[Chicago] Trac?

Garrett Smith garrett at mojave-corp.com
Sat Sep 30 15:30:26 CEST 2006


This is an interesting thread, at least to me. I've long been interested
in applying OSS development patterns to more general cases -- business
management and research being good examples. I think there's a new,
compelling theory of collaboration and information management that's
largely unarticulated (at least formally) that we can derive from the
astounding success of various OSS initiatives. Folks like the
signatories of the Agile Manifesto (http://agilemanifesto.org) have
written from the stand point of software development, but I think this
stuff applies well to more general cases.
 
(Okay, I just lumped OSS and Agile together without so much as a hand
wave -- but these two forces routinely bump into one another.)
 
Some of the principles that I see applying well to collaboration in
general:
 
 * Empower individuals, or at least individual workgroups
 * Work hard to keep things simple
 * Provide safety nets for experimentation -- and in particular,
mistakes
 * Make change transparent
 
In my experience, these four items are very rarely present in most
organizations. In fact, the opposite of each generally holds. On the
other hand, all four are thoroughly baked into most OSS projects (maybe
not the second point so much...this is more common in Python OSS
projects :-)
 
Wikis and source management systems are key technical enablers for each
point:
 
 * Wikis allow anyone to change anything. (Insecure managers: dont knock
it 'til you try it!)
 
 * As you pointed out SharePoint is a complex beast -- complex to
implement and maintain. Wikis are much simpler and get the job done.
(KISS helps beat down people's tendency to become too clever.)
 
 * Wikis and source control systems are experts in backing out or
amending changes. Be bold and plow ahead knowing full well that what you
add today can be amended at any time. How empowering is that!
 
 * Wikis and source control systems are expert in tracking change. For
those who have experienced this degree of transparency, it's a powerful
social dynamic that simultaneously empowers people (there's no
centralized information gatekeeper) and keeps contributors accountable
(everyone can see what everyone does).
 
So, without making this reply any longer, I think you're on the exact
right track, or Trac.
 
One other point, then I'll be done :-) My teams have started using
Subversion for managing all document types -- e.g. Word, Excel,
PowerPoint, etc. -- in non software development projects. For example,
we just finished up a large proposal that used lots of misc Office
documents -- and all are under source control. GUI clients like
TortoiseSVN let non-technical users update and check in changes with
minimal training. Also, I was ''very'' pleased to learn that Word
supports diffing -- e.g. you can double click a revision of a Word
document in TortoiseSVN and view the changes in Word's "track changes"
mode. Sweeet.


________________________________

	From: chicago-bounces at python.org
[mailto:chicago-bounces at python.org] On Behalf Of Ted Pollari
	Sent: Thursday, September 28, 2006 4:45 PM
	To: The Chicago Python Users Group
	Subject: Re: [Chicago] Trac?
	
	
	
	
	On Sep 28, 2006, at 2:45 PM, Garrett Smith wrote:


		If you have specific ideas for how you're thinking of
using Trac, post them and we can do some brainstorming on how it could
be used.


		In brief, one of the many things my group does is work
with many collaborating researchers, analysts and data-managers  in
clinical and social science research and help them manage data and other
research related tasks -- much of the time it's a byproduct of working
more efficiently ourselves =) -- one of the central themes of our work
of late has been examining and adapting tools from software development
and the open-source software environment to the university research
environment.  One great example of this is getting researchers and
bio-statisticians to do their data analyses more or less completely
programatically and version both their datasets and their data
management/manipulation code.  As fundamental as these processes may be
to many in the software development community, it's a completely new
concept to many in clincal trials and social sciences, among many other
fields -- the 'track changes' option in ms word is the closest that
anyone seems to get to versioning and even that is used inconsistently
and poorly (a straw-man, I admit as it can't be used well, IMHO ;)  With
my own experience in research psychology, I can say that solid data
management processes are just not commonly taught in these fields -- and
if they are, it's not done in any sort of formal manner.

	So, anyway, web-based collaborative tools are big and we're
looking for ways to satisfy a number of needs without becomming
full-blown plone developers or allowing the heavily sharepoint wielding
MCSE types around here make a "solution" for the research groups with
which we're affiliated.  So, anyway, trac's wiki functionality,
skinability and interconnection with SVN have been making it look like a
good option for us to move some of our projects to rather than some of
the home-grown zope based options we've been using for a few years,
which need updating and improvement.  


	So, yeah, that's the gist of why I asked.

	-Ted


			-- 
	Ted Pollari
	Research Programmer
	Department of Health Studies
	The University of Chicago
	tcp at uchicago.edu
	773.834.0559


	



More information about the Chicago mailing list