forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Bart Guijt" <b.gui...@chello.nl>
Subject Re: [RT] my Forrest dream-list
Date Thu, 21 Feb 2002 20:20:24 GMT
From: "Stefano Mazzocchi" <stefano@apache.org>
> Ok guys,
>
> I think we have enough community critical mass to make this happen, but
> one thing is for sure: I wanted this project and I want it to be
> focused.
>
> So I need to give more focus.
>
> I call this a 'dream-list' because a wish-list is not enough, but as Sam
> once stated, "some people have no vision" and even if I believe the
> people that subscribed here are those *with* the vision, it doesn't hurt
> to throw more dreams in and get people excited.

<snipped lots of beautiful stuff>

Hi,

I'd like to jump in and propose another idea which is probably significant
in terms of structuring data for graphs, reports and other stuff.

        Dream: Create a datawarehouse which stores every single bit of
information somehow related to the software project.

There are weblogs, CVS commit messages, JUnit test results, code metrics and
perhaps other sources of information. All this information applies to the
following project 'level' dimension:
- Project itself (weblogs)
- directory / java package (CVS commit msgs, test results, metrics)
- file / java class (CVS commit msgs, test results, metrics)
- java class member (test results, metrics)

And so 'time' is another dimension:
- year
- quarter
- month
- day
- ...

So we've got two dimensions structuring project data, aggregated at each
level. This should create huge capabilities for report and graph designers,
since all information is cleanly structured in a (relational) database
rather than formatted output from some obscure logging tool.

Thoughts?




Mime
View raw message