Monday, April 11, 2005

Blogging Hiatus

I haven't been posting much recently. I'm busy, but not more so than usual (for me or for anyone else). I've been busy writing, be it ever so slowly, documentation for my code, papers for publication, and a thesis; and I've been busy reading, and doing mathematics, and programming. But this blog has been largely independent of those things in the past; so why so few posts?

I'm curious about the world, and about other people's work and ideas. I enjoy learning from my friends, by listening or by asking questions or just by observing; and in my turn, I enjoy explaining things. I don't just think my interests are interesting; many are interesting. It's a sort of gut belief: I realize abstractly that not everyone would sympathize with my interest, but that realization has no permanent home in my mind, and so I sometimes forget. I started blogging because I wanted to write about these interesting things, to set them in concrete form to share with others.

However interesting a thing may be, though, it interests different people in different ways. English is full of good words for physical phenomena: interfaces, vortices, breaking waves, vibrations, pressures, currents, orbits, attractions and repulsions. But though I can speak to anyone of a traveling wave and invoke a mental picture, I can invoke the same picture in relatively few with the words essential spectrum. Audience matters. So when blogging about technical matters, I find myself wondering why? Why not write a technical note for a technical audience, either to publish or to post on my web page? Or why not write a less technical article and spend enough time on it (and use enough pages) that I can really craft a good, accessible description of what it is that so interests me about my topic? Or why not grab one of my colleagues or professors and spend fifteen minutes to ask him what he thinks? The blog format doesn't lend itself to any of these expository modes.

For the non-technical things, too, a blog is a poor audience. I can write for myself, or for a few friends I know, or for my family, and the words I write will have meaning for some others. But I'm a sufficiently private person that what I write in a public forum is not the same as what I'd write in a letter to friends or family; and increasingly, I'd as lief write a letter or e-mail as write a blog entry. When placed in the context of a day, it is a fine thing to write that I took a walk in the sun or that I read a great article. If I don't feel like describing the context, though, those statements become banal, or at least far less fine than the experiences of walking and reading. The right response to a brisk walk, for me, is a grin; and I'm far more satisfied when people smile back than when I write I walked, and was happy.

So I'm taking a break from blogging. I do not expect to post here again until the end of the month, and possibly until the end of the summer. I'll take stock and decide whether or not I want to blog any more, and if so, in what form. With a little luck, I will have a thesis draft done by the summer's end, and will not feel so guilty or grouchy about getting distracted by this blog. I will probably start posting articles to my web page involving technical topics (some of them elementary, some of them less so), and, as ever, I will try to keep up with my correspondence.

May all be well with you -- whoever and wherever all of you are.

-- David

  • Currently drinking: Lychee-flavored black tea

Friday, April 08, 2005

Recent reading

I've had a great week for pondering and reading, some of which is reflected in my seminar talk on complex symmetric matrices this Wednesday, and some of which may show up in more detail later. Want to know what a Prony series is? What an Evans function is? What the Correspondence Principle is? Why mixed precision arithmetic is a Good Thing for even simple computations? Okay, maybe you want to know none of these things, but that won't necessarily keep me from telling you.

On the other hand, I made little progress in coding or writing this week. During the breaks when perhaps I should have been writing code, I did some extracurricular blog reading:

I'm going to read more about Evans functions now.

Electronic Paper

Remember the paperless office? The name dates the idea: if it was conceived in the past decade, it would have been the e-office, or if Apple were involved it might have been the iOffice. However it's called, the idea hasn't been realized, and for good reason. Electronic document management still has a ways to go.

  • Display

    I refuse to read books on CRT monitors. When I program or write, I spend time staring into space and thinking; when I read material which is not too technical, I spend almost all my time looking at what I'm reading. The difference in eye strain is tremendous.

    I have fewer troubles with LCD monitors, and I do sometimes read books from my laptop screen. But I have difficulty reading from my laptop in bright light or when the screen is at an angle.

    I think I would be far less inclined to print out papers if I could read them from a sharp display -- monochrome would be fine -- which didn't flicker, was easy to read when tilted to nearly horizontal (the angle at which I usually read and write), and remained easy to look at in a brightly-lit workspace.

  • Portability

    I read at the desk, in bed, in planes, in libraries, in cafes, and in other places as the whim strikes. I can carry a book or paper with me under my arm, or possibly even in my pocket. They don't get hot; they aren't too heavy to hold in my lap; and I can easily hand over whatever I'm reading for someone else to glance at a choice passage (I don't mind someone reading over my shoulder, but only for a short period). It's possible that some of the e-book readers would meet similar requirements, but my laptop certainly doesn't.

