why isn't python more popular?

[Tim Peters]

[Glenn R. Williams]
> And now we have a different impediment: this flap about the Python
> license, the delay in releasing 1.6 (and 2.0) makes Python look
> unstable, a work in progress, and generally not ready for prime time!

Actually, Python 2.0 is now scheduled to ship several years *ahead* of its
original schedule <wink>.  The delays are partly due to moving to a new
company, and partly to the license haggling.  The former is over, and the
latter is nearing an end.  Since this has hardly been Nightly News material,
I don't expect it will have any long-lasting effect on Python's prospects.
BTW, development is already much more efficient than it was before, as the
move to SourceForge eliminated so many narrow procedural bottlenecks in
contributing code.

> This is probably not a factor in the slow adoption to-date, since it
> happened quite recently. However, in general the Python site changed
> very little (on the surface) over the last few years.

Well, it always was a site designed for adults <0.9 wink> -- the *content*
expanded and improved dramatically over that span.  No rotating 3D metallic
letters, and never will be.  It's been dead in the water for several months
now, though, and that sucks.

> In that vein, the future may be problematic, at least for developing a
> wider user base. The plans announced so far indicate code breakages,
> rapid release cycles, and a big "change of direction".

Where did you hear *that* stuff?!  I'm tempted to say I thought we've been
keeping all that a secret <wink>, but it's actually news to me.  In the
nearly 1.5 years(!) since 1.5.2 was released, it was decided to change the
semantics in a few places to finally match the documentation, and to make
one genuinely backward-incompatible change (having str(unbounded_int) drop
the trailing "L", a change requested by masses of users for years).  I
haven't heard anything about a big change of direction.  I personally *am*
pushing for a less sloth-like release schedule, but compared to the gap
since 1.5.2-- which didn't even have a single bugfix followon
release! --even a yearly release would be "rapid" (btw, if 9 months is long
enough to make a baby, IMO it's long enough to wait for a release).

> That really does affect stability.

If it were true, sure.  A far greater threat to stability is the "bottleneck
removal" I mentioned at the start:  it's never been so easy for people to
submit code changes, and I think having this many active developers grates
some against Guido's personal style.  There are procedural questions here
that have barely been addressed yet.  But so far it's been working out
great, and I'm confident 2.0 will be a peach of a release.  After 2.0 it's
going to require some serious attention to keep it sane, though.

> It's ironic that all those goodies we Python programmers welcome with
> open arms cause many corporate IT departments to shudder! What if it
> breaks? What if my applications no longer run?  Such is I.T.

Do you think Python is in a different boat than Linux there?  Guido still
keeps much firmer control over Python than does Linus over Linux, and while
that may not even be a *good* thing in the end, I think it's more what IT
types want to hear.  Python definitely needs professional marketing to make
inroads into Corporate IT, but then so does anything else.

> The final observation I have is on Grail. This was promoted as an
> example of how Python could be used in "serious" development.

By whom?  It was invented as a demo platform for CNRI's "handle" system
(which keeps popping up in the strangest places during Python's lifecycle
<wink>), and had the usual range of Grand Plans that never materialized.
Take a trip down memory lane:

    http://www.python.org/workshops/1995-12/minutes.s3.html

I especially like the last quote from Guido:

    "I would prefer to keep the Python language pure, and throw away
     all that Java s(tuff)"

Heh.  I don't think Corporate IT ever heard of this -- it was a
quasi-academic research project.

> Then it stops at version 0.6. From the inside, we know how these
> things happen.  But to the unconverted, it looks like just another
> half-finished project.

I respectfully disagree:  to the uncoverted, Grail was an invisible project.
I'd be surprised if the majority of current c.l.py readers even heard of it
before.  Guido's stalled CP4E plans are much more current, inspiring and
visible than Grail ever was (btw, if anyone with vision and deep pockets
<wink> would like to rekindle the CP4E flames, BeOpen.com isn't hard to
find!).

> I hope I haven't offended anyone - Python is my favorite language by
> far! I just don't want to see it lose credibility because of internal
> "issues".

No offense taken, none intended.  The separation from CNRI has been a
supernaturally protracted drag, and that can't be denied.  But other than
that, there's nothing I see going on here that isn't faced by every other
growing & so-far successful software project.  Been there, done that, doing
it again -- it's a hoot!  Have a little faith.  I'm not worried at all.

then-again-some-people-think-i'm-a-flake<wink>-ly y'rs  - tim