1
00:00:00,080 --> 00:00:02,430
The next process model I want to discuss is evolutionary

2
00:00:02,430 --> 00:00:05,786
prototyping, which works in four main phases. We start

3
00:00:05,786 --> 00:00:08,393
from an initial concept, then we design and implement

4
00:00:08,393 --> 00:00:11,509
a prototype based on this initial concept, refine the prototype

5
00:00:11,509 --> 00:00:14,002
until it is acceptable, and finally we complete and

6
00:00:14,002 --> 00:00:17,550
release the prototype. Therefore, when developing a system using

7
00:00:17,550 --> 00:00:22,330
evolutionary prototyping, the system is continually refined and rebuilt.

8
00:00:22,330 --> 00:00:25,340
So it is an ideal process when not all requirements

9
00:00:25,340 --> 00:00:28,330
are well understood. Which is a very common situation. So, looking

10
00:00:28,330 --> 00:00:30,370
at this in a little more details, what happens is that

11
00:00:30,370 --> 00:00:33,760
developers start by developing the parts of the system that they

12
00:00:33,760 --> 00:00:37,690
understand, instead of working on developing a whole system, including parts

13
00:00:37,690 --> 00:00:40,520
that might not be very clear at that stage. The partial

14
00:00:40,520 --> 00:00:43,900
system is then shown to the customer and the customer feedback

15
00:00:43,900 --> 00:00:47,480
is used to drive the next iteration, in which either changes

16
00:00:47,480 --> 00:00:50,340
are made to the current features or new features are added.

17
00:00:50,340 --> 00:00:53,060
So, either the current prototype is improved or the

18
00:00:53,060 --> 00:00:56,270
prototype is extended. And finally, when the customer agrees that

19
00:00:56,270 --> 00:00:58,980
the prototype is good enough, the developers will complete all

20
00:00:58,980 --> 00:01:01,410
the remaining work on the system and release the prototype

21
00:01:01,410 --> 00:01:03,930
as the final product. So let's discuss as we did

22
00:01:03,930 --> 00:01:06,780
for the previous process models, what are the main advantages

23
00:01:06,780 --> 00:01:10,580
and disadvantages of evolutionary prototyping. In this case, the main

24
00:01:10,580 --> 00:01:15,440
advantage is the immediate feedback. Developers get feedback immediately as

25
00:01:15,440 --> 00:01:17,560
soon as they produce a prototype and they show it to

26
00:01:17,560 --> 00:01:21,050
the customer and therefore, the risk of implementing the wrong system is

27
00:01:21,050 --> 00:01:25,150
minimized. The main negative is the fact that it's difficult to plan.

28
00:01:25,150 --> 00:01:29,070
When using evolutionary prototype it is difficult to plan in advance how

29
00:01:29,070 --> 00:01:31,240
long the development is going to take, because we don't know how

30
00:01:31,240 --> 00:01:34,550
many iterations will be needed. And another drawback is that it can

31
00:01:34,550 --> 00:01:37,120
easily become an excuse to do kind of do cut and fix

32
00:01:37,120 --> 00:01:40,530
kind of approaches in which we hack something together, fix the main

33
00:01:40,530 --> 00:01:43,640
issues when the customer gives us feedback, and then continue this

34
00:01:43,640 --> 00:01:46,780
way, until the final product is something that is kind of

35
00:01:46,780 --> 00:01:49,830
working, but it's not really a product of high quality. Something

36
00:01:49,830 --> 00:01:51,910
else I want to point out before we move to the next

37
00:01:51,910 --> 00:01:54,490
software process model is that there are many different kinds of

38
00:01:54,490 --> 00:01:56,700
prototyping, so evolutionary prototyping is just

39
00:01:56,700 --> 00:01:58,010
one of them. For example, throwaway

40
00:01:58,010 --> 00:02:02,100
prototyping is another kind of prototyping in which the prototype is

41
00:02:02,100 --> 00:02:05,580
just used to gather requirements, but is thrown away at the end

42
00:02:05,580 --> 00:02:08,710
of the requirements gathering, instead of being evolved as it happens here.