Last night we decided to have a few friends over for dinner. Somehow we hit on Mexico as the culinary theme, it being a few days before cinco de mayo.

We made fajitas with chicken, red & green peppers, and sliced portobello mushrooms. I marinated the chicken and veggies (separately) in a mixture of finely chopped cilantro, garlic, lime juice, and a touch of olive oil. It’s best to let the chicken marinate for at least an hour: it will start to really break down from the acid, and it turns greenish from the cilantro. I grilled everything on a miniature one-burner grill pan. (Outdoor grill would be ideal, but we don’t have one yet.)

To go with it, I made a fresh pico de gallo with finely diced tomatoes, chopped cilantro, garlic, lime juice (again, the holy trinity), and red onion. Scallions probably would have worked well in place of the red onion. A can of corn would have gone well too, but we were out.

Although not particularly Mexican, we had an eggplant to use up, so I roasted it dry in a 500° oven (20 or 25 minutes), scooped out the innards into the food processor, and blended it with garlic, allspice, ‘Italian seasoning’ (fresh parsley would have been ideal), and a touch of olive oil to smooth it out. This made a nice spread for French bread or crackers. I enjoyed eating it still slightly warm, but I must admit the flavor intensified a bit after refrigerating overnight.

I believe that the best approach to cooking is to start with the freshest, highest quality ingredients you can find (or afford), and then don’t mess them up! Conveniences like pre-cut garlic in a jar or lime juice in a plastic bottle don’t seem worthwhile to me. What’s so time-consuming about squeezing a lime? (I did use packaged tortillas though… just heated them for a minute on the oven rack before serving.)
□
16:11  / Cooking § 1 comment

As I age it seems I’m becoming…shudder…a morning person! Of course, many people will think it’s crazy that I’m alarmed at being up at 7 on a Saturday. But I’ve fully embraced the “academic lifestyle” since I was 18: my time has always been flexible, and I used to do my best work at night. The last time I was obligated to attend anything before 10 was an 8AM physics lecture in my freshman year…and I skipped it rather often.

But now as I’ve made dreaded transitions in age, I often start getting sleepy at 22:00 (that’s 10PM for you folks that are bad at math). Fortunately, I’ve found that I can now be alert and do good work at 7 through 9 in the morning to compensate.

I recall a strange (and perhaps embarrassing) email I wrote in college about productivity and times of day. It turns out it was about this time of year and about this time of day… in 1992.

      From:   Christopher A League <league_c@jhunix>
      Date:   Wed, 15 Apr 1992 07:18:26 -0400
      Subject: why ask why?

      Why, you ask, do I stay up all night working on
      a program that isn't due yet?  I don't know.

      Maybe it's the thrill of knowing that I'm doing
      something productive while the rest of you clods
      are sleeping.  ...Nah.

      Maybe I love the "thrill of the chase" and i
      can't stop programming until i get something
      done.   ...Nah.

      Maybe I'm evil and i love staying up thru the
      "witching" hours with my hair flowing everywhere
      and a glazed look in my eyes.   ...Yup.

      I am the Lizard King.  I can do anything.
      --Christopher League

I’m on the hunt for sophisticated task management software for Linux & Mac. Lots of ‘productivity suites’ include a task component, but it is usually a dirt simple sequence of lines, each with a check box, and perhaps with some priority. The calendar or scheduling component of these systems is usually more complex, but my scheduling needs are very simple. In my life, sophisticated task management might help with the following types of questions: what should get my focus for the next 2 hours? When should I switch my attention from project A to project B? When should I start to worry about looming paper/ proposal/ bureaucratic deadlines—not all deadlines are equal in priority or in the amount of work needed to meet them.

I’ve used the paper Franklin Planner for many years. I still (mostly) like the model that Franklin uses, but for various reasons I don’t take as much advantage of it as I should. I’ve been thinking of taking the process online… mostly so that I can archive and search my records easily, not so I can look hip with a Palm Pilot. But anyway, I never found quite the right tools.

