[python-advocacy] _The Python Papers_ Online Magazine

[Jeff Rush]

[This is an open response to the editor of The Python Papers, Tennessee 
Leeuwenburg, to give him some feedback and make suggestions on how the online 
magazine can be improved.  Viewpoints other than my own are encouraged -- 
Tennessee is a member of the advocacy mailing list.]

See:
   http://pythonpapers.org

Tennessee, thanks for contacting us regarding your issues getting content - as 
advocacy coordinator I very much want to assist you in providing to the Python 
community a quality online magazine focused on our interests.

I've had time now to fully review your site and first issue.  Let's start with 
the first issue:

  - I'd like to see a de-emphasis on the peer-reviewed aspect; while for
    some it denotes quality content, it puts off others who believe that
    your focus is academic research or perhaps carefully researched case
    history articles.  When I first found your site, I thought, "hey, I
    should write an article... hmm, looks like I'd need to carefully
    polish anything I write, complete with references to support my
    arguments.  Too much work for now, maybe later."  While the Python
    community certainly has academic researchers, such as in the SciPy
    sub-community, many of us are far more casual about writing.

  - The _MontyLingua_ seemed a rather dry article, especially with the
    very long bibliography at the end.  I say drop most of the bibliography
    and provide links to web content.  In my world, albeit not the only
    one ;-), if it isn't on the web, it doesn't exist.  While that may seem
    rather narrow-minded, the reality is that many of us don't have time
    to dig up the printed journals and prefer to click and explore a
    topic when your articles have motivated us.

  - You need more diagrams, especially to explain difficult concepts.
    Perhaps you don't have a volunteer graphics designer to help out
    -- if not, we'll need to help you find one.  Regarding your staff,
    while I saw a long list of names, much of the content, of both the
    site and the magazine, appear to be by you.  Hopefully in the next
    issue that will be broadened, since you're just getting started.

  - Since you're not constrained by number of pages, adding photos of
    authors would be good.  I saw in your call for papers that you mention
    this, so perhaps it was just the first issue that was lacking in them.

  - Continue a strong focus on showing Python source code.  We all love
    the readability of Python and its very-high-level nature.  This makes
    it a better candidate than many languages for inclusion in magazines
    - a concisely expressed view into complex but understandable ideas.
    On the other hand, you don't want to show reams and reams of library
    code, because gone are the days people typed in source from Dr. Dobbs
    Journal.  The source is to demonstrate the ideas, not be be a
    distribution medium.  I'd rather be _shown_ than _told_ the concept.

  - Being an electronic journal, I will read your issues online.  The PDF
    format is a problem for that - being in two-column format and sized
    for a letter/A4 paper size, I have to keep scrolling up and down the
    page to follow an article.  And I imagine the paper size is an issue
    for your global audience - someone cannot print it out on their
    regional size of paper.  The PDF is also a problem because I cannot
    link to a particular article -- so you lose the exposure via Digg
    and other such social sites.

    I would like to toss out the idea of adopting reStructuredText as
    the underlying format, and mechanically provide issues in various
    formats.  If you haven't checked out reStructuredText, it is quite
    powerful.  You can embed author credentials within each article
    and, by tagging articles with an :audience: or :difficulty:
    indicator, mechanically catalog articles on different web pages.
    It also becomes possible to associate keywords with articles, and
    later provide an advanced search facility.  You could even develop
    sub-community issues, where pythonpapers.org/zope presents one
    list and pythonpapers.org/numpy presents another.  reStructuredText
    also provides for pull-boxes (sidebars) and custom tags for other
    forms of magazine-specific styles of presentation.

With regard to the website:

   - IMHO, The entry page lacks sizzle.  You should highlight the table
     of contents, with pull abstracts, so people are quickly drawn into
     the magazine.  I would also add a photo of Python people or some
     really cool technical diagram or clipart, changing every quarter.
     The site also doesn't come across as 'fun', but rather serious.

   - Add *prominent* links to the entry page:

         Article submissions gratefully accepted at:
               submissions at pythonpapers.org
         The editors may be contacted via email at:
               editor at pythonpapers.org

     along with a link to writer guidelines your policy of article
     acceptance I can peruse.  As a small time article writer and
     advocacy coordinator, I've checked into what is involved in
     writing for various magazines.  It is so frustrating to dig out
     writer guidelines, to find out which formats they require,
     what rights I retain, whether they pay for content and whether
     they accept content from outsides or only staff writers.  Add
     a large "Write for Us!" graphic icon at the top.

   - I confess I'm unfamiliar with cgpublisher but if issues are
     free, what is the shopping cart for?  A leftover artifact?

   - I realize you're reworking your hosting arrangements, so I'll
     just say to stabilize and clarify your domain structure.  I
     got lost a bit:

       http://pythonpapers.org/diary
       http://pythonpapers.cgpublisher.com
       http://archive.pythonpapers.org

     I saw the announcement about the archive site and wondered whether
     I had to go there to read issues and if there was different content
     there.  No big deal really.

   - The site badly needs a calendar -- to draw in both authors and
     readers, I need to know deadlines, when to check back for new
     content.  I realize this depends significantly on your supply of
     content, but perhaps a regular publishing schedule regardless, so
     that good articles don't get held back waiting for additional content.