    I read a lot of technical material at a desk, though. So for the purpose of removing some of my paper clutter, this is not such a big deal.

  • Browsability

    I have several tens of books that I can navigate by feel: given a topic or a passage, I can get to within a few pages without looking at the pages at all. For most books, I can navigate to pertinent information by rubbing my thumb along the outside edge to reveal a bit of each page in turn. The pages go past far too quickly for me to read anything in detail, but the structure of the text -- paragraphs, inset equations, section breaks, figure layout, and so forth -- is distinctive enough that I don't need to read any details to find what I want.

    The ability to quickly riffle through the pages of a book in order to see where some bit of information lies -- or to verify to myself that I have picked out the wrong book -- is perhaps the most significant reason I still print out papers and read paper versions of technical manuals. Dragging a mouse along a scroll bar isn't nearly as effective, particularly when the rendering engine in the reader software I use (Acrobat or Ghostview) can't keep up. This is a shame, because there are other ways in which electronic documents are far easier to browse than are paper documents: hyperlinks and full-text search are wonderful things.

  • Archival accessibility

    I regularly visit the engineering and math libraries in order to get books and articles which aren't available yet in electronic form. And when I do get them, I make paper copies. These paper copies are often terrible: the text is blurred near the page edge by the binding, or it is cut off at the top or bottom of the page, or it is too dark or too faint. But I still make paper copies, because it's currently the most convenient thing for me to do.

    Perhaps I should invest in one of those hand-held copy devices?

    Besides making paper copies when I access other archives, I'm inclined to make paper copies of my own important documents for my personal archive. Technology does change, as I've been reminded while helping one of my professors convert his old Epson printer files -- on 3.5 inch floppy disks -- into more modern formats.

  • Markup

    I don't like to make notes in my books. I do take notes while I read; but they go onto separate sheets of paper, which I then fold in half and use as bookmarks. A few of my best-beloved books are now probably ten percent note-scrawled bookmark (by volume). For papers, however, I have no such compunctions: when I read a paper carefully, or when I revise a paper in progress, I fill the margins with comments and diagrams (usually in a nearly-illegibly small hand).

    I compose carefully organized notes which are essentially archival; I sometimes never look at those notes again, as the very act of organizing them keeps them in my mind, often for months or years. I also jot notes temporary which consist largely of diagrams and equations, with only a few fragments of text. Sometimes I compose my archival notes directly on the computer, though usually I prefer to write them longhand -- I type too quickly for my thoughts to keep up. But I almost never do any sort of jotting at the keyboard. It's possible to write equations, draw diagrams, and make proofreading marks on the computer, but it's inconvenient.

    (As an aside: I use LaTeX when I typeset my notes. My notes contain equations; and, system preferences and aesthetics aside, I find the MS equation editor too balky to be convenient. But even Word looks good when compared with HTML, which is a terrible markup language for mathematical composition. Were HTML slightly more convenient for mathematical markup, or were extensions like MathML more broadly available, I would no doubt post more mathematical blog entries. I leave it to you to decide whether this would be a good thing.)

It is for good reason we use paper rather than papyrus, vellum, clay tablets, wax tablets, slates, or sand tables. Paper is flexible (both literally and figuratively), inexpensive, durable, and lightweight. Even if it were possible, I would not wish to see to see paper completely displaced. But I do wish electronic document management were a little more advanced, if only so that I could access my technical library with equal ease at home and at work without overloading my backpack.

Monday, April 04, 2005

Not Written

In the past week, I wrote one entry on a quote from William James.

In the past week, I wrote three letters, about twenty pages of miscellaney in my log/research notebook, and more e-mails than were probably necessary. I swapped a new cartridge into my fountain pen, and wrote it dry again within five days. I renewed my library books, and recommended to someone else that they might recall a particularly useful book on Microtransducers CAD which sat on my shelf for about six months a while ago -- I used it often enough that I was sad when it was recalled, but not enough that I was willing to pay $240 for it. I did wash and accidentally put my partly-broken watch through the dryer; fortunately, I'd already bought a new watch. I re-set my clock, which had slipped so that it was nearly a quarter hour fast. I then forgot about daylight savings.

