MINERVA superseeded IF/Prolog.
Please see
http://www.ifcomputer.co.jp/MINERVA
for details.
We discontinued to sell IF/Prolog Dec 31. 2003.
Dedicated technical support for IF/Prolog ended Dec 31 2008.
This site is maintained as a community service only.
This example concerns the planning of a simple project.
Tasks are only dependent on the completion of other tasks. Further analysis
reveals the ``critical path'' - those tasks in which a delay would be
critical to the completion of the project on time.
The tasks A through G comprise the planned project. Each task requires a
certain amount of time to complete and is dependent on the completion of
other tasks according to the following diagram:
:- import(const_domain).
program :- Jobs = [A,B,C,D,E,F,G],
Durations = [Da,Db,Dc,Dd,De,Df,Dg],
Durations = [ 4, 2, 4, 3, 4, 4, 3],
A ?>= 0 + Da, D ?>= B + Dd, F ?>= D + Df,
B ?>= A + Db, D ?>= C + Dd, G ?>= F + Dg,
C ?>= A + Dc, E ?>= C + De, G ?>= E + Dg,
label(Jobs), print(Jobs).
[user] ?- program.
[4,6,8,11,12,15,18]
yes
In the IF/Prolog program, inter-dependencies between the jobs (A...G)
are represented as arithmetic constraints, for example D ?>= B + Dd can
be read as Job D is completed after the start and duration of Job B.
The constraints themselves (?>=/2) do not activate to determine a solution
to the variables they contain, until they are passed to one of the solver
predicates which then solves the complete problem. The label/1 predicate
finds values for all variables in the list Jobs.
In the above, label/1 is quickly able to determine that a value for variable
A exists as its start time and duration are fixed. Then, through
``constraint propagation'', values are calculated for the remaining
variables.
| 
