Older blog entries for jfleck (starting at number 296)

the canyon
There is a close parallel for me in standing before Nude Descending a Staircase and the Grand Canyon of the Colorado River. It's a sort of gut-clunching, teary feeling, as if I am looking upon something importantly beautiful but slightly beyond me, something difficult to understand. Great feeling, to be repeated whenever possible. In other words, mom and dad and I had a great trip to the Grand Canyon.

Like icing on the limestone layer cake (he said, brutally mixing his metaphors) I saw a California condor, so large and very endangered but serene sitting on a tawny outcrop just below the rim with no understanding of his special place in the world.
crowd size
telsa: Crowd size is a bugagoo for both cops and journalists. I did stories years ago about the size of the crowd at the Rose Parade in Pasadena, Calif., on New Year's Day. Every year, the cops would say there were a million people watching the parade, and I wondered how they knew. They told me they flew over in a helicopter and said, in essence, "Yup, looks like about a million people.". We did some elementary arithmetic, and concluded that it was physically impossible to fit a million people onto the sidewalks of Pasadena. There just wasn't room. So my colleagues and I made an attempt at rigorous crowd counts, measuring sections of street at various points along the route and counting the people and doing the math. The result was a crowd of about 300,000, give or take 50,000. We wrote the story for the newspaper gleefully. And now, every year, I read the wire stories out of Pasadena and every year the cops still say there are a million people at the Rose Parade. Oh well.

Last fall, I was covering an anti-war rally at the university here, and one of the organizers complained about the crowd size estimate we had run in the newspaper for a demonstration they'd held the week before. So when the group went on the march, I just counted them all. Didn't take too long, and I had a real number.
road trip
Taking off tomorrow for a long weekend at the Grand Canyon with Mom and Pop. I love the canyon, but this is most about spending some time with the folks.
error messages
For everyone who has ever bemoaned or mocked those terse and unhelpful error messages software spits out at us (and who hasn't?), there is this, from a talk by high performance computing pioneer Roger Lazarus. He's talking about SEAC, an old behemoth with just 1024 words of memory and an apparently unpredictable nature:
Los Alamos users also trained the cantankerous machine to print out a tell-tale warning in case of major errors: "CALL A CAB."

"When that happened, we'd go home and leave it to the engineers," Lazarus said.

no wimps here
Now that is a bike race.
I've spent a good part of the weekend in Nuclear Reactions, a book about U.S. nuclear waste policy. I've gotta get it done today - I've promised a book review, which I must turn in tomorrow. I was a bit afraid it would be turgid, but the review is a labor of love (written by an old friend on a topic I've covered a lot myself), so I'm duty-bound to finish it. But it has not at all been a chore. It is very well written, and sheds some very useful new light for me on the seemingly intractable problem of nuclear waste disposal (more specifically on what makes it intractable).
good news
" It won't be too long - that's all I'll say!"
I responded a couple of days a go to an appeal from Havoc Pennington for help cleaning out the bugzilla reports for the new GNOME terminal. It's full of misfiled reports from the old terminal, and duplicates, and such, which is pretty straightforward to take care of, but which Havoc didn't have the time to deal with, so I offered to help. It's given me a chance to take a much more detailed look at the seemingly simple terminal, which seems particularly well crafted.

Of special interest has been the repeated requests for different preferences/options. Havoc's very articulate on the issue of preferences. There's a tendency in many free software GUI apps to add preferences galore to deal with issues that come up, ignoring the fact that every preference comes at a cost, both in terms of usability (GUI complexity not being free in terms of usability) and in terms of code complexity and the resultant possibility of bugs. So Havoc fights off a lot of enhancement requests for this or that added preference, with the basic argument that there is a limited amount of real estate in the preferences dialogs and he is not willing to squander it.

bike racing

Jaime and I raced a front end loader yesterday morning.