In the past week, I answered questions about SUGAR, about FEAPMEX, and about CLAPACK; I asked questions about VTK; I tried to test out features of OpenDX, gmsh, and GnuPlot. I added a few critical features to the MATLAB wrapper generator I use, including one which allows me to add documentation strings so that MATLAB's help says something better than "No documentation available yet" for my HiQLab commands. I wrote test cases for HiQLab.

In the past week, I listened to a very interesting talk about companion matrices, and started to work on my seminar talk on complex symmetric eigenvalue problems, which I will give Weds. I had interesting technical conversations with people about finite elements and singularity-capturing shape functions, about different interactions between thermal, mechanical, and electrical fields in different types of materials, about dielectric losses and about how to choose a starting node in the RCM ordering algorithm.

In the past week the pope died, and Terry Schiavo died, and Paul Wolfowitz was approved as the World Bank leader. Student government elections are coming up, and the campus has been swarming with people advertising one or the other of the candidates. They have avoided me.

In the past week, I saw Bulletproof Monk and Hero and Once Upon a Time in China II. I read from A Stroll with William James, and have added A Short History of Nearly Everything (Bryson) and Principles of Psychology (James) to my short-term reading list. I read an interesting article on self-plagiarism in the Communications of the ACM, and another article on whether or not CS is really a science. I listened to interesting pieces on All Things Considered about the neo-conservative movement, about John Danforth's views on the religious right's increasing role in the direction of the Republican party, about the pope, and about the return of the radio essay series This I Believe. I read interesting articles on Joel on Software about the FogBugz bug-tracking software. I bought a thesaurus. I read my favorite blogs.

In the past week, I tried Filipino food for the first time, and had pho for the first time in a while. I ate bread and cheese and yogurt for dinner, Mexican for lunch, steamed spinach buns for breakfast, Chinese dough balls for dessert, yuan yang for tea time. I had crayfish with pasta at Ikea while helping Winnie shop for a computer desk. I had pears and chips and shortbread. I probably averaged 2.5 meals a day, but I ate well.

In the past week, I finally got around to ordering my new bus pass, and I signed up for the CityShare car sharing program. I took the train to Fremont once and got stuck across from a man who smelled foul and who kept ranting about the pope. I walked to El Cerrito once, and I took a few walks through parts of Berkeley I'd never visited before.

In the past week, I did all the ordinary everyday things that I do. They were neither dramatic nor profound, but they filled up my days and kept me happily busy.

Friday, April 01, 2005

Sifting Studies

You can give humanistic value to almost anything by teaching it historically. Geology, economics, mechanics are humanities when taught with reference to the successive achievements of the geniuses to which these sciences owe their being. Not taught thus, literature remains grammar, art a catalogue, history a list of dates, and natural science a sheet of formulas and weights and measures. The sifting of human creation! -- nothing less than this is what we ought to mean by the humanities.

-- William James
(quoted from A Stroll with William James by Jacques Barzun)

I need a verb to describe the accidental transfer of knowledge or ideals from one person to another. To rub off only partly captures the idea. I shake my head the same way my mother does, and apparently the same way my grandmother did; the trait rubbed off when I was young. But when I learned how boring a monotonous voice can be in lecture, it didn't rub off. While I may or may not have learned more by watching the nodding heads class synchronize and de-synchronize than by listening to the lecture, the memory of the class's reaction to a dropped book is far more thoroughly fixed in my memory than anything the teacher said. The right sentence for how I learned that day is not she rubbed off on me; it is there but for the grace of God go I.

I have my heroes, as anyone does. There's a pantheon for every field of study and every branch of literature: great ancients, recent stars, and fictional characters. It's a little sad (and puzzling) that anyone would find Paris Hilton more interesting than Gauss. But recently I've come to wonder when students get to learn about -- and from -- the heroes of all the subjects we study in school. The gods of Greek mythology that I learned about in a literature class had more human character than the handful of idols and demons we learned about in history classes. It was a deficiency in my education through high school, one which was remedied partly by leisure reading and partly from college professors who were enthusiastic about the heroes in their fields. Men of Mathematics is one of my favorite books because the pages are filled with characters. I think it is far more satisfying to know something of how mathematics developed and who did the developing than it would be to think that all of modern analysis was created full-fledged from the superhuman efforts of Newton, Liebnitz, and Gauss, who vanished into nothingness upon the completion of their work, and who will undoubtedly show up as ghostly figures in the sky at the end of Episode III.

Gauss with a lightsaber? Okay, maybe Galois. Beware the dark side of Cauchy-Schwartz!

  • Currently drinking: Black tea with vanilla