1
00:00:12,320 --> 00:00:16,880
welcome back everyone we're ready for

2
00:00:14,559 --> 00:00:18,400
the first talk of the day here in the

3
00:00:16,880 --> 00:00:19,920
devops track

4
00:00:18,400 --> 00:00:22,880
and the first one we have today is an

5
00:00:19,920 --> 00:00:24,240
amazing talk by jonathan hall about how

6
00:00:22,880 --> 00:00:26,240
a single bit

7
00:00:24,240 --> 00:00:27,760
changed his career i've been looking

8
00:00:26,240 --> 00:00:29,920
forward to this talk for a while and i

9
00:00:27,760 --> 00:00:31,039
can hardly wait to hear what he has to

10
00:00:29,920 --> 00:00:33,360
tell us

11
00:00:31,039 --> 00:00:35,440
so without further ado let's kick off

12
00:00:33,360 --> 00:00:37,680
the devops track we talked by jonathan

13
00:00:35,440 --> 00:00:37,680
hall

14
00:00:40,000 --> 00:00:44,559
hi everyone i'm jonathan hall thank you

15
00:00:42,239 --> 00:00:46,239
for attending my presentation

16
00:00:44,559 --> 00:00:49,520
uh i hope you enjoy it let me jump over

17
00:00:46,239 --> 00:00:49,520
to the slides and we'll get started

18
00:00:50,000 --> 00:00:53,120
so as you all saw from the program the

19
00:00:51,760 --> 00:00:56,160
title of my talk of course is the

20
00:00:53,120 --> 00:00:59,680
butterfly effect uh how a single bit

21
00:00:56,160 --> 00:01:02,160
changed my career

22
00:00:59,680 --> 00:01:04,400
my agenda for this talk of course is uh

23
00:01:02,160 --> 00:01:06,320
to tell you the story of how i

24
00:01:04,400 --> 00:01:08,560
uh sort of started my career and in

25
00:01:06,320 --> 00:01:10,799
software delivery

26
00:01:08,560 --> 00:01:12,720
i want to encourage you on your own path

27
00:01:10,799 --> 00:01:14,080
in that regard to whatever extent you

28
00:01:12,720 --> 00:01:15,280
were involved in software delivery which

29
00:01:14,080 --> 00:01:16,960
i expect

30
00:01:15,280 --> 00:01:19,600
most of you are as you're probably

31
00:01:16,960 --> 00:01:21,520
developers of some sort

32
00:01:19,600 --> 00:01:22,799
and then maybe give you some pointers on

33
00:01:21,520 --> 00:01:25,680
how you can improve in some of these

34
00:01:22,799 --> 00:01:27,680
areas if that's interesting for you

35
00:01:25,680 --> 00:01:30,720
of course my name is jonathan hall and

36
00:01:27,680 --> 00:01:32,320
uh i am uh the tiny devops guy if you

37
00:01:30,720 --> 00:01:34,560
google for that you'll find me

38
00:01:32,320 --> 00:01:36,960
i have a podcast by that name

39
00:01:34,560 --> 00:01:38,720
and uh i do devops consulting primarily

40
00:01:36,960 --> 00:01:40,960
for small companies

41
00:01:38,720 --> 00:01:44,159
i'm i'm a lifelong developer uh

42
00:01:40,960 --> 00:01:46,799
operations person uh i now do coaching

43
00:01:44,159 --> 00:01:49,119
i'm a trekkie a podcast host i also

44
00:01:46,799 --> 00:01:50,880
dance salsa i'm a father many many other

45
00:01:49,119 --> 00:01:53,280
things i can say about myself but you're

46
00:01:50,880 --> 00:01:54,720
not here to learn about that i i think

47
00:01:53,280 --> 00:01:57,280
most of you are more interested in the

48
00:01:54,720 --> 00:01:59,119
technical part of my career so

49
00:01:57,280 --> 00:02:01,119
i started i got into computers actually

50
00:01:59,119 --> 00:02:04,159
at the age of eight uh when i started

51
00:02:01,119 --> 00:02:06,079
programming on my dad's commodore 64.

