Re: [Planner Dev] Duration Calculation



On Wed, 2006-05-31 at 12:43 -0700, Andrew De Ponte wrote:
> Hi all. I have a question really. I have checked out the latest code
> from the gnome cvs server and have built and installed it and noticed
> that the Work time for summary tasks calculation has been corrected to
> be the summation of the work of the sub-tasks which makes complete sense.
> 
> However, I also noticed that the Duration calculation does NOT seem
> correct, well to me at least. In my opinion the Duration should be
> calculated by adding up durations of the subtasks which share a specific
> resource (this will no doubt require using the unit to calculate
> percentage of the work that a specific resource is assigned). Then after
> doing that compare that total to the duration of each non associated
> task and use the largest duration out of them all for the summary task's
> duration.
> 
> Please let me know what all your inputs is on this. I have looked at the
> code a little and see the new addition
> 
> mrp_task_manager_calculate_summary_duration()
> 
> in revision 1.18 however as the comments above the function denote it
> ignores assignments.
> 
> Why is it ignoring the assignments? The reason this is problematic for
> me is because the gantt chart uses the durations so that a visulazition
> of the time needed for certain tasks can be seen. In this case it is
> ignoring the fact that if a single resource is applied to two different
> tasks the duration is going to be the sum of the tasks that resource is
> assigned to, due to the fact that that one resource can't do both things
> at once.

Well, the trick really isn't to get the summary duration to understand
the resource assignments, but to implement resource leveling so that
tasks can be rescheduled automatically when there's a resource conflict.
Once the conflicts are resolved, the duration becomes accurate.  Right
now you need to do this manually using the resource usage view, and
either create relationships between the conflicting tasks or set
constraints such as a 'start no earlier than' date.

Here's the bug on this subject:
http://bugzilla.gnome.org/show_bug.cgi?id=132917

-- 
Kurt Maute <kurt maute us>




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]