We were coming out of Corrales, a little rural community on the edge of Albuquerque, riding the village's narrow, winding main road, when the front end loader saw his opening and tried to make his move. The loader attacked where the road widens a bit at the base of the hill out of town. I was ready to let him go. But when I saw that he faltered as the road tipped up, I saw our chance, and launched the counterattack. We played cat and mouse with the loader up the hill, riding furiously while a line of cars poked along behind our little train. Funny what a difference an engine makes. In the end, I dropped off the back, unable to match the loader's blistering pace, but Jaime launched a furious attack at the top of the hill and won the stage. He was grinning from ear to ear when I caught up with him.
Pat Costello sent me a draft of the Sun team's proposed updates to the GNOME Documentation Style Guide Friday, and I spent some quality time yesterday poring over it. It's terrific stuff. They've added a lot of good material, and its organization is better, making it more useful.
Spurred by Luis' effort to get more of us involved in triaging bugs in GNOME's bugzilla, I've spent a bunch of time this weekend sorting out Nautilus bugs. The Nautilus bugzilla has become a bit of a tangle - I'm not sure how to untangle it. Mostly, what I've been doing in recent months is closing obvious non-bugs and duplicates, and leaving the rest alone, but it's clear more than that is needed. Guess I'll give it a shot.


Nora's been out every evening the last week, working as a member of the "house crew" at the play at her school. She's greatly excited by the whole experience, so we are too, of course. But she's at an age where the presence of her parents is normally completely uncool. When I picked her up Friday night after the play, she suggested Lissa and I might enjoy the play. This was invitation in her way, so we took her up on it last night. Of course, she was much too busy to spend much time with us, but I was completely flattered that she was unembarassed enough to even have us there at all. Great play, too. Bonus: "Do you think Grandma and Grandpa would like it too?" So my parents joined us for our night on the town.

And then there was the after-play party. A bunch of kids not old enough to drive meant I ended up at the Frontier late last night, trying to somehow be a responsible adult while staying completely out of the kids' way, so as not to embarass them. Nora suggested I looked pretty pathetic sitting at the table in the corner by myself, eating fries and reading free liberal weekly newspapers. Whatever. It's new terrain.
you go, Buzz!
Somebody needs to start a Buzz Aldrin defense fund.
10 Sep 2002 (updated 10 Sep 2002 at 21:12 UTC) »
anonymous sources
Much discussion here a couple of weeks back about anonymous sources and the case of Steven Hatfill, the poor sap who has been publicly named by the FBI as a "person of interest" in the anthrax cases, and the reporting on the case, especially a torrid Newsweek account of bloodhounds going apeshit at Hatfill's apartment and a Denny's. I made the "peer review" argument, that anonymous sources makes peer review by other journalists damn near impossible. Now comes David Tell in the Weekly Standard with a terrific account of the Hatfill case, in particular the way the poor sap has been railroaded by government leaks. Tell's discussion of other reporters' followup on the bloodhound story is instructive. Lots of other media have repeated the story, quoting Newsweek. But no one who actually did their own reporting was able to confirm it. Proving a negative is hard. But Scott Shane of the Baltimore Sun has come the closest, showing as best he can that no such bloodhound search ever happened at a Denny's in Louisiana (as claimed by Newsweek) and that it is unlikely (according to the best experts he could find) that any scent would have been left on the anthrax envelopes to be of any use to bloodhounds. But here's the problem. Without knowing the identity of Newsweek's source(s), there is no way to challenge them directly about the claim. Here's where "peer review" breaks down in the anonymous source game. If a tale is true, other reporters will quickly be able to get their own anonymous source confirmation. But if it's false, it is much harder to use the same system to refute it. In the meantime, other media keep slavishly repeating the Newsweek bloodhound tale.
flame war
Got in a fun, if a bit time-consuming flame war on the Linux Documentation Project mailing list about SpamAssassin and the issues arising surrounding spam filters. A new correspondent noted that his list confirmation missive when he signed up was snagged by SpamAssassin. That was fine, but he then went on to suggest in a later post that it was somehow dmerrill's responsibility to tweak his mailman settings to avoid SpamAssassin's evil grasp. This is the post that sent me into flame drive.

As I explained here, the whole notion that senders should have to tweak their email so it is not caught by spam filters is abhorrent to me. It suggests that the spammers and, by extension, the spam filter authors have become defacto censors. The responsibility lies with spam filter authors building in the flexibility so users can tweak settings to avoid false positives. It then is incumbent on the user to do the tweaking.

I can die happy.

Sitting at my desk at 5:30 yesterday afternoon, Jamie walked up and handed me two tickets to sold out Lyle Lovett/Bonnie Raitt show. There was a moment when the lights went down and she started pickin' a quiet acoustic and there was this:

If dreams were thunder
and lightnin' was desire
this old house it would've burned down
a long time ago
I 'bout melted into my grass amphitheater cool late summer evening lightnin' in the distance lights of the city my baby by my side lawn chair. Man that girl still knows how to rock.

287 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!