52
00:02:04,159 --> 00:02:09,440
uh and for many years i just played

53
00:02:06,079 --> 00:02:11,120
around made games in various uh silly

54
00:02:09,440 --> 00:02:13,920
little programs on the computer

55
00:02:11,120 --> 00:02:14,720
and then in 2006 is when i got my first

56
00:02:13,920 --> 00:02:16,800
uh

57
00:02:14,720 --> 00:02:18,879
professional programming job

58
00:02:16,800 --> 00:02:21,520
and the the interview was interesting

59
00:02:18,879 --> 00:02:23,440
the the owner of this small company

60
00:02:21,520 --> 00:02:25,520
drew three boxes on a whiteboard and

61
00:02:23,440 --> 00:02:27,520
said these boxes represent the three new

62
00:02:25,520 --> 00:02:30,000
servers we just bought

63
00:02:27,520 --> 00:02:32,959
we would like to hire somebody maybe you

64
00:02:30,000 --> 00:02:35,599
to come in and build these servers out

65
00:02:32,959 --> 00:02:37,040
to run our spam filtering software

66
00:02:35,599 --> 00:02:38,400
and up until that point their spam

67
00:02:37,040 --> 00:02:40,239
filtering software that they that they

68
00:02:38,400 --> 00:02:42,000
had written in-house

69
00:02:40,239 --> 00:02:44,000
they would install it on an appliance

70
00:02:42,000 --> 00:02:45,120
meaning like a 1u server along with the

71
00:02:44,000 --> 00:02:47,200
operating system and they would ship

72
00:02:45,120 --> 00:02:48,560
that to the customer's

73
00:02:47,200 --> 00:02:50,480
office and the customer would install

74
00:02:48,560 --> 00:02:54,239
this physical server next to their

75
00:02:50,480 --> 00:02:56,959
probably physical exchange server

76
00:02:54,239 --> 00:02:58,879
and they had the idea of switching to a

77
00:02:56,959 --> 00:03:00,480
cloud-based model what we now call sas

78
00:02:58,879 --> 00:03:02,400
or software as a service it was

79
00:03:00,480 --> 00:03:03,680
relatively new at that point

80
00:03:02,400 --> 00:03:06,319
so that's what i was hired to do i was

81
00:03:03,680 --> 00:03:08,000
hired to uh take this existing spam

82
00:03:06,319 --> 00:03:10,159
filtering software and turn it into a

83
00:03:08,000 --> 00:03:11,920
sas that we could sell uh as a single

84
00:03:10,159 --> 00:03:13,840
installation rather than all of these

85
00:03:11,920 --> 00:03:15,360
countless installations we had

86
00:03:13,840 --> 00:03:17,440
physically installed around the world

87
00:03:15,360 --> 00:03:19,120
the author of this spam filtering

88
00:03:17,440 --> 00:03:21,760
software had just left for for another

89
00:03:19,120 --> 00:03:23,680
job uh up until that point he had been

90
00:03:21,760 --> 00:03:26,000
the only software developers technical

91
00:03:23,680 --> 00:03:26,959
support person network engineer all of

92
00:03:26,000 --> 00:03:28,480
that

93
00:03:26,959 --> 00:03:30,000
in one person

94
00:03:28,480 --> 00:03:31,920
and so when he left they decided to

95
00:03:30,000 --> 00:03:34,000
replace him with with two they hired a

96
00:03:31,920 --> 00:03:37,200
developer to take over the day to day

97
00:03:34,000 --> 00:03:38,799
and then hired me uh to to help turn it

98
00:03:37,200 --> 00:03:41,440
into a sass

99
00:03:38,799 --> 00:03:43,840
jump ahead just a month and after i had

100
00:03:41,440 --> 00:03:46,000
started and it was thanksgiving day and

101
00:03:43,840 --> 00:03:47,360
uh for those of you who have not

102
00:03:46,000 --> 00:03:48,640
experienced thanksgiving day in the

103
00:03:47,360 --> 00:03:50,080
united states let me just give a little

104
00:03:48,640 --> 00:03:51,760
bit of context here

105
00:03:50,080 --> 00:03:53,439
of course thanksgiving day is a national

106
00:03:51,760 --> 00:03:55,599
holiday in the united states so

107
00:03:53,439 --> 00:03:57,200
everybody uh practically everybody was

108
00:03:55,599 --> 00:03:58,959
not working that day even restaurants

109
00:03:57,200 --> 00:04:01,120
closed down

110
00:03:58,959 --> 00:04:03,280
it's almost impossible to buy a stick of

