Task Priority to 1000 or constraint on end date

[Javier]

Hi all,
I have build a project plan with 50 independant subprojects not linked
between them but sharing resources. Some of the subprojects can not be
delayed and some can. My question is: for leveling purposes, is it better to
put a date constraint "Finish no later than" on the final task for the
subprojects that cannot be delayed, or set Priority 1000 to this task. Also
if all the tasks are linked with a "finsih to start" relationship is it
enough to set the priority to 1000 on the last task or should I do it in all
tasks.
Thanks in advance for your help

 

[Jan De Messemaeker]

Hi Javier,

First and foremost, Finish no later then will be respected by leveling only
of the option "Level only within available slack" is ON
Tis being said, in that case I would think a constraint is more realistic
than a 1000 priority because if necessary your subproject can be delayed
till its latest date if necessary; with 1000 priority it will not be moved
at all.
If teh option about available slack is ON, it is enough to fix the final
task becasuse when the project becomes critical all tasks will have total
slack 0 because of the FS links and will no longer be moved.

Hope this helps,

--
Jan De Messemaeker
Microsoft Project Most Valuable Professional
+32 495 300 620
For availability check:
http://users.online.be/prom-ade/Calendar.pdf

 

[Steve House]

What factors would cause a task be scheduled past some arbitrarily chosen
completion date? Either the predecessors that physcially MUST complete
before the task in question can start will take too long to do or the
resources required for the tasks won't be available. These are physical
processes, not something that we get to vote on whether we'll obey or not.
Our delivery requirement is the 10th on July but the predecessor work won't
even be finished until August 1st - does it do us any good to publish a
schedule that promises delivery on July 10th? If you prevent the scheduling
engine from moving the task past that date by setting a constraint or
setting a priority or whatever other method you devise, all you've really
done is cripple Project's ability to warn you of impending problems early
enough to fix them. We can't arbitrarily establish dates and force the
schedule to print them - we have to manipulate the physical forces such as
resource distribution, etc, that drive the task's dates until they come into
compliance with our business needs. In my example, you don't lock the
completion task to July 10th - rather you work on fixing the problems making
the predecessors take so long so it CAN finish on the 10th.