Son of Devil's Advocate

Son of Devil's Advocate, June 2001

Click here to make UNIX performance tuning easier!

Previous SODA Index of SODA Columns Next SODA

Son of Devil's Advocate

Stan Kelly-Bootle

Picture of Stan Kelly-Bootle

Clog de Web

As you know, the World Wide Web is even wider than Al Gore or Berners-Lee ever thought possible. Not to mention unworldly wider and webbed beyond the most advanced spiders' dreams.

The naive belief is that, given an "affordable"[ref 1] powerbook and modem, we can all readily access the total sum of real-time human knowledge. It is, without a doubt, all on the Web, incremented hourly as we browse.

I rather like the adverts for the exhaustive "Yellow Pages." If it's not in this compendium, they claim, it doesn't exist. Yet it's clear that no paper-ridden publication can match the up-to- dateness of instant on-line, electronic-content access. Indeed, "Yellow Pages" has the sad connotation of faded "yellowed" pages. Or "slightly foxed" as they say in the manuscript auction brochures.

Browsing, of course, implies a sort of mindless ewe-vine, cow- vert munching, distracted by this or that clump of greenery[ref 2] . And why not? Pre-web scholarship, after all, has long had the equivalent of random footnote serendipity.

But, rather than fill in hand-written slips at the British Museum Librery[ref 3] and wait for physical retrieval, we can now click away at endlessly embedded links, yea nested unto or beyond the big-bangs of human language and writing. Real sad events, some say. I long for the Neanderthal flute-tooting days of yore when harmonic analysis came naturally [ref 4].

Thumbing through grubby index cards is replaced by diverse "search" strategies. Alas, another potential kettle of worms. Apart from the intrinsic "mass of greppage" (aka "pot of message" or "mess of potage" as reported in the Hebrew Testament), whereby "genuine" matches range from 0 to throw-more-disk-exce[p]tion, it appears that we cannot trust these grep algorithms.

Flaws have been found in some of Knuth's most "proven" string- matching programs. Can Dijkstra be far behind?

Even worse, we have, horror, deliberate scams that can modify the results of various commercial search-engine searches.

It seems that (my lawyers are coy) if you pay enough to the major (no names) web portals, any search for any "X," however circumscribed with those damned Booleans, will locate your home page. This adds a new meaning, perhaps, to "finder's fee?"

Yes, I know about content-search and meta-tags, but rather interesting, even useful that "Bill Gates" can be made to string-match "Larry Ellison"?

Final challenge: should a search for "X AND NOT-X" find everything or nothing? Over to Shannan Hobbs, my still favorite pistemologist.

ref 1:"Affordable" is one of my favorite Dickensian non- predicates, loaded beyond belief. E.g., I'm Sudanese with just a dollar to feed the kids. I'm offered the latest laptop for a mere two dollars.

ref 2: Covet not poor Jean Buridan's Ass, nor his wife's Ass. Faced with two equal temptations, one can die of indecision or self-castrate using Occam's razor?

ref 3: A place dripping with blood. I see Karl Marx, protected by his hated Kapitalists, plotting away to murder unborn millions.

ref 4: See "The Mathematics of Musical Instruments," Rachel W. Hall & Kresimir Josic [sorry about the spelling -- Ed], The American Mathematical Monthly, April 2001, Vol 108, #4. A sublime, must-read!

This Column 15 Years Ago

The Devil's Advocate: UNIX Review June 1986

To Err is Human...You're Fired! If this be error and upon me proved, I never writ, nor no man ever lov'd (Bill the Quill, Sonnet CXVI)

The deliberate mistake is an ancient ploy adopted by writers great and small, anxious to avoid that bowel-rending condition known as reader-indifference. Readers just love that warm, quiet glow of superiority (pronounced "Yahoooo!") as each carefully planted solecism swims into their ken. It has been known to backfire, of course. A 17th century English Bible, now remembered as the "Adulterers' Bible," playfully rendered Exodus ch 20 v 14 as "Thou shalt commit adultery." The missing "not" certainly attracted attention, to the extent that the typesetter and proofreader were promptly hung, drawn and quartered, then barbecued as an added bonus for the righteous mob.

More recently, writers of software manuals have developed the deliberate error into a fine art. In the early days, when programmers wrote up their own stuff, documentation was pure and mirrored reality precisely. So, no one called the system house hot-lines. 800-line counselors doodled, atrophied, and eventually left in search of more exciting jobs. The lack of user feedback was clearly inhibiting program development and refinement.

