Hello Julien, John, and Asa,

I appreciate your interest in new tools that help open source project management. At this point it is important to map the requirements different people have and to measure how much real public interest is there for such a project (Just say +1 if you agree but do not have anything else to contribute). IMHO this particular area is somewhat neglected in the open source community; maybe this is done in the spirit of "it needs to be free because otherwise nobody will want to attend". Whereas there probably are words of wisdom here, this just means that there are right and wrong ways to encourage people to do work that is meaningful from the project perspective. Just because project development resources might be free, and sometimes they are not like here at OSAF and at Mozilla Foundation where we have hired employees that receive regular salary, it does not mean that you should waste people's time and energy -- especially if it is given voluntarily. We all know that thousand individual workers are not able to build a pyramid but the same thousand workers under proper project management just might be able to do it and the same holds for building larger open source applications. Whereas there are a lot of really good tools, many would say even commercial quality, for revision control, issue tracking, etc. for some reason there seems to be no single tool that would address the need of project management. The tools that were mentioned above do contain a lot a valuable information but currently it is too fragmented to be of real value in seeing the big picture. Here at OSAF we do project metrics from Bugzilla as curves that show how the amount of bugs develop as a function of time. This yields a rough estimate of the remaining work for a release and hence helps in preparing timetables and work load balancing. I have prepared an automated tool for drawing such curves that can be found from our wiki page http://wiki.osafoundation.org/bin/view/Projects/BugzillaProjectManagementTool. This is however just a tip of an iceberg that could be archived in this area.

Here is a brief list of theses that I have been thinking about:
* We should have an SQL database for storing information from various sources and for making quick and easy queries. * We should have automatic data collection to the database from various information sources like issue tracking, revision control, tinderbox, mailing lists etc. * We should have a prebuild and easily extensible set of existing queries (project metrics) that answer the most typical questions in project management. * We should have some kind of data mining tool that allows exploratory data analysis of the database. This tool would be used to find interesting phenomena in the database and by that to solve questions that the user's did not even know they had. * We should have all the above usable through an easily usable user interface and project metric diagrams directly linkable from any webpage.

In my design we would have a web application that is build on Python-TurboGears-SQLAlchemy-ActiveMapper that uses any SQL database -- I currently use MySQL -- for data storage and queries, and RPy or an equivalent environment for data analysis. The automatic data collection can work either through a webinterface, where the system mimics the actions of a regular user in a browser interface and then parses the returned web pages for information (see for example twill/BeautifulSoup), or we can use the already existing libraries to access the information in these other services directly (like pysvn). This way the new system can easily be integrated to an already existing infrastructure.

I will take a look at the projects that you suggested. There is also an existing commercial tool called Deskzilla that has some part of the functionality that has been discussed. Also see my comments inline:

Cheers,
   Markku Mielityinen

Julien BETI wrote:
I do develop a little software with one feature that might partially answer your request: It updates a Imendio Planner file (Cross platform and Open Source Planning Management Software) using time stracking and status information held in Bugzilla. It also update Bugzilla with deadlines identified in the planning. It send a status report with warnings, and send notification to Bugzilla users when a plannified task is about to start. You'll find more information here: http://freshmeat.net/projects/jujunie-integration/
Regards,
Julien.

John P. Fisher wrote:

I looked at your OSA site briefly. Chandler may be the answer to my question, haven't looked at it yet.

Chandler is not a project management tool -- at least at this point.

For your question, I would build off XOOPS, or egroup etc.

But I would love to work on this application:
a Bugzilla descendant written in PHP with major support for custom fields and a more mature bug-tracking model. SVN and CVS cooperation would be built in. I would write it on top of Xoops or one of the other popular PHP frameworks.

What's "more mature bug tracking?"
Version found-in, version fixed-in
Cloned bugs across differing hardware, software
Support for branches of software
Proper user-group permissions
Project management left to other modules
Modular additions supported in the plug-in or mod model.
SVN CVS standard, others optional ( perforce etc...)

It seems that we think alike. The only difference is the implementation platform as I have learned to like Python and do think that I is better able to meet the various needs in this kind of a project.


thanks

John
Bugzilla customizer,build manager and so forth


Markku Mielityinen wrote:
Dear Group,

Does anyone know if there is an open source project management tool that can import project information from Bugzilla (and maybe from Subversion) and preferably supports some project management metrics? We may be interested in developing one...

Best Regards,
   Markku Mielityinen
   Open Source Applications Foundation
   http://www.osafoundation.org/
_________________________________


_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Open Source Applications Foundation "General" mailing list
http://lists.osafoundation.org/mailman/listinfo/general

Reply via email to