1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
|
1
00:00:00,100 --> 00:00:03,060
There are some important considerations I want to make on final state
2
00:00:03,060 --> 00:00:06,410
machines. And more in general, on model based testing. The first one
3
00:00:06,410 --> 00:00:10,600
is about applicability. Testing based on final state machines is a very
4
00:00:10,600 --> 00:00:13,890
general approach, that we can apply in a number of contexts. And in
5
00:00:13,890 --> 00:00:16,510
particular, if you are working with UML, you have state machines for
6
00:00:16,510 --> 00:00:19,720
free. Because state charts are nothing else but a special kind of
7
00:00:19,720 --> 00:00:22,210
state machine. So you can apply the technique that we just saw
8
00:00:22,210 --> 00:00:26,010
directly on state charts, and try to cover their states and their transitions.
9
00:00:26,010 --> 00:00:29,980
Another important point is that abstraction is key. You have to find the
10
00:00:29,980 --> 00:00:33,280
right level of abstraction. The bigger the system, the more you have to
11
00:00:33,280 --> 00:00:36,130
abstract if you want to represent it with a model, and in particular, with
12
00:00:36,130 --> 00:00:38,710
the final state machine. So it's like having a slider, and you have
13
00:00:38,710 --> 00:00:41,840
to decide where you want to move on that slider. The more you represent,
14
00:00:41,840 --> 00:00:44,700
the more complex your system is going to be and the more thorough your
15
00:00:44,700 --> 00:00:48,110
testing is going to be but also more expensive. The less you represent the
16
00:00:48,110 --> 00:00:51,330
less expensive testing is going to be, but also testing might not be as
17
00:00:51,330 --> 00:00:53,500
thorough as it would be otherwise. So you have to find
18
00:00:53,500 --> 00:00:56,150
the right balance between abstracting the weight too much and abstracting
19
00:00:56,150 --> 00:00:59,840
the weight too little. And finally there are many other approaches.
20
00:00:59,840 --> 00:01:02,840
So we just scratched the surface, and we just saw one possible
21
00:01:02,840 --> 00:01:05,760
approach. But for instance, other models that you can use are
22
00:01:05,760 --> 00:01:09,780
decision tables, flow graphs and even historical models. Models that can
23
00:01:09,780 --> 00:01:13,560
guide your testing based on problems that occurred in your system
24
00:01:13,560 --> 00:01:16,500
in the past. And also, in this case, I'm going to put pointers
25
00:01:16,500 --> 00:01:18,460
to additional materials in the class notes.
|
