[reportlab-users] Trac

Jonas Borgström reportlab-users@reportlab.com
Thu, 29 Apr 2004 21:46:41 +0200

Andy Robinson wrote:

>>If you're using Subversion and want a wiki, I'd recommend Trac
>><http://projects.edgewall.com/trac/> - it's an integrated wiki,
>>Subversion repository browser and bug tracking system written in
>>Python with a great user interface.  I've been using it at home and
>>work for a month or so and love it.
> Could I ask you to save us a few hours by answering some
> questions?  Setting it up and playing won't happen before the weekend
> due to the subversion 'excitement'.


Nicholas pointed me to this mail, I'll do my best to answer your
questions. (please cc me if you reply to this mail, I'm not
subscribed to this list)

> - does one database and CGI have multiple projects, or do you
> set up one CGI per project?
One Trac installation (on disk) can be used for multiple projects.
Each project has their own wiki pages and tickets and is associated
with a subversion repository (but multiple trac projects can use the
same repository).

> - could it also be used in a design stage to quickly map
> out subprojects and tasks and register all of them as to-do
> items?
I don't think I understand this question...

> - we already have a MySQL support database.  Any idea if there are 
> deep fundamental ties to SQLite or if this would be a relatively
> minimal change?
There isn't any fundamental ties to sqlite. Although Trac won't
currently work out of the box with mysql, Trac has been ported to
postgresql as a proof of concept but these changes haven't been
merged into trunk yet. The bottom line is: Trac will become database
independent in the future but because sqlite works very well for this
type of application this hasn't been our highest priority.

> - is there any kind of security model so I could let let
> BigBank employees see the BigBank/Statements project
> wiki and issue list,  without even being aware of the existence 
> of the rival HugeBroker/Statements project we are working on
> at the same time? ;-)
> Rationale:  an integrated system which works for open source
> users and commercial customers on projects, and which keeps all 
> our staff issues in one database will be a big time saver.
> Creating separate databases for every project might not be 
> (although we could do it for the core code...)
Trac lets apache take care of the actual authentication so you
can authenticate using any authentication method apache supports.
Each Trac project has a set of permission that control what users
are allowed to do or see. But it's currently not fine-grained
enough to configure different permissions for different wiki pages
or tickets in the same Trac project.

I hope I was able to answer a few of your questions.

/ Jonas
Jonas Borgström    | Edgewall Software
jonas@edgewall.com | Professional GNU/Linux & Open Source Consulting.
                    | http://www.edgewall.com/