     And I'll be radical here, and wonder why there are issues at all.
     In the  days of printed content, it was necessary to batch content
     in order to optimize postage and printing schedules.  In today's
     online world, it seems you could run a continual publishing cycle,
     with an RSS feed to which your readers could subscribe to stay on
     top of content.  It sounds like a blog but an edited, peer-reviewed
     blog, in a consistent format - I would like that myself especially
     if I would subscribe to topic tracks, like Zope and Twisted but not
     Django or TurboGears.  Access to such subscription information
     could also guide you in selecting topics for publication, with a
     carefully designed topic taxonomy.

   - Are you tracking web traffic?  How do you measure the success of
     your endeavor, the popularity of an issue?  And if you keep articles
     together in a single PDF, you'll be unable to track the popularity
     of individual articles.

Other suggestions:

   - Take a stand on whether you are focusing on beginner or advanced
     topics.  While you want some of both, you risk boring one audience
     while addressing the other.  In the Linux market they forked and
     provide separate beginner and advanced magazines.  In your case,
     an approach is to insure a sufficient article count per issue to
     provide something for everyone.  You may want to consider adopting
     a simple icon scheme in the table of content or even an automated
     system of generating unique tables of content for each audience,
     like:  pythonpapers.org/beginners  and  pythonpapers.org/advanced.
     You're not constrained to paper so you can provide different views
     into your content.

   - Be sure to blog the table of contents of _each_ issue, not just
     the overall magazine.  It is those pull titles and pull abstracts
     that will get people to click through to your site.

   - Could you explain a bit more about your use of cgpublisher and your
     hosting arrangements?  If you're looking to change hosting, some
     idea of your infrastructure requirements would help -- I run a
     colocated server and provide hosting to others, sometimes for money
     and sometimes for free.  Is cgpublisher something that can be
     installed elsewhere or a service portal?

   - For discussion here, could you list the titles of the articles you
     have selected so far for inclusion into the second issue?  I'm
     curious how rich your submission queue is and what we can do to help.

     I would also suggest adding a link from your entry page and/or writer
     guidelines to the advocacy page of article ideas, as a suggestion
     for those who want to write for you but can't think of anything.
     You can find it at:

       http://wiki.python.org/moin/ArticleIdeas

     I plan to do some promotion to try to get people to submit topics
     which they would like to read, even if they themselves cannot do the
     writing.  Having a single pool of writing ideas would help us all.

If you are interested in the reStructuredText approach, with a continual 
publish cycle in different formats, perhaps I can help a bit.  My advocacy 
infrastructure work is built on just such a system, although still in the 
developmental stages, at advocacy.python.org.  The idea there is similar, to 
have medium-sized chunks of writing dynamically split out into different 
formats and organizational layouts.  Of course I understand you may prefer the 
cgpublisher approach, with cool features I'm unaware of.

Above all - don't get discouraged!  As Stephan Deibel said, it takes time to 
establish a presence in the Python community.  One key lesson I've learned in 
my work for PyCon and advocacy is that the community is fragmented, and you 
have to go into the various sub-communities and talk to them.  I've joined 
many Python mailing lists in the past year, just so I can converse with 
various groups, which was reflected in the quantity of talk submissions for 
PyCon.  Although we have the general python-users mailing list/newsgroup, 
along with the python-announce list/newsgroup, the vast majority of Python 
developers are on neither.  Here in Dallas, hardly anyone in the DFW 
Pythoneers reads those lists, despite repeated recommendations they do so.  It 
can be very frustrating, that you can't just fire a press release at 
python-announce and assume it reaches people.  Heck, I keep running across 
individual Python developers locally who are on no mailing lists and know no 
one else who uses Python.  They are so relieved to find our user group.  Give 
it time, as we reach out and work to gather them together.

-Jeff



> ---------- Forwarded message ----------
> Date: Wed, 10 Jan 2007 12:16:50 +1100
> From: Tennessee Leeuwenburg <tennessee at tennessee.id.au>
> To: psf at python.org
> Subject: [PSF-Board] The Python Papers: Collaboration with the Python
>     Software Foundation
> 
> I am currently working as Editor-In-Chief on a free journal called "The
> Python Papers" (http://pythonpapers.org). I had intended to spend more time
> building up the interest in the journal before contacting the PSF, however I
> now believe that we will need the help of the PSF sooner in order to make
> the project and ongoing success.
> 
> We are currently approaching our second edition. Our content is primarily
> generated by the team involved in creating the journal, and I had been
> hoping for more interest from outside our group. Unfortunately, while many
> people have expressed their congratulations on the first edition, this has
> not been followed by contributions of articles. We have enough content to
> put out one more edition, but after that I think we will need to re-examine
> whether the journal is feasible unless we find a sharp increase in interest
> from the community.
> 
> There are a few specific things we need help with:
>  1: Gaining greater community contribution. I believe that greater
> recognition would assist this.
>  2: Closer ties with prominent Python groups, such as SIG and PUG leaders
> and notable members of the community
>  3: Technical assistance moving and maintaining our web page for better
> outreach (currently a web-redirect prevents Google from indexing the site)
> 
> I have contacted many PUGs on this matter, although I have yet to contact
> SIGs individually. While there may be good reasons for the lack of response,
> I think it is a shame that the Python community is not more closely
> connected. I had been hoping to feature articles covering the various SIG
> groups, thereby showing somewhat of the diversity and reach of Python around
> the world.
> 
> In addition to this, we provide academic publishing with a full peer-review
> process.
> 
> It seems to me that the goals of The Python Papers are very much in the
> Python community's best interests. Obviously, people may disagree, but to
> date the response has been very positive.
> 
> I would be very interested to hear what the PSF might think of this effort
> and pursue a closer relationship if possible.