Webware / WebKit / Docs / Future.txt

Future Work

Webware for Python

:Version: 1.1
:Released: 08/03/11


You can find more information about known bugs and future work
in the Wiki__ and in the `SourceForge tracker systems`__,
both accessible from the `Webware Home Page`__.
The SourceForge task manager is currently not used since these systems should
be sufficient and we do not want to scatter issues in too many systems.


Future Work/Limitations

Sprinkled throughout the code are comments tagged with ``@@`` which are
hopefully accompanied by a date and someone's initials.  These comments
represent things to be done.  The double at-sign (``@@``) convention was
chosen because it doesn't appear to be used for anything else.

In addition to the inline comments, some significant items have been
recorded below.  These are future ideas, with no commitments or timelines
as to when/if they'll be realized.  The Python WebKit is open source,
so feel free to jump in!

Known Bugs

All major known bugs that existed previously have been fixed.

To Do

Major Items

* Some of our old style guidelines are in conflict with newer Python
  conventions (PEP8, properties etc.). We have already changed the
  Webware convention of using tabs, but most naming conventions cannot
  be changed so easily without breaking the API.

* Think about a migration path towards Pyhon 3, check code with 2to3 tool.

* Use Distutils/setuptools/Distribute instead of our own installer, use
  Python eggs and maybe namespace packages instead of our own plug-in concept.

* The installer should try building and/or installing mod_webkit and wkcgi

* Use Sphinx instead of our own tools in DocSupport, use reStructuredText
  as the format for all docs and maybe also for the docstrings.

* Separate more clearly and systematically which messages are printed to
  stdout and which to stderr.

* Make use of the logging module (so far it is only used with unit testing)
  instead of conditional print statements with "debug" and "verbose" flags.

* Role-based security and user-authentication.  Goal is to eliminate,
  as much as possible, developer-written security logic.
  This should be provided by the WebKit and be configurable.

* Better support and documentation for distribution, load balancing and
  fault tolerance.

* More sophisticated admin tools including password protection,
  clearing logs, displaying a maximum of information at a time, etc.
  Consider using module 'resource'.

* Investigate case insensitive URLs, especially for the Windows platform.

* Support unicode strings and different encodings.  HTTPContent should get an
  encoding attribute with 'utf-8' as the default (set in Application.config).
  This could be output in Page.writeMetaData() and used as the default for
  an optional encoding parameter in HTTPContent.write(), so you can write
  unicode strings to output.  So far, Webware works best if you use ordinary
  strings with a consistent encoding throughout the whole application.

* Integrate support for i18n and l10n.

* Plug-ins:

  * In ExamplePage, automatically support examples of any plug-in
  * Better docs
  * Properties.config. 'Load', 0, 1 or the name of the required op sys


* Hunt down: ``@@`` tags (which signify "To Be Done"s), ``FUTURE`` items
  in class doc strings, ``NotImplementedErrors``, ``--`` tags

* Code clean up. Use bin/checksrc and pylint.

* Right now, if the Application itself (as opposed to Servlets) throws
  an exception, it doesn't get captured nicely.  However, it is displayed
  in the app server's console.

* The exception handler is pretty nice and has features like logging,
  e-mail, gathering debugging info, etc.
  However, on occasions it can throw exceptions too.
  There should be a simpler, secondary exception handler for when this happens.

* Review the timestamp caching logic and its relation to .pyc files if any.

* Add "Last-modified:" to generic files that are served via WebKit.

* If a Python file has only one class that inherits from Servlet,
  then use that as the Servlet class
  (rather than requiring the name be the same as the file).


* Provide testing web page where people can report their testing results
  including version numbers, etc.

* Provide higher level automation of testing.  For example, a testing script
  should be able to launch various app servers multiple times.

* Provide highly automated benchmarking so we can track changes in performance.

* Add more unit tests, expand the regression test suite.


* Use Sphinx instead of our own tools in DocSupport, use reStructuredText
  as the format for all docs and maybe also for the docstrings.

* Add a Getting Started Guide and a screencast.

* Beef up the User's Guide and Tutorial.

* User's Guide: Create a caching section to discuss the virtues of doing so
  (the Color example became 12 X faster on the server side).

* Clean up the Wiki, removing all the obsolete cruft and summarizing
  everything that is still useful, moving it to the Sphinx documentation.

Food for thought, considerations, reviews

* Consider including FormKit, FunFormKit or FormEncode:
  A plug-in to aid the construction and validation of forms.

* Consider adding a simple helper lib for generating HTML
  (such as SimpleHTMLGen) to the WebUtils package.

* Better support for `WSGI <>`__.
  Currently we only have the WSGIAdapter interfacing to the app server.

* Consider this statement from the FastCGI docs:
  "Redirects are handled similar to CGI. Location headers with values
  that begin with "/" are treated as internal-redirects; otherwise,
  they are treated as external redirects (302)."

* FastCGI app server:
  The idea is that if the app server itself supports FastCGI,
  then it can be used directly with FastCGI enabled web servers
  sans the infamous "adapter".
  Dan Green has brought this up in Webware-discuss.

* Consider if we need to support ``<form action="" method="post">``
  where you will have both a query string and posted data.

* Application modifies sys.path so that servlets can say
  ``from SuperServlet import SuperServlet`` where SuperServlet is located in
  the same directory as the Servlet.  We'd prefer a more sophisticated technique
  which does not modify sys.path and does not affect other servlets.  (Or maybe
  this would go away with a new one-process-per-application architecture.)

* The ThreadedAppServer is not optimal for multiprocessor systems and modern
  multi-core CPUs, due to the Python GIL. On Python 2.x, the GIL implementation
  is particularly bad. Check whether we need to build something new such as
  a MultiprocessAppServer, or use existing app server or tools such as Twisted.

Check out

* `Python Paste <>`__

* `Django <>`__,
  `TurboGears <>`__,
  `Pylons <>`__,
  ` <>`__

* `Aquarium <>`__




* FastCGI related: