The purist in me comes out. First, Julie's post will absolutely get you what
you need, and I use this method -- a lot. Permit me a moment of
pontification.
By using negative lag, you actually imply with absolute certainy that the
following task (starting the in future) will happen simply becuase the
predecessor takes place. Here is an example: Meeting Agenda to be sent 15
days prior to meeting. Using negative lag, if I submit an agenda the meeting
will take place. Even if none of the drawings are ready for reveiw, etc.
What I try to do.... Use the deadline date. Set a deadline on "Send
Agenda" to 15 days prior to "Conduct Meeting". You can use the deadline
field for the date. Unfortunately it is a manual entry field so if your
meeting date is driven to the left, you will need to revise your deadline
date. Same if the meeting moves to the right.
This way, you can use proper schedule logic and let other tasks drive the
start of the meeting such as "prepare drawings." With a little thought, you
can also say, "Once I have the list of drawings available, I know it will
take me 25 days to complete the drawings. Then I can set the send agenda task
milestone to 'List of drawings avaialble FS+10 days. Then Agenda Sent can
have the successor "conduct meeting" as FS+15d. Now what happens is the
meeting will not occurr until at least 15 days after the agenda is sent...
the task "make the drawings" will also drive the "conduct meeting" task and
could push it out beyond the 15 days.
Anyway, the "best practice" does try to avoid negative lag. However, we all
use it all the time.
--
If this post was helpful, please consider rating it.
Jim Aksel, MVP
Check out my blog for more information:
http://www.msprojectblog.com