Some unknown genius of a technical writer (so far, at least, no group has claimed responsibility) must have thought, "Is anyone reading our stuff? Let's shove a few random numbers in the index. Or what if we introduce a teeny-weeny typo -- let's spell cd as rd, for example, that should rouse the buggers a bit." The phone-lines were soon a-buzzing, and irate letters clogged the correspondence columns. Rival houses were drawn into heavy competition. "I hear that Lotus got 3,000 calls last week, Jones, while we got a lousy 500. We are not paying you to spread inertia and apathy. Our upgrade sales are slipping. How can all those Version 1.2 users be coping out there? I thought you said the manual was a fine mess."

I must confess that I am not a fan of the deliberate error, although I occasionally indulge in a mild tease or two to check if my esteemed readership is paying due attention. The letter from C. McCarthy querying my use of the word "idempotent" to characterize the operator spoon() (see "Spoonerisms," UNIX Review, March 1986) is a welcome sign that someone cares. I sent him the following valid, but tongue-in-cheek, defense of my choice of words:

Dear C. McCarthy,

Not sure how your bible defines "idempotent," but mine says:

"When an expression raised to a square or higher power gives itself as the result, it may be called 'idempotent.'"

This is from Peirce, no less, the man who first formally defined the word! (B. Peirce, American Journal of mathematics, 1881, Vol IV, 104). They don't write papers like that anymore.

PROOF that Peirce would allow spoon() to be called "idempotent."

Let Z+ be the set of non-zero, positive integers. We can express Peirce's definition as:


For a given n in Z+: IF { n>=2 AND (X)^n = X } THEN X may be called idempotent (whatever other appellations it may attract)


1.1 (spoon)^2 = (1), the identity operator; 1.2 En passant, (1) is trivially idempotent;

This is left as an exercise for the reader (wake up at the back!)


2.1 3 >= 2;

This is also left as an exercise for the reader (I'm tired). (Hint: G. Frege's Grundgesetze der Arithmetik (1893))

LEMMA 3 very pretty and the converse pretty too, But only God and Fermat know which of them is true.

3.1 (spoon)^3 = spoon; 3.2 Indeed, (spoon)^[2n + 1] = (spoon) for all n in Z+

This is left as a final exercise for the reader.

THEOREM Follows from Definition 1 with n=3: (spoon) may be called idempotent in Peirce's sense.

McCarthy rightly claims that (spoon) is an "involution", a name I tend to reserve for conic sections, Desargues and such. Involution has overtones of entangled complexity that are scarcely appropriate for my simple (spoon). All involutions are Peirce-idempotent, of course, but not vice-versa.

Class dismissed!

Postscript: Paul Canniff's criticism (The Last Word, UNIX Review, March 1986) is well taken. I plan to develop LEGOL, a Modula-2-based language that will allow well-structured legalese without the scoping problems Paul refers to. The secret will be the MODULE STATUTE, from which we can IMPORT ProvisoString, a new VAR of TYPE ARRAY of DICTA, where DICTA is INFINITE_ARRAY of CHAR.

Incidentally, may I take this opportunity to congratulate Paul Canniff for submitting a prize-winning Chess program to Scientific American! My spies are everywhere!

Stan's bio:

Liverpool-born Stan Kelly-Bootle has been exposed to computing, on and off and vice-versa, since 1953 when, after graduating in Pure Mathematics at Cambridge University, he switched to impure post-grad work on the wondrous EDSAC I. After some trenching with IBM and Univac in the 1960s and 70s, Stan opted for self-employment as a consultant, writer, folk-song revivalist, after-dinner entertainer, and cunning linguist.

His monthly DA ("Devil's Advocate") column ran and ran in UNIX Review (aka Performance Computing) from 1984 until January 2000 (a date that will live in infamy) but lives on as SODA ("Son of DA") via the homepage devoted to UNIX performance.

The latest of his umpteen books are "The Computer Contradictionary" (MIT Press) and "UNIX Complete" (Sybex). More on his biblio- and disco-graphy can be found on soon due for its millennial update.

Stan welcomes reader reaction:

To the SarCheck home page

The URL of this page may change in the future. Please bookmark the home page instead of this one.

Portions © copyright Stan Kelly-Bootle 2001.
Portions © copyright Aurora Software Inc. 1997-2001, all rights reserved.
This column is sponsored by Aurora Software Inc., makers of SarCheck. The opinions of Stan Kelly-Bootle are his own and do not necessarily reflect the opinions of Aurora Software Inc. Other products and companies referred to herein are trademarks or registered trademarks of their respective companies or mark holders.