The most interesting approach to advanced task management that I’ve seen is embodied in the LifeBalance software for Mac, Win, and Palm OS. You specify a hierarchy of tasks with deadlines, relative priority, effort, lead times, and repetition. The software automatically orders your tasks to take an amazing number of factors into account. Tasks will float to the top as deadlines approach, but only to the extent that you’ve declared they are important to your life as a whole. The software takes into account the effort spent on completed tasks, so as to re-balance your priorities. Spent too much effort recently on teaching and service? LifeBalance will automatically give higher priority to your research tasks to compensate.

This seems like a really great model, but a few things stop me from using LifeBalance itself. First, it doesn’t run on Linux: I have only one Mac laptop that I don’t carry with me often, but I have Linux desktops at home and at work. Second (and perhaps this is more important), I hate the overly ‘widgetized’ interface, where you have to click on six different tabs in two different panes and then move the focus between a dozen different fields. Ugh. I’d so much rather have a free-form editor (like Emacs, naturally) and a language for specifying what I want:

     C4 _ Schedule haircut [every 6w, lead 1w]
     B2 _ Weekly project meeting [every Tue@10, lead 1d]

Here, ‘B2’ would mean that this task is ‘rather’ important relative to its parent, and the effort required to prepare for these meetings is ‘average.’ In brackets, we see that this task repeats, and I need ‘1 day’ to prepare. This means the task would appear on my list 48 hours before the meeting, and gradually rise to its maximum priority 24 hours before.

It shouldn’t be too hard to write a simple prototype of this model, where I type these specs in Emacs and have some tool analyze them and spit out a properly prioritized task list.

Of course, looking for a purely technological solution to my broken time management skills is probably not the best use of my time!
□
12:09  / Software § 1 comment

I’ve been meaning to set up a web log for some time. (Hmm…do you think one day ‘web log’ will sound as quaint as ‘violincello’ sounds today?) This is a way for me to quickly and easily publish various thoughts, insights, and announcements on my web site. The reverse chronological order imposed by a ’blog is helpful—previously if I wanted to record some random thought, I’d have to find a permanent place for it in the hierarchical structure of my site. With a ’blog, each entry has a ‘permalink’ so it can be indexed and retrieved forever, but on the primary page, older entries naturally sink down, then fall off the bottom of the page.

This project was delayed for months because I was stymied by the incredible number of choices out there for ’blogging and more sophisticated content management. I knew I wanted software that could be installed and run easily on a server under my control—rather than a service like LiveJournal. Another high priority was to avoid doing most of the editing in my browser’s text boxes! I needed a system that could process text files on my own disk, or perhaps receive them by email. Furthermore, I was reluctant to write my own system from scratch.

Anyway, I decided to go with Blosxom. It assembles a collection of text files into a ’blog with both chronological and hierarchical structure. Its architecture is extremely simple, yet it’s very easy to extend. It can create the ’blog either dynamically (on demand) or statically (ahead of time). I chose the latter option, but integrated my own dynamic comment mechanism using PHP. I hacked some of the standard Blosxom plugins, and wrote a few of my own. My plugins do the following:

• Pipe entries through GNU recode to convert UTF-8 characters to HTML entities. (I configured my Emacs so I can type “curly quotes”—dashes—ånḋ alł kíndß of other things like Γ ⊢ e : τ→σ!)
• Insert thumbnail photographs or amazon.com product images in the left margin, according to meta variables set in the entry text file.
• Provide a ‘ROOT’ construct that can be used within entries and layout templates to refer to the document root using relative links. This is helpful because blog entries can appear at different levels of the URL hierarchy, but I needed some way to refer to other fixed paths on the site. I prefer not to use absolute links because they make mirroring and local browsing a bit more difficult.
• Pipe entries through txt2html to do minor markup including even

      preformatted 
      text.

As I gain confidence in using these new plugins, I’d be happy to make them available to others.

Any ’blog entries that appear chronologically before this one have been imported from other sources and back-dated.

Enjoy!
□
21:13  / Web / Changes § comments?

