[omaha] Idea for Multi - meeting/person/topic presentations
Jeff Hinrichs - DM&T
jeffh at dundeemt.com
Wed Mar 14 05:47:56 CET 2012
There have been requests for a beginning Python topic for each meeting in
addition to the current type of topics. From prior presentations on
beginning topics, they are usually quite dry, running down a laundry list
of topics from dictionaries, to PEP8 to iterators, etc. The presentations
are ok but I was thinking about a way to have a multi-focus project that
would cover the spectrum. Something completely different...
As a thought experiment, pick a slightly more than trivial project -- say
Gina Trapani's Todo.txt http://todotxt.com/ -- a nifty idea for a
simplistic but workable todo list app. But built on a hodge podge of bash,
awk, sed with plugin's using programming langauges all over the universe.
Say we agreed to re-implement the app in Python. It's GPL we can do that.
It is flat file based with lots of string processing and user interaction
-- just the kind of stuff that make for a decent first program. We can
show off file handling, iterators, etc. Pretty good topics for beginning
Now we have also had requests for presentations on testing. This would
make a good vehicle for demonstrating unit tests, smoke tests, integration
tests and acceptance testing.
Every meeting we would have someone volunteer to give a talk on topic X and
use this on-going project as the guinea pig for demonstrating.
- We would host the code up on a public repository, say bitbucket and
then people could demo using mercurial and git.
- It would be a candidate for using Tox http://pypi.python.org/pypi/tox to
assist in making it 2/3 compatible
- Documenting your project with sphinx and/or read the docs
- checking your documentation examples with doctests
- Packaging your app for Linux, Mac, Windows and putting it up on pypi
- Using nosetest w/wo coverage.py
- The list goes on, webify it, make a gui, etc, etc.
Now not every talk would be using this "group project" as the example -- a
lot of times it just wouldn't make sense, but when it did make sense, we
would have a common code base that all were familiar with or could get
familiar with quickly.
I look forward to your responses -- I hope you get my angle on this, but I
think we could make this or something like this a real win for the group.
More information about the Omaha