[python-advocacy] Marketing Python - An Idea Whose Time Has Come

[Todd Grimason]

* Tim Parkin [2006-04-21 10:47]:
 
> The Ruby propoganda has travelled mostly on the back of rails as far as
> I can tell (everyone I know that isn't a programmer seem to think Ruby
> is Rails and vice versa).
> 
> Python.org can never be (and never should be) a
> http://www.rubyonrails.org/ because it is a language and not a product.
> 

[snip]

Tim's post is dead-on -- and the "language not a product" observation is
key. 

That said, Ruby itself (at least among developers) has been getting much
more attention than it previously was, mostly if not entirely driven by
Rails -- a product. 

Now of course there are lots more "big" products written in Python, but
a key thing about Rails is it's a "web" product. I get the sense many
Python-folk spend time writing massively parallel, scientific or crazy
math stuff I could never hope to understand -- nor could the many
work-a-day developers out there who drive book sales, corporate
adoption, and then academic adoption. They are more and more working
with web development (or at least bolting on a web frontend), and for
example Tim O'Reilly recently posted they saw the biggest increase in
web dev books, around 20% (somewhere on radar.oreilly.com). Ian Bicking
also described this in a weblog post a few months ago. If you're doing
scientific calculations in a uni lab, book sales and visibilty are
probably(?) not so important to you.

So I'm making sweeping generalizations and meandering a bit -- but
clearly web development has a very large impact on "public" awareness of
a language. Bittorrent, Google using it inside, NASA, etc., are nice and
well, but do not provide visibility in the "check out this app, it's
really well done, and written in Python" sort of way that public web
apps do. And saying 8 gazillion people use Bittorrent is kind of
meaningless in that most of them have no idea it's written in Python and
couldn't care less what it's written in as long as they can find that
song they heard on the radio.

As it's unlikely any library/framework in the web space will ever be
"blessed", whatever the pros or cons of doing so, the product approach
seems a non-starter. So there's the other successful path -- ubiquity --
a la PHP. This I think is was Ian was getting at, and I believe efforts
at WSGI are (partially at least) aimed at -- vastly improving the
"deployability" and ease for ISPs and such to deploy Python webapps.
This seems the best possible way, esp considering Rails has helped push
lighttpd and more flexible Apache setups into more ISPs, but Rails
itself is still no picnic to setup for beginners. They are working on
this ('mongrel' for example), and at this point it looks like they might
have a widely-used, supported (it is getting funding from some companies
for development as well) deployment solution before Python. Leapfrogging
Rails on deployment, and/or "composite apps" (lots of little apps
without 5 entire Rails setups for example) might be a fruitful path.

Anyhow, that's some thoughts from a more casual Python user, which while
I can't argue on the best implementation in core for immutable sets or
whatever, I think I can provide some decent "how it looks from here"
input. Or at least I hope so. Of course, it's short on answers. But
maybe some ideas can come of it.



-- 

________________________________
toddgrimason*todd[ at ]slack.net