111
00:04:01,120 --> 00:04:07,200
gum in the united states on thanksgiving

112
00:04:03,280 --> 00:04:07,200
day uh because everything is closed

113
00:04:07,280 --> 00:04:10,799
but what you may also know is that the

114
00:04:08,879 --> 00:04:13,120
day after thanksgiving day is a big deal

115
00:04:10,799 --> 00:04:14,959
because that's known as black friday and

116
00:04:13,120 --> 00:04:17,199
that's when everybody rushes out to get

117
00:04:14,959 --> 00:04:19,199
the best deals they can on electronics

118
00:04:17,199 --> 00:04:21,680
and on vacuum cleaners and on whatever

119
00:04:19,199 --> 00:04:24,880
else stores might be trying to sell

120
00:04:21,680 --> 00:04:28,320
for the holiday shopping season

121
00:04:24,880 --> 00:04:30,960
this particular black friday uh while

122
00:04:28,320 --> 00:04:33,040
the entire country was outside shopping

123
00:04:30,960 --> 00:04:35,520
malls fighting over tv sets

124
00:04:33,040 --> 00:04:37,040
uh we had a half day of work uh at the

125
00:04:35,520 --> 00:04:38,639
office and our

126
00:04:37,040 --> 00:04:40,720
our newest developer not me the other

127
00:04:38,639 --> 00:04:42,800
guy who joined at the same time

128
00:04:40,720 --> 00:04:44,160
was pushing his first change out to

129
00:04:42,800 --> 00:04:45,919
customers

130
00:04:44,160 --> 00:04:48,320
what could possibly go wrong here right

131
00:04:45,919 --> 00:04:48,320
nothing

132
00:04:48,400 --> 00:04:52,800
this is what our phone lines maybe

133
00:04:49,600 --> 00:04:52,800
normally look like uh

134
00:04:53,199 --> 00:04:55,520
soon they look more like a christmas

135
00:04:54,560 --> 00:04:58,560
tree

136
00:04:55,520 --> 00:05:01,520
every single customer we had 150 plus

137
00:04:58,560 --> 00:05:02,479
appliances installed in the wild every

138
00:05:01,520 --> 00:05:05,280
one of them

139
00:05:02,479 --> 00:05:05,280
stopped working

140
00:05:05,440 --> 00:05:10,720
our phone lines were jammed and not just

141
00:05:07,680 --> 00:05:12,720
that day this lasted into the next week

142
00:05:10,720 --> 00:05:13,919
everyone in the company not just the

143
00:05:12,720 --> 00:05:15,759
people working on the spam filtering

144
00:05:13,919 --> 00:05:17,840
software the sales representatives the

145
00:05:15,759 --> 00:05:19,360
president of the company the janitor

146
00:05:17,840 --> 00:05:22,960
were taking phone calls from from

147
00:05:19,360 --> 00:05:24,560
customers whose mail was not working

148
00:05:22,960 --> 00:05:26,320
what happened

149
00:05:24,560 --> 00:05:29,120
well

150
00:05:26,320 --> 00:05:32,160
our developer had made a small a single

151
00:05:29,120 --> 00:05:32,160
bit mistake

152
00:05:32,800 --> 00:05:38,520
he had failed to set the execute bit on

153
00:05:35,039 --> 00:05:38,520
a startup script

154
00:05:38,639 --> 00:05:42,080
so this prevented of course the mail

155
00:05:40,560 --> 00:05:43,360
from flowing but it also prevented

156
00:05:42,080 --> 00:05:44,560
future updates from working so we

157
00:05:43,360 --> 00:05:46,800
couldn't even

158
00:05:44,560 --> 00:05:48,080
create a patch and put it on the on the

159
00:05:46,800 --> 00:05:49,360
central server for the servers to

160
00:05:48,080 --> 00:05:51,600
download because the all these

161
00:05:49,360 --> 00:05:55,360
appliances wouldn't wouldn't even start

162
00:05:51,600 --> 00:05:57,840
enough to download a new update

163
00:05:55,360 --> 00:05:59,360
so we had to actually log in by ssh to

164
00:05:57,840 --> 00:06:02,400
every customer machine

165
00:05:59,360 --> 00:06:04,080
and do it mod plus x on this particular

166
00:06:02,400 --> 00:06:05,840
script

167
00:06:04,080 --> 00:06:07,440
for every customer now you might think

168
00:06:05,840 --> 00:06:10,000
couldn't you script some of this well we

169
00:06:07,440 --> 00:06:11,520
could and we did to an extent

170
00:06:10,000 --> 00:06:13,919
