From: vincenegri (vince.negri_at_gmail.com)
Date: Tue Nov 15 2005 - 13:19:17 CET
From: vincenegri <vince.negri@gmail.com> Message-ID: <181f.4379d245@www.taskjuggler.org> Date: Tue, 15 Nov 2005 13:19:17 +0100 (CET) Subject: [taskjuggler-devel] Meaning of "duration"?
The behaviour of the "duration" column in the interactive reports seems strange to me, but I'm willing to believe there is an explanation.
With the default working day, (9am to 6pm, one hour lunch) if I create a task with defined effort of 1 day, then the duration column shows "9h". That makes sense, since the elapsed time from start to end of the task is 9 hours.
If I create a 2d effort task (so that it starts at 9am on one day and finishes at 6pm on the next day) then the "duration" column says "11h". So... 8 hours of extra effort result in 2 hours extra duration?
Checking the source code, the duration column output is obtained by getting the task's getCalcDuration(), which is a value in *calendar* days, and then passing it to scaledLoad(). I think that's where the issue creeps in, because scaledLoad expects input in man-days. A task that runs from 9am Monday to 6pm Tuesday lasts (in calendar time) ((24+9)/24) = 1.375 calendar days. If you suddenly pretend that is man-days, then it is equivalent to (1.375 * 8 ) = 11 hours.
I did an experiment where I created a new function unscaledLoad(). This function performs the same "work out the best units to use" as scaledLoad() but based on calendar days, not man-days. I then changed the "duration" output to use unscaledLoad() instead of scaledLoad(). The duration column now shows the calendar duration of the task, which is what I wanted.
But maybe the current implementation is intentional? If so, what is the meaning of "duration" and how would it be used in planning?
-- To unsubscribe, email: taskjuggler-devel-unsubscribe@suse.com For additional commands, email: taskjuggler-devel-help@suse.com
This archive was generated by hypermail 2.1.7 : Tue Nov 15 2005 - 13:19:20 CET