1
00:00:06,837 --> 00:00:08,678
- So let's talk about some ways

2
00:00:08,678 --> 00:00:12,845
we can reduce the time it
takes to complete the end game.

3
00:00:13,988 --> 00:00:16,591
One of the things that I like
to do is use retrospectives,

4
00:00:16,591 --> 00:00:19,563
and we mentioned that before, is to say,

5
00:00:19,563 --> 00:00:23,736
"What were some of the problems,
and how did we do that?"

6
00:00:23,736 --> 00:00:26,276
Lisa, I know that you have at least

7
00:00:26,276 --> 00:00:28,635
one story to contribute to this.

8
00:00:28,635 --> 00:00:30,183
- Many, many stories, but this is

9
00:00:30,183 --> 00:00:32,561
the most recent story on my team.

10
00:00:32,561 --> 00:00:34,890
We really would like to go to

11
00:00:34,890 --> 00:00:38,723
a daily delivery to
production, and currently,

12
00:00:39,604 --> 00:00:42,132
our manual regression
checking that we have

13
00:00:42,132 --> 00:00:44,659
tens of thousands of automated tests,

14
00:00:44,659 --> 00:00:47,135
we still have a lot of checklists

15
00:00:47,135 --> 00:00:49,480
to go through manually at the end,

16
00:00:49,480 --> 00:00:52,934
and that could take us
anywhere from 45 minutes

17
00:00:52,934 --> 00:00:55,812
to longer if we find some problems,

18
00:00:55,812 --> 00:00:57,832
and we don't want to do that every day,

19
00:00:57,832 --> 00:01:01,165
so we took a step back and thought about

20
00:01:02,624 --> 00:01:05,267
what risks do we want to
make sure that we cover,

21
00:01:05,267 --> 00:01:07,529
we drew our architecture on a whiteboard,

22
00:01:07,529 --> 00:01:09,136
we started looking at the areas

23
00:01:09,136 --> 00:01:11,275
we thought were the most
risky, and we put them

24
00:01:11,275 --> 00:01:13,598
in a spreadsheet and prioritized them,

25
00:01:13,598 --> 00:01:16,534
and we've written stories to make sure

26
00:01:16,534 --> 00:01:19,291
we have automated test coverage
for these different things,

27
00:01:19,291 --> 00:01:21,219
so in some cases we might
have to write some new tests,

28
00:01:21,219 --> 00:01:23,123
we're using a bit of a different approach,

29
00:01:23,123 --> 00:01:25,554
and in some cases we can
still use our old tests,

30
00:01:25,554 --> 00:01:29,768
so we can take advantage of
that, but we really turn that

31
00:01:29,768 --> 00:01:32,335
into actual feature
development, if you will.

32
00:01:32,335 --> 00:01:33,326
- Okay.

33
00:01:33,326 --> 00:01:35,581
That really makes a lot of sense.

34
00:01:35,581 --> 00:01:36,801
By the way, I really like the way

35
00:01:36,801 --> 00:01:38,550
you use your virtual whiteboard, too.

36
00:01:38,550 --> 00:01:40,558
It makes me feel much better.

37
00:01:40,558 --> 00:01:44,076
So the very first day I
walked into this one team,

38
00:01:44,076 --> 00:01:47,233
they had just released to the customer

39
00:01:47,233 --> 00:01:49,606
to do their user acceptance testing.

40
00:01:49,606 --> 00:01:53,336
I walked in and I start
asking my questions, and I go,

41
00:01:53,336 --> 00:01:56,156
"So when are we going to get it back

42
00:01:56,156 --> 00:01:58,748
"from the user acceptance
testing so that we can

43
00:01:58,748 --> 00:02:01,183
"get it ready to put into production?"

44
00:02:01,183 --> 00:02:04,372
And they said, "Six weeks,"
and I went, "Six weeks?

45
00:02:04,372 --> 00:02:06,684
"But our release cycle is three months.

46
00:02:06,684 --> 00:02:08,801
"We're going to be halfway through."

47
00:02:08,801 --> 00:02:12,968
It made no sense to me, and so
I asked why it took so long.