but many of these machines were behind

171
00:06:11,520 --> 00:06:16,000
firewalls uh or they the the owners had

172
00:06:13,919 --> 00:06:18,400
changed the password or deleted ssh keys

173
00:06:16,000 --> 00:06:19,759
for security reasons so in some cases we

174
00:06:18,400 --> 00:06:21,919
actually had to call customers over the

175
00:06:19,759 --> 00:06:23,440
phone and walk them through the process

176
00:06:21,919 --> 00:06:25,039
you know if they weren't able to give us

177
00:06:23,440 --> 00:06:26,560
access to their through their network

178
00:06:25,039 --> 00:06:28,080
then we would

179
00:06:26,560 --> 00:06:29,360
walk somebody through going into their

180
00:06:28,080 --> 00:06:30,800
server room

181
00:06:29,360 --> 00:06:33,360
plugging a monitor into the physical

182
00:06:30,800 --> 00:06:35,199
machine logging in with root on the

183
00:06:33,360 --> 00:06:36,880
console and typing these commands to get

184
00:06:35,199 --> 00:06:38,319
things working again it was quite an

185
00:06:36,880 --> 00:06:40,960
ordeal it took the better part of the

186
00:06:38,319 --> 00:06:43,039
following week to get everything fixed

187
00:06:40,960 --> 00:06:46,000
over the following months uh in the

188
00:06:43,039 --> 00:06:48,800
aftermath of this uh this incident

189
00:06:46,000 --> 00:06:50,319
uh we had some other problems as you

190
00:06:48,800 --> 00:06:52,960
might imagine you know problems come up

191
00:06:50,319 --> 00:06:56,160
all the time on software uh one that

192
00:06:52,960 --> 00:06:58,240
jumps to mind is uh we had some reports

193
00:06:56,160 --> 00:07:00,800
that were taking 20 minutes or so to

194
00:06:58,240 --> 00:07:02,639
report to to generate from our database

195
00:07:00,800 --> 00:07:04,639
we were using my sql

196
00:07:02,639 --> 00:07:07,280
and uh

197
00:07:04,639 --> 00:07:09,599
the developer uh the the lead developer

198
00:07:07,280 --> 00:07:11,599
and remember my my job was mainly uh

199
00:07:09,599 --> 00:07:13,599
getting the sas ready his job was the

200
00:07:11,599 --> 00:07:14,960
core development so the core developer

201
00:07:13,599 --> 00:07:16,319
uh and i had a

202
00:07:14,960 --> 00:07:18,080
disagreement on how to do this i

203
00:07:16,319 --> 00:07:19,599
suggested partitioning by date so that

204
00:07:18,080 --> 00:07:21,759
we could only query the dates we cared

205
00:07:19,599 --> 00:07:23,840
about and he said that would be too

206
00:07:21,759 --> 00:07:26,160
complicated and take too long

207
00:07:23,840 --> 00:07:26,960
well our mutual boss came to me

208
00:07:26,160 --> 00:07:28,560
uh

209
00:07:26,960 --> 00:07:30,720
in private and said jonathan you think

210
00:07:28,560 --> 00:07:33,759
you get that partitioning working

211
00:07:30,720 --> 00:07:36,240
and uh i threw together a prototype in a

212
00:07:33,759 --> 00:07:37,680
few hours and it proved that it would be

213
00:07:36,240 --> 00:07:38,560
worth investing in

214
00:07:37,680 --> 00:07:40,160
um

215
00:07:38,560 --> 00:07:41,919
so that upset the developer um that

216
00:07:40,160 --> 00:07:44,720
wasn't the only thing of course um but

217
00:07:41,919 --> 00:07:47,919
he ended up leaving the company so i

218
00:07:44,720 --> 00:07:50,080
became the lead developer of the project

219
00:07:47,919 --> 00:07:52,080
so i essentially set out on my own of

220
00:07:50,080 --> 00:07:54,160
course we did over time hire other

221
00:07:52,080 --> 00:07:56,720
people to join the team and i remained

222
00:07:54,160 --> 00:07:59,280
the team lead for for several years and

223
00:07:56,720 --> 00:08:00,240
during this time i was reading

224
00:07:59,280 --> 00:08:01,840
a lot

225
00:08:00,240 --> 00:08:04,240
of course every hero's journey needs a

226
00:08:01,840 --> 00:08:06,240
guide and i had no shortage of guides

227
00:08:04,240 --> 00:08:08,560
um my guides were people like uh like

228
00:08:06,240 --> 00:08:11,360
