Subject: Re: logical pathnames - lack of info?
From: Erik Naggum <erik@naggum.net>
Date: Fri, 22 Feb 2002 20:30:15 GMT
Newsgroups: comp.lang.lisp
Message-ID: <3223398617285191@naggum.net>

* Erik Naggum
> The obvious eludes explanation.

* Kenny Tilton
| Not to a good teacher.

  Heh.  This is probably true, but it takes a lot of work to find the way
  back to how you found something out.  Most of us discover something in a
  round-about way if we have not learned it from someone else.

| Good teachers remember what it was like not to know, and they remember
| how they got from there to knowing.

  However, this can be their most serious problem in teaching others.
  Spending years and years on something until it finally makes sense is not
  something you would want your students to go through in order to learn
  it.  On the contrary, "obvious" is extremely hard to pin down before the
  fact, and, well, obvious after the fact.

| Someone else has noted that CLTL2 goes from a formal description to bad
| examples--nothing useful there to a student.

  It depends on the experience of the student.  As a friend commented in
  mail, having used and known TOPS-10/20 makes it so much easier to figure
  out how pathnames work and why they are designed the way they are, with
  what they humorously called "pathological devices" which were very close
  to logical pathnames in design (and which caused me to misunderstand the
  word "pathological" as used elsewhere for a few years :).

  For someone who has only seen Unix or MS-DOS, pathnames are incredibly
  complex beasts, but for the more likely audience of CLtL2, which was not
  published at a time when it was clear that Unix would be only file system
  paradigm to survive (whatever Microsoft does is just copying Unix).

| Well, some students can read a formal description and "get it".  I need
| examples, and that is the big problem I have with this subject (and
| throwing in the variability of OS- and implementation-dependence
| massively does not help, tho mebbe that cannot be helped).

  It would probably be more useful to find examples of how filenames were
  named in VMS, TOPS-10, TOPS-20, Multics, PrimeOS, ITS, etc, all of which
  I had used or read about at the time I faced the pathname specification.

///
-- 
  In a fight against something, the fight has value, victory has none.
  In a fight for something, the fight is a loss, victory merely relief.