[Pydotorg-redesign] Domain-specific content

[Skip Montanaro]

I suspect actually coming up with domain-specific content will be the
biggest barrier to contribution, even for domain experts. Picking a person
and domain at random, if we asked Andrew Dalke to write a white paper on
Python and Bioinformatics, I suspect the major effort on his part would be
writing the paper, not stuffing it into the web site. Sure, massaging that
content and making it available to others to edit would be more difficult
with ht2html than Zope, but at the moment we have no content to manage. The
issue of how we manage the content is probably a secondary concern. I think
we'll be very lucky if content management as opposed to content creation
becomes the primary barrier to a better site.

I think there are several of us who agree that adding domain-specific
content to the Python website would be worthwhile, at least from a marketing
standpoint. Can we spend some time on that? I'd like to flesh that out a
little to see if we can identify a structure for such content, what domains
we want to go after first, and maybe sound out some people for potential
submissions. Once we're drowning in content we can see how hard it is to
maintain.

I'll toss out some thoughts about what I think such pages should/might
contain:

* brief overview of the domain (just a paragraph or two, with perhaps a
couple links to other introductory material)

* applications in the domain which are written in Python

* core and third-party modules and packages in Python which support the
domain

* references to domain-specific websites which support the use of Python
in the domain

Back in April I (probably prematurely) came up with a possible starter list
of domains and domain experts:

Scientific Computing Eric Jones
XML Uche Ogbuji
Data Visualization/3D Graphics Prabhu Ramachandran or Michel Sanner
Web Services Mark Pilgrim
Relational Databases Andy Dustman or Federico di Gregorio

Here are a couple more:

Web Application Servers Paul Everett, Andrew Kuchling,
Geoffrey Talvola, many others
Bioinformatics Andrew Dalke

There are obviously many other people who could contribute in these areas.
The names above are just those that popped into my head.

I don't know how many domain-specific pages or sections we want on the site.
I think the above would be a good start. If we can get content for four or
five areas I think that would encourage other people to contribute.

Skip

[Roy Smith]

Post by Skip Montanaro
I suspect actually coming up with domain-specific content will be the
biggest barrier to contribution, even for domain experts.

Bingo. Having been involved in a few web projects, it's painfully
obvious that the technical issues of building a web site are the easy
stuff. Content is the hard part.

[Dylan Reinhardt]

Post by Skip Montanaro
I suspect the major effort on his part would be
writing the paper, not stuffing it into the web site

[...]

Post by Skip Montanaro
at the moment we have no content to manage.

An astute observation. Also fairly circular.

As is stands, there is no *broad* invitation to submit content, so it's
not surprising that little content is contributed. Low amounts of
content suggest that there is no need for content management.

The key is that a CMS encourages submissions of a quality that falls
somewhere between Barely Useful and Highly Decent. When the author
stops getting negative feedback about the content, it is Good Enough.

With a higher barrier to contributions, it's hard to imagine running
something that's only Good Enough. If it's a big deal to have something
published, you'll *treat* it like a big deal. At least 9 out of 10
times that means it won't get finished. End result? No content.

It's a bit like e-mail. If you keep re-writing your e-mail until it's
good enough to carve in granite, you'll never send any. But it's got
your name on it, so you'll expend enough effort to avoid looking like an
idiot. That's the same general quality proposition of a CMS.

I can already hear you wondering where the high barrier for contributing
content to python.org comes from. After all *you* are actually
*encouraging* people to write stuff. Look a the site. Most of what
content it presents is authored by the undisputed luminaries of the
language. Like it or not, that sets the bar pretty high.

Let's say I have a moderately clever 10-line recipe for solving an
uncommon but difficult problem. It's nothing brilliant, but other
people might benefit from seeing it.

You would never think to ask me if I had something to submit to
www.python.org because you don't know me and couldn't possibly be
familiar with what I'm working on (though by now, you are no doubt
beginning to suspect it has something to do with Zope). :-)

If most of the content on the site is Guido's high-level philosophy of
language design, it's pretty obvious to me that www.python.org isn't the
place for my trivial little limited-audience hack. I'll just post it to
the list and within a couple weeks, nobody will ever find it again.

