Description
Part 1: Estimation
Your team has successfully developed 8 systems in the last 2 years, all in the
same domain: e-Commerce applications, using over the years 3 different
programming languages+technology (noted A, B and C). You’ve collected some
information about these 8 projects, as part of your attempts to do Function Point
Analysis:
# of inputs
# of outputs
# database entities accessed
# of system users
effort in person-days
# of pages in the final users guide
programming language & toolkit (A, B, or C)
Project Inputs Outputs Entities Users Pages Lang. Effort
1 210 420 40 10 35 A 30
2 469 1406 125 20 10 A 85
3 513 1283 76 18 9 B 108
4 660 2310 88 200 75 B 161
5 183 367 35 10 5 C 22
6 244 975 65 25 32 C 42
7 1600 3200 237 25 12 B 308
8 582 874 111 5 3 C 62
X 180 350 40 20 B ? ??
Y 484 1190 69 35 B ? ??
You are now planning to develop two new projects, X and Y. From their current
descriptions — a requirement document– you have derived some estimates of
the number of input, output and data entities.
Your task is to estimate the effort for these 2 projects X and Y.
a) what are the parameters that seem to drive the productivity (= Function Point
per person-day) in our team?
b) what is the productivity (in FP per person-day) for language A, B and C?
c) what would be the estimated effort for X and Y using a Function Point
approach?
d) what would be the estimated effort for X and Y using just an analogy
approach (similar project, or “close enough” project)?
e) Right now projects X and Y are planned to be done in language B? Should
you change?
Notes:
1) An unadjusted Mark II function point (FP) count can be estimated by the
formula
FP = 0.58 x (#of inputs) + 1.66 x (#of DB entities) + 0.26 x (#of output)
2) The Euclidean distance D between 2 projects u and v, for which you have a
number n of significant parameters a, b, c, …n is:
D = (au − av )
2 + (bu − bv )
2 + …+ (nu − nv )
2
Source: Hughes and Cotterrell 2003
Part 2: Scheduling
Your project consists of 8 major subprojects or tasks, for which you now have
some estimates in person-days, and some dependencies. By dependencies, we
mean that (in the table below) task T6 can only start after T3 and T4 are
completed. Assume you cannot split tasks between individual developers (that
is, have 2 persons or more working together on the same task).
Task # Duration Dependencies
T1 8 days None
T2 4 d None
T3 6 d T4
T4 4 d None
T5 10 d T1, T6
T6 4 d T3, T4
T7 6 d T6
T8 2 d T2
a) Identify the sequence of tasks that constitutes the critical path.
Suggestion: Draw a project activity network (i.e., a directed graph with
tasks on the arcs) for this project.
b) What is the minimal time required to complete this project?
c) How much “slack” does task T1 have? (in other words: How late can it be
started or delayed during execution without affecting the final delivery
date?)
d) How many people at minimum do you need to complete the project in the
minimum time you gave in (b). Show one possible task allocation.
e) The developer of task T7 reports one morning that it will take her twice
the expected time to complete: 12 days instead of 6. How does this affect
the critical path, and what is the new time to complete the project?
Submit you assignment in PDF via Vista by Tuesday February 7th at 1:00pm
and place a printed copy in the mailbox located in McLeod 4th floor, between
rooms MCLD422 and 426. This is an individual assignment.
