yet another "Here's to 1.5.2" post

[Michael Hudson]

weeks at vitus.scs.agilent.com (Greg Weeks) writes:

> Actually, Python 1.5.2 was already a bit too large.  The functional
> programming constructs (lambda, map, reduce, filter) are fluff.

Yes, but that doesn't mean all later constructs are fluff too.

scheme is over there ---------------------------------->

> List comprehension is cutesy and practically Perlish, an idiosyncrasy
> reminiscent (in flavor and aroma) to Perl's ".." operator.

Strange, seeing as they were inspired my Haskell...

> "print" has always been something of a blemish.

But an important one:

    I mean, just take a look at Joe Strout's brilliant little "python
    for beginners" page. Replace all print-statements with
    sys.stdout.write( string.join(map(str, args)) + "\n") and you'll
    surely won't get any new beginners. And That Would Be A Very Bad
    Thing.
                    Fredrik Lundh, 27 Aug 1996 

(from http://www.amk.ca/quotations/python-quotes/page-2.html)

[...]
> The so-called "nested scopes" added in 2.1 are a mistake. 

Says you.

[...]
> With somewhat less confidence -- and really, what's the point of being sure
> about this (like it matters!) -- I think the redefinition "/" and the
> addition of a boolean type were bad choices.

Oh well.

> Now, if only the energy that was put into adding the above blemishes had
> been placed into making Python more robustly installable.  On my system
> (HP-UX 10.2) with my reasonable efforts at building a system:
> 
>     in 1.5.2, there is no FCNTL module
> 
>     in 2.2, "configure" fails due to erroneous C pre-processor code
> 
>     in 2.2, "make test" fails for reasons I haven't had the time to pin down
> 
> I'll keep using Python, so long as Python fills the Perl niche better than
> Perl.  But creeping TIMTOWTDI and installation problems make me worry,
> worry, worry.

Hey, if you want HP-UX 10.20 support, you are probably going to have
to do it.  I wouldn't mind spending a while on it, but I can't.  FWIW,
it seems Python 2.2.1(c1) on 11.00 actually works.

Anybody want to post a HP-UX box to me?

Cheers,
M.

-- 
  I also feel it essential to note, [...], that Description Logics,
  non-Monotonic Logics, Default Logics and Circumscription Logics 
  can all collectively go suck a cow. Thank you.
              -- http://advogato.org/person/Johnath/diary.html?start=4