What if there were a place where small (but useful) things could be
stored and actually *found*? Well, I'd put my recipe *there* instead,
wouldn't I? So would other people who had recipes, tips, links, book
reviews, opinions, and other things that are useful but not important or
weighty enough (individually) to work into the current site.

Let's say I'm not a person who can (or would) write a white paper on
bioinformatics. What if I have a couple good algorithms I'm willing to
share? Wouldn't it be great if my small contributions could be included
in the total body of information collected for that domain? Wouldn't
that make our case for using Python in bioinformatics stronger?

Looked at another way, what if 5% of the energy put toward c.l.python
were directed toward making the same information available in a manner
that was better organized and more usefully searched? Wow. There's a
heck of a lot of genius on that list... it's a shame how short-lived
most of the contributions really are.

Whew! That was more than I intended to write. :-)

Dylan

[Aahz]

Post by Dylan Reinhardt
Let's say I have a moderately clever 10-line recipe for solving an
uncommon but difficult problem. It's nothing brilliant, but other
people might benefit from seeing it.
You would never think to ask me if I had something to submit to
www.python.org because you don't know me and couldn't possibly be
familiar with what I'm working on (though by now, you are no doubt
beginning to suspect it has something to do with Zope). :-)
What if there were a place where small (but useful) things could be
stored and actually *found*? Well, I'd put my recipe *there* instead,
wouldn't I? So would other people who had recipes, tips, links, book
reviews, opinions, and other things that are useful but not important or
weighty enough (individually) to work into the current site.

Just in case there's anyone on this list who doesn't know the correct
answer to this example:
http://pythoncookbook.activestate.com/

(Yes, I agree that we should move python.org in this direction.)

[Dylan Reinhardt]

Post by Aahz

Post by Dylan Reinhardt
What if there were a place where small (but useful) things could be
stored and actually *found*? Well, I'd put my recipe *there* instead,
wouldn't I? So would other people who had recipes, tips, links, book
reviews, opinions, and other things that are useful but not important or
weighty enough (individually) to work into the current site.

Just in case there's anyone on this list who doesn't know the correct
http://pythoncookbook.activestate.com/

Just to be clear, I'm well aware of the cookbook and frequently thankful
it exists. The recipes I've gotten there have been top notch.

The cookbook site was a great tool for what it was designed to do (fill
a book) and still has significant enduring value. It's barely
scratching the surface of what's possible in the CMS space.

Post by Aahz
(Yes, I agree that we should move python.org in this direction.)

Cool.

Dylan

[Skip Montanaro]

Dylan> Let's say I have a moderately clever 10-line recipe for solving
Dylan> an uncommon but difficult problem. It's nothing brilliant, but
Dylan> other people might benefit from seeing it.
...
Dylan> If most of the content on the site is Guido's high-level
Dylan> philosophy of language design, it's pretty obvious to me that
Dylan> www.python.org isn't the place for my trivial little
Dylan> limited-audience hack. I'll just post it to the list and within
Dylan> a couple weeks, nobody will ever find it again.

Dylan> What if there were a place where small (but useful) things could
Dylan> be stored and actually *found*? Well, I'd put my recipe *there*
Dylan> instead, wouldn't I? So would other people who had recipes,
Dylan> tips, links, book reviews, opinions, and other things that are
Dylan> useful but not important or weighty enough (individually) to work
Dylan> into the current site.

The correct place for that would be ActiveState's Python Cookbook:

http://aspn.activestate.com/ASPN/Python/Cookbook/

Lot's of people contribute to it. A large collection of the early ones were
even bound up into an excellent book by the esteemed Mr. Martelli.

Dylan> Let's say I'm not a person who can (or would) write a white paper
Dylan> on bioinformatics. What if I have a couple good algorithms I'm
Dylan> willing to share? Wouldn't it be great if my small contributions
Dylan> could be included in the total body of information collected for
Dylan> that domain? Wouldn't that make our case for using Python in
Dylan> bioinformatics stronger?

Sure. But at the moment we have no hook for you to hang that stuff on. I
want to talk about what we need in the way of contributions to get things
started. Let's see if we even need a content management system. It may
turn out that ReST is good enough.

Skip