Christopher Lenz wrote:
Am 21.01.2006 um 18:20 schrieb Justin Francis:
Three months ago I created a patch that provided an extension point
for milestone statistics so that plugins could add or change the way
statistics are calculated. We use this at work currently to use time
spent on tickets as opposed to ticket counts for statistical
information.
Is there a reason why the patch I submitted in ticket #2314 has
still not been applied to the repository? Is it a bad idea? Is it
not useful? Is it a time thing?
At this point, it is probably too late to apply the patch as it is
outdated and many changes have been applied to milestones in the
last three months. I don't want to generalize, but it seems like
this development team is not very open to contributions. It is very
discouraging, and I personally believe that trac development could
be much faster, but the core team is very hesitant in growing the
project.
I am not talking about new features here -- I am talking about
adding extension points. I think that if the team is serious about
not wanting the feature set to grow huge, it is critical to have as
many extension points as there are customizable aspects of trac.
As the developer most familiar with the roadmap code, this is mostly
my fault... at the time when you sent your patch, I was busy on my
thesis, and had very little time for reviewing patches and the like.
Sorry about the lack of feedback there.
Anyway, the case with the roadmap is that the code is pretty awful (I
can say that, I wrote it). The accumulation of ticket statistics is
pretty inefficient, and my impression at the time was that your patch
didn't address that concern, but rather made the inefficiency part of
the extension API.
I assume you are referring to the fact that the API defines a function
that takes a list of ticket objects and returns a data structure with
the stats information.
You are correct that this does not address the inefficiency, but I am
not convinced that is a valid concern. How efficient should it be? I
have not heard complaints about its current efficiency. There is only
one hit to the database for each milestone (and this can be improved by
some intelligence in processing the query).
Moreover, I thought that the general idea was to move toward an
object-oriented domain model architecture for the project. I can't see a
more efficient way of gathering stats info without coupling the
extension point more tightly to the app (like directly to the database).
This is the trade-off of inefficiency vs maintainability.
Justin
Which doesn't mean the patch was bad... that was just a concern I had
when looking over the code without much time.
Cheers,
Chris
--
Christopher Lenz
cmlenz at gmx.de
http://www.cmlenz.net/
_______________________________________________
Trac-dev mailing list
[email protected]
http://lists.edgewall.com/mailman/listinfo/trac-dev
_______________________________________________
Trac-dev mailing list
[email protected]
http://lists.edgewall.com/mailman/listinfo/trac-dev
