1
00:00:00,120 --> 00:00:03,220
After recapping the three major issues that characterize a software crisis

2
00:00:03,220 --> 00:00:05,626
let's see what was the evidence that there was indeed a

3
00:00:05,626 --> 00:00:07,900
crisis. So what I want to discuss now is the result

4
00:00:07,900 --> 00:00:11,060
of a study performed by Davis in 1990s. So in even

5
00:00:11,060 --> 00:00:13,670
more recent times than the 60s and the 70s. And the

6
00:00:13,670 --> 00:00:17,280
study was performed on nine software projects that were totaling a

7
00:00:17,280 --> 00:00:20,990
cost around $7 million and I'm going to show you how this

8
00:00:20,990 --> 00:00:25,190
projects went using this representation, this pi representation, in which I'm

9
00:00:25,190 --> 00:00:27,520
going to discuss what each of the segment of the

10
00:00:27,520 --> 00:00:30,010
pi represent. So let's start looking at the first one.

11
00:00:30,010 --> 00:00:32,920
This is a software that was usable as delivered. Other

12
00:00:32,920 --> 00:00:36,590
software was delivered, and usable, either after some changes or

13
00:00:36,590 --> 00:00:41,080
after some major modifications, so within additional costs involved.

14
00:00:41,080 --> 00:00:43,530
But the striking piece of information here is that the

15
00:00:43,530 --> 00:00:46,890
vast majority of the software, so these two slices, were

16
00:00:46,890 --> 00:00:50,250
software that was either delivered but never successfully used or

17
00:00:50,250 --> 00:00:53,730
software that was not even delivered. And this corresponded

18
00:00:53,730 --> 00:00:57,500
to five over the seven total million dollars for

19
00:00:57,500 --> 00:01:00,050
all the projects. So clearly, this shows a pretty

20
00:01:00,050 --> 00:01:03,910
grim picture for software development and its success. In short,

21
00:01:03,910 --> 00:01:06,410
there was clear evidence the software was becoming to

22
00:01:06,410 --> 00:01:08,990
difficult too build and that the software industry was facing

23
00:01:08,990 --> 00:01:11,190
a crisis. And this is what led to the

24
00:01:11,190 --> 00:01:15,130
NATO Software Engineering Conference that was held in January 1969,

25
00:01:15,130 --> 00:01:19,100
which is what we can consider the birth of software engineering. And what

26
00:01:19,100 --> 00:01:23,080
I'm showing here is a drawing of the proceedings for that conference. And if

27
00:01:23,080 --> 00:01:26,020
you look at the class notes you can see a link to the actual

28
00:01:26,020 --> 00:01:27,640
proceedings, in case you are interested in

29
00:01:27,640 --> 00:01:29,180
looking at the issues that were discussed.