[Edu-sig] Goofing with Groups

[Tim Peters]

[Guido, mentions Python's indirect connection to SETL, via Lambert
 Meertens and ABC]

Guido (and everyone else only optionally <wink>), you *have* to take a few
minutes to check out David Bacon's 1997 survey paper, "SETL for Data
Processing on the Internet" (confusingly, this has the same title as his
later doctoral dissertation, which is also on this site):

http://www-robotics.eecs.lehigh.edu/~bacon/survey-pap/survey-pap.html

At

http://www-robotics.eecs.lehigh.edu/~bacon/survey-pap/node2.html

in particular, you'll find Python, ABC, Lambert and you mentioned *by name*
<wink>.  Oddly, though:

    Indeed, in the realm of data processing, SETL still has no peer.
    Its closest competitor is probably Python, an object-oriented
    language that in terms of CWI's support might fairly be called
    the successor to ABC.  Semantically, Python has many of the same
    high-level conveniences as SETL, though it falls short of SETL's
    abhorrence of pointers.

We really have to do something about all those pointers <ahem>.  I think
he's obliquely referring to SETL's (like ABC's) "by value" semantics, but
phrasing it misleadingly.

    Syntactically, however, it is rather a mess,

Don't worry, Lispers hate SETL's syntax too <0.5 wink -- and Bacon returns
the favor>.

    and has nothing like the set formers that give SETL so much of
    its flavor.

It does now (i.e., Python 2.0's list comprehensions were adapted from
Haskell, which in turn borrowed them from SETL).  SETL was gloriously
innovative, and still worthy of study.

it's-a-small-world-if-not-entirely-a-vicious-one-ly y'rs  - tim