For the fun of it, I decided this year to compute how much our combined federal taxes would drop if Art and I could get married. The answer is US$1200.  We are overpaying by 8%. GWB could buy a new laptop with that! I think I’ll call this “the gay tax.”
□
10:00  / LGBT § 1 comment

Our Australia pictures are now available. I’d still like to provide more captions, but they’re organized well enough now to make them public.
□
10:00  / Web / Changes § 1 comment

Just a few photos from a summer weekend in Snowmass and Aspen, Colorado. Art was attending an M.D./Ph.D. conference.
□
08:00  / Web / Changes § comments?

Today I dug up the email from ’99 in which I suggested the Wish List feature to amazon.com.
□
08:00  / Web / Changes § comments?

I did some searching today, and discovered that my Erdős number is 4.
□
10:00  / Web / Changes § comments?

I published the results of some personality surveys… just for fun!
□
10:00  / Web / Changes § comments?

I wrote a short feature about a remarkable ten-year-old undergraduate programming project. You can download the code and play the game on any DOS box, or just browse the code online…
□
12:00  / Web / Changes § comments?

I established a page to document some ideas for M.S. thesis projects for my students. More to come in the future, I hope.
□
10:00  / Web / Changes § comments?

I wrote a tiny Perl script to include National Weather Service forecasts into my Emacs diary.
□
10:00  / Web / Changes § comments?

Before I established this ’blog format, I kept a separate list of changes.
□
12:00  / Web / Changes § comments?

Good even as no-nonsense overview.

I was looking for a concise, no-nonsense book about the various XML technologies, including XPath, XSLT, and schemas. Although this book is billed as a reference, it was also a quick way to get up to speed on these ideas, without having to skim hundreds of pages of hype. The schema reference does not have as complete an overview as the other chapters, or I would have given 5 stars.
□
12:00  / Books § comments?

Brilliant, but weak as a reference.

I bought the TeXbook two years ago, but finally spent a few days reading it cover to cover—and I am impressed. As many others, I started exploring plain TeX because I wanted more from LaTeX. I was surprised to find how simple and, yes, elegant TeX is in comparison. I guess TeX is to LaTeX as C is to C++. Certainly do not buy this book if you just want to use LaTeX!

The writing is superb, full of fine detail and more than a few clever jokes. Why can’t recent books about modern systems be so delightful? Maybe David Pogue’s Missing Manual series comes close, but the topics are not quite as technical.

As a reference, the TeXbook is weak because each command or concept is scattered across so many places: one introductory chapter, one summary chapter, in exercises, in “dangerous bend” passages, and so on. I believe the book is best organized for front to back reading, although probably in two or three passes if you include the dangerous bends. For reference, I prefer TeX by Topic by Victor Eijkhout. It is out of print, but available for download on his web site.

The paperback edition of the TeXbook is spiral bound. I appreciate that it lays flat, but the back pages are always falling out of the binding!
□
12:00  / Books § comments?

Fascinating but uneven.

The thesis is fascinating, and really got me (oops.. I) thinking about the role of conscious thought vs. non-conscious action in my own experience: all the places my conscious mind wanders while I’m (somehow) playing a familiar (but not memorized) piece on the piano, or while driving through traffic to the grocery store.

The analysis depends crucially on scientific experiments by Benjamin Libet, whose methodology may be open to criticism. Nørretranders defends the methodology (and I believe it) but the arguments are far from air-tight.

I have long believed that consciousness is an illusion, a subjective property that can potentially emerge (and be useful—even adaptive) in any sophisticated information processing system. I do not, however, buy the argument of Jaynes that consciousness is only a few thousand years old (and may have disappeared in the Middle Ages).

This view of consciousness is of course problematic for the notion of free will. If my brain initiates a movement half a second before I consciously “decide” to move, how can I be in control of myself? Nørretranders tries to rescue free will with a conscious “veto.” The connection he makes to Christian vs. Jewish theology here is interesting but unconvincing—but then, I’m a determinist/atheist.

My biggest complaint: did Nørretranders have to meet a page quota? Part one, about thermodynamics, computation, and information theory introduces some requisite concepts, but they drag on too long. I would prefer that he clearly explain the thesis and some if its ramifications up front; then guide us through some of the prerequisites, periodically tying them back to the thesis. Also, most of part four was irrelevant. Stop reading after chapter 12 and skip to the last subsection of chapter 16.
□
12:00  / Books § comments?