use writing to communicate, I might get more stuck the in gray area between presentation and structure that you describe. (If I was skilled at the graphical side of communication, I also could write television commercials and make a boat load of money.)

Consider the table. That is an example where presentation is critical to content. The structural markup languages like DocBook and the one developed in my industry, the News Industry Text Format have explicitly included it. So the gray area is inevitable. But I continue to believe that presentation should be extracted into a separate process, using a structural markup to create content and then a variety of transformation tools to present it. I think one of the reasons people insist on sending me the MSWord versions of their news releases at work is because they spent too much time making it pretty, when all I care about is the words. Send me the words!