48
00:02:14,023 --> 00:02:17,055
What the problem was was
that the person who was doing

49
00:02:17,055 --> 00:02:20,276
the user acceptance
testing was fitting it into

50
00:02:20,276 --> 00:02:24,158
her regular job, and it
turned out that she was doing

51
00:02:24,158 --> 00:02:27,132
bits and pieces, because
it was her regular job

52
00:02:27,132 --> 00:02:29,145
that she was trying to work around.

53
00:02:29,145 --> 00:02:30,498
Well, I came up with an idea.

54
00:02:30,498 --> 00:02:32,809
We started talking about
it in our retrospective,

55
00:02:32,809 --> 00:02:36,829
and I said, "How about," and thankfully,

56
00:02:36,829 --> 00:02:39,320
everybody bought into
it, including Julie--

57
00:02:39,320 --> 00:02:43,487
that was our UAT person-- and
so, every Friday afternoon,

58
00:02:44,417 --> 00:02:48,041
Julie started coming for an
hour, one hour every week,

59
00:02:48,041 --> 00:02:50,647
and she would pair with
me, and I would show her

60
00:02:50,647 --> 00:02:52,916
what we had been doing for the week,

61
00:02:52,916 --> 00:02:55,582
so at the end of the next cycle,

62
00:02:55,582 --> 00:02:59,419
when we went to give it to
her, she said, "You know what,

63
00:02:59,419 --> 00:03:02,103
"I don't need six weeks,
I only need three weeks,"

64
00:03:02,103 --> 00:03:04,621
because she was much more
confident, and I went,

65
00:03:04,621 --> 00:03:07,081
"Not good enough, not good enough,"

66
00:03:07,081 --> 00:03:09,285
so we talked a little bit more about it.

67
00:03:09,285 --> 00:03:13,627
What we did was we brought her
own desk, her own computer,

68
00:03:13,627 --> 00:03:15,498
and every Friday afternoon, she would come

69
00:03:15,498 --> 00:03:18,031
and she would spend a couple of hours

70
00:03:18,031 --> 00:03:21,540
and test herself what we had been doing.

71
00:03:21,540 --> 00:03:24,829
When we asked her the next release cycle,

72
00:03:24,829 --> 00:03:28,624
how long she needed, her answer was,

73
00:03:28,624 --> 00:03:33,237
"I will take a day off work,
put it in our test system,

74
00:03:33,237 --> 00:03:36,314
"and I will finish it
by the end of the day,"

75
00:03:36,314 --> 00:03:39,582
the reason being that she
had confidence in what it was

76
00:03:39,582 --> 00:03:40,763
because she had been doing it,

77
00:03:40,763 --> 00:03:43,263
so don't wait to the end game.

78
00:03:44,551 --> 00:03:47,640
If you can bring that kind of
testing in earlier, please do.

79
00:03:47,640 --> 00:03:50,373
It can make all the
difference in the world.

80
00:03:50,373 --> 00:03:52,545
Let's continue with a summary now.

81
00:03:52,545 --> 00:03:55,631
This lesson was about
planning for testing.

82
00:03:55,631 --> 00:03:58,306
It's important to understand your context.

83
00:03:58,306 --> 00:04:00,479
How much is enough?

84
00:04:00,479 --> 00:04:03,545
- [Lisa] We talked about
models to get the whole team

85
00:04:03,545 --> 00:04:07,479
thinking and provide visuals
for our quality attributes

86
00:04:07,479 --> 00:04:09,781
and the testing that we want to get done.

87
00:04:09,781 --> 00:04:12,945
We talked about how the
agile testing quadrants

88
00:04:12,945 --> 00:04:15,338
help us think about all
different types of testing

89
00:04:15,338 --> 00:04:17,899
beyond just the functional
testing that we most likely

90
00:04:17,899 --> 00:04:20,021
are going to think about first,

91
00:04:20,021 --> 00:04:23,200
and we talked about planning
release activities early

92
00:04:23,200 --> 00:04:26,196
and doing what's
necessary in the end game.

93
00:04:26,196 --> 00:04:28,009
- [Host] Next, we'll take about ways

94
00:04:28,009 --> 00:04:32,176
to build shared understanding
through early testing.