I hate to say it but that's not a very good way to develop a schedule. As
Late As Possible means just what it says - the tasks are scheduled to
finish, hence start, as late as they possibly could without missing the
deadline. But the corollary of that method is that if a task gets delayed
for any reason it becomes impossible to meet your requirements since they're
already scheduled up against the wire. And the one thing you can count on
with absolute certainty in any real-world project is that SOMETHING is going
to be late! So scheduling backwards from required finish dates is an almost
certain guarantee that your project will fail by finishing late.
Why don't you have a valid start date to supply in the project information?
You're the project manager, it's your call. You decide that start date ...
choose one based on what you think is reasonable. Making such decisions is
why it's called "project management" - decision making is the most
fundamental aspect in the job description of "manager." So decide. Now you
might want to do some "what-if" kinds on experiments on the schedule to aid
in making that decision - roughing in the tasks, estimating their durations,
and setting the linkages that the physical process requires, scheduling
backwards from your required delivery date - this tells you the latest date
you might possibly start in a perfect world. But then you decide how much
of a cushion, how much wiggle room, you're going to need based on your
judgement of the risks and benefits, then pick a rational start date well in
advance of the latest possible date as the actual project kickoff date.
Develop your schedule by scheduling forwards from that chosen date
