Welcome to Orcmid's Lair, the playground for family connections, pastimes, and scholarly vocation -- the collected professional and recreational work of Dennis E. Hamilton

Recent Items
 
Agreeing While Disagreeing
 
Addressing Customer Demands?
 
Crushing Supply-Push Under Demand-Pull
 
The fearful state of on-line students
 
The Final Page
 
Oh, the tragedy, the humanity: Donate Now (amazon link)
 
The Asia Disaster: Aiding the International Relief Effort
 
Tsunami - Earthquake - Supporting International Relief
 
The Heart of Trust
 
Accountability: Lessons from Engineering and Medicine

Don Knuth on Science as Art and Scholarship as Community Effort

NPR : Donald Knuth, Founding Artist of Computer Science.  Here's the kind of podcast that I couldn't not listen to.  My contemporary (we're a year apart) talks about how he looks at problems and why the kitchen wastebacket is in the center of the room.  The interview wanders through Donald Knuth's office at home, his filing system (behind him in the photo), and how he toils in the fields of Computer Science.  I was awestruck and moved by 00m30s into the NPR host's lead-in.

Partway through the interview, DEK refers to the dependability and trustworthiness of the Art of Computer Programming and how, knowing there would be mistakes, he engaged the community of computer science to find corrections and refinement.  He mentions that 90% of the pages in the original volume 1 had changed in the second edition, and we're now in the third (plus on-line errata).  He's paid out over $20,000 in checks for bugs, although people cash those checks less and less these days, making a problem for balancing his checkbook.  I must find the framed nit-picker award that I held onto all these many moves from place to place after letting him know which one I wasn't cashing.

The interview ends with one of Knuth's inspiring chapter-lead quotations.  I still look at The Art of Computer Programming (reverently spoken of as "TaoCP" in geek code) as art.  The volumes have owned a prominent place on my bookshelves long since the two volumes of Courant's calculus and the three of Principia Mathematica were sacrificed to the gods of traveling light.  Many, especially wary instructors, find the books inacessible in our math-challenged age, and yet there is much to reward the steadfast learner.  Like profound art, it may take some devotion to the craft before the beauty of the work becomes apparent.

---

Knuth's gentle message and playful spirit were brought to my attention at the perfect moment this morning.  I am failing to maintain even the simplest schedule for my M.Sc in IT.  I can't find a simple C-Language bibliographic-entry addbib program like the one I compiled and used on MS-DOS 20 years ago, and I've been obsessing over that problem instead of keeping up with my commitments.  I hate do-overs and don't trust myself to quickly build an off-hand replacement for the program I've misplaced on some old computer somewhere and haven't found in Gnu or Linux code anywhere.  I've been making myself crazy installing and removing poor substitutes.

In the midst of that maelstrom, I was thinking, just last night, that Knuth's taking 10 years out for TeX couldn't have been worth it.  How stupid and arrogant of me.  I've witnessed firsthand the power of Knuth getting his tools in order ahead of a project.  I watched him build his own assembler so he could perform faster run-debug-fix-reassemble-run ... cycles for a FORTRAN compiler he was building.  And he arranged for that assembler to take hints into minimum-latency optimization that it provided; then he had it produce flowcharted documentation from his specially-annotated comments.

Recalling that inspiring feat, I suddenly felt much lighter, caught my breath and took a calmer approach. I'm adopting a custom *.bib template that I can honor by hand in a text editor. Later, when I have more space in my schedule, I'll make a little tool to provide guided entry and updating of annotated-bibliography files.  Even later, I'll worry about extracting citations and references in the different formats I use:  My idiosyncratic annotated bibliographies and the Harvard style for the dissertation.  What a relief.  I've been productively purposeful all day.  Many thanks to anderbill for the heads-up about the NPR interview.

posted by orcmid at 3/15/2005 08:31:00 PM