BKHKRPNT.RVW   20041023

"Hackers and Painters", Paul Graham, 2004, 0-596-00662-4,
U$22.95/C$33.95
%A   Paul Graham
%C   103 Morris Street, Suite A, Sebastopol, CA   95472
%D   2004
%G   0-596-00662-4
%I   O'Reilly & Associates, Inc.
%O   U$22.95/C$33.95 800-998-9938 fax: 707-829-0104 nuts@ora.com
%O  http://www.amazon.com/exec/obidos/ASIN/0596006624/robsladesinterne
  http://www.amazon.co.uk/exec/obidos/ASIN/0596006624/robsladesinte-21
%O   http://www.amazon.ca/exec/obidos/ASIN/0596006624/robsladesin03-20
%O   tl i rl 2 tc 2 ta 3 tv 1 wq 3
%P   258 p.
%T   "Hackers and Painters: Big Ideas from the Computer Age"

The preface states that this book is intended to explain to
non-computer people what goes on in the world of computers.  I'm not
quite sure why Graham thinks he's done that.  There is very little in
the book that explains the computer "world" with all its
ramifications.  Possibly the closest he comes is to give a bit of a
glimpse into the mind of one particular programmer and entrepreneur.

I was really prepared to like this book, to begin with.  I mean, a guy
who starts off by stating that " ... innovation and heresy are
practically the same thing" can't be all bad, and anyone who can work
C. S. Lewis quotes into a book on technology is doing OK.

Chapter one is an insightful examination into the fact that "nerds"
are generally unpopular, particularly in school.  As a former school
teacher, I have to say that Graham's characterization of school
society as having a lot in common with prison society is more valid
than most people are comfortable accepting.

I agree with Graham initial position in chapter two: we do not yet
know the form that programming will finally take.  But I do not feel
he proves his contention that structured programming is completely
wrong.  After all, he himself points out that programs should be
readable, and structured programming provides an environment that
supports the documentation and clarity of thought that produces such
code.  In chapter three the author challenges us to examine our own
thinking, but his techniques are not necessarily useful.  If you
decide what someone in search of "cool" would reject from his parents'
values, what does that indicate about who is right?  It has, in fact,
been noted that incompetent people are not competent to determine
their own incompetence.  And what is the point of this proceeding? 
Simple mental agility?  A good thing, yes, but is it truly a big idea? 
Hackers are an unruly bunch, but they are hardly the automatic fascist
detectors that chapter four makes them out to be.  Server-based
software has the advantages that chapter five points out, but those
blessings have downsides, too: you can release as often as you like to
improve the software, but you will annoy users who liked it as it was. 
I can agree that wealth does not equal money, but much else that is
stated in chapter six sounds like a free enterprise fantasy, and seven
is even more so.

Chapter eight is basically an opinion that statistical or Bayesian
spam detection is the best, and seems to ignore the fact that fraud is
not the only ill that spam creates, what with the drain on bandwidth
and resources it consumes.  Good design is beautiful, but chapter nine
neglects the fact that many times the better product does not win:
Graham includes an odd return to support for disparities in pay, and
thus implies that the marketplace will identify the elegant solution. 

Chapter ten's explanations of aspects of programming are clear, but
because they are extremely selective the overall impression would
likely be misleading.  Non-technical readers will probably believe
that they understand what the author is saying in chapter eleven about
the design of a properly timeless programming language, but I doubt
that they will comprehend the full implications.  (One of which is
that, contrary to his earlier statement, assembly language is best.) 
Graham says in chapter twelve that he likes the Lisp programming
language.  In chapter thirteen he derides Pointy Haired Bosses for not
understanding technology, but still insists on explaining, to this
presumably non-technical audience, the advantages of Lisp in terms of
extremely technical aspects such as the internals of the language and
relative code sizes.  Chapter fourteen describes the ideal programming
language--from Graham's perspective.  More good things about Lisp are
listed in chapter fifteen.

Graham certainly hasn't explained the world of computing, and he
hasn't even recounted all that much about programming.  His arguments,
such as they are, lack compulsion: just because his startup succeeded
does not mean either that Lisp is a good language or that he created
wealth (as opposed to obtaining it at someone else's expense).  There
are some interesting points in the book, but the purpose of the work
escapes me.

copyright Robert M. Slade, 2004   BKHKRPNT.RVW   20041023