[FoRK] A brief thought about a historical oddity in programming language evolution

Dave Long < dave.long at bluewin.ch > on > Tue Oct 31 03:02:13 PST 2006

> One thing that amuses me somewhat about how programming languages
> have evolved over the last 30 years or so is this...  the whole
> "functional programming languages" concept per-se was kicked off and
> into high gear by Jim (Fortran) Backus's 1977 Turing Award lecture,

Would you believe 40 years*?  All of the languages you mention below:

> It's a bit odd that the entire family of functional languages that
> this lecture almost directly gave birth to --- Hope, Miranda, Clean,
> ML and friends, Haskell, and the ilk ...

strike me as having descended more from ISWIM than FP/FL.  (wasn't ML 
also late 70's?)

Landin, Peter J. (March 1966). "The next 700 programming languages". 
CACM 9 (3): 157–166.
http://www.cs.utah.edu/~wilson/compilers/old/papers/p157-landin.pdf

(In which Landin, among other things, discusses referential 
transparency and introduces the offside rule)

For more a more direct FP descent, see <http://www.plasm.net/>.  For 
expressing the point-free (pointless?) aspect of Backus' vision, maybe 
J, or Apter's functional array language experiments with K would also 
be more in line.

On the other hand, I don't know the true history, and, apart from 
Backus, the only pre-1980 items in the bibliography of "Algebra of 
Programming" are math and OR works.  Maybe "inspired" more than 
"sired"?

-Dave

* Many plants grow from their tips out, "standing on the shoulders of 
giants" as it were.  Grasses grow from their base up.  This is somewhat 
like the situation of s/w people, who stand on top of the hardware, and 
patiently wait decades for machines to get fast enough to actually use 
the abstractions they having been dreaming about.  (compiling Haskell 
vs. compiling C sucks up about 20 years of Moore, in my limited 
experience -- or, compare Multics with a current OS)


More information about the FoRK mailing list