joel spolsky who you may know as the

229
00:08:08,560 --> 00:08:14,000
co-founder of stack overflow uh y'all he

230
00:08:11,360 --> 00:08:16,160
also had a blog uh i guess he still does

231
00:08:14,000 --> 00:08:18,240
it's called joel on software he used to

232
00:08:16,160 --> 00:08:20,319
blog very frequently now it's

233
00:08:18,240 --> 00:08:22,639
about once a year maybe

234
00:08:20,319 --> 00:08:25,120
but i read his blog

235
00:08:22,639 --> 00:08:27,840
front to back uh maybe more than once

236
00:08:25,120 --> 00:08:29,520
during this time

237
00:08:27,840 --> 00:08:31,599
michael feathers

238
00:08:29,520 --> 00:08:33,919
he's the author of the famous book

239
00:08:31,599 --> 00:08:35,760
working effectively with legacy code i

240
00:08:33,919 --> 00:08:38,240
would say more than any other book this

241
00:08:35,760 --> 00:08:39,360
book has changed my career it is an

242
00:08:38,240 --> 00:08:42,959
excellent book especially if you're

243
00:08:39,360 --> 00:08:45,200
working with legacy code and who isn't

244
00:08:42,959 --> 00:08:47,360
kent back he's the author of extreme

245
00:08:45,200 --> 00:08:50,000
programming he's also the so-called

246
00:08:47,360 --> 00:08:52,560
inventor of test driven development and

247
00:08:50,000 --> 00:08:56,080
uh he's a very influential person one of

248
00:08:52,560 --> 00:08:58,399
the co-authors of agile manifesto

249
00:08:56,080 --> 00:09:00,959
along with bob martin another co-author

250
00:08:58,399 --> 00:09:02,640
of the agile manifesto bob martin wrote

251
00:09:00,959 --> 00:09:05,040
a number of books in the clean coding

252
00:09:02,640 --> 00:09:06,959
series clean code clean architecture

253
00:09:05,040 --> 00:09:09,360
he has some great youtube talks all of

254
00:09:06,959 --> 00:09:11,279
these people and many many others were

255
00:09:09,360 --> 00:09:12,800
my guides during the the years that i

256
00:09:11,279 --> 00:09:13,920
spent on this project

257
00:09:12,800 --> 00:09:15,360
uh to

258
00:09:13,920 --> 00:09:17,760
to help me

259
00:09:15,360 --> 00:09:20,640
take this project to the next level

260
00:09:17,760 --> 00:09:22,480
one of the lenses i have used to sort of

261
00:09:20,640 --> 00:09:23,519
evaluate the project i was working on at

262
00:09:22,480 --> 00:09:26,080
the time

263
00:09:23,519 --> 00:09:28,080
became the joel test which is a a blog

264
00:09:26,080 --> 00:09:29,519
post that joel spolsky wrote on his joel

265
00:09:28,080 --> 00:09:31,440
on software blog

266
00:09:29,519 --> 00:09:34,480
i think the post was written in 2000 so

267
00:09:31,440 --> 00:09:36,800
i read it six seven years later

268
00:09:34,480 --> 00:09:39,120
but he goes through 12 uh

269
00:09:36,800 --> 00:09:40,640
steps and and 12 questions you can ask

270
00:09:39,120 --> 00:09:43,680
about your project

271
00:09:40,640 --> 00:09:45,680
to decide the maturity or the the

272
00:09:43,680 --> 00:09:48,000
usefulness of your project

273
00:09:45,680 --> 00:09:50,640
um many of the of the items on this list

274
00:09:48,000 --> 00:09:52,399
are timeless they're still valid um do

275
00:09:50,640 --> 00:09:54,800
you fix bugs before writing new code for

276
00:09:52,399 --> 00:09:56,320
example or or does your team

277
00:09:54,800 --> 00:09:58,000
get distracted with new things before

278
00:09:56,320 --> 00:09:59,680
they finish the old things

279
00:09:58,000 --> 00:10:01,600
and leave bugs that sit for months or

280
00:09:59,680 --> 00:10:03,600
years or forever

281
00:10:01,600 --> 00:10:05,360
can you make a build on a single step

282
00:10:03,600 --> 00:10:07,040
this was kind of profound back then at

283
00:10:05,360 --> 00:10:08,399
least

284
00:10:07,040 --> 00:10:10,480
on the other hand some of the items on

285
00:10:08,399 --> 00:10:11,360
his list are a little bit outdated

286
00:10:10,480 --> 00:10:13,120
um

287
00:10:11,360 --> 00:10:15,279
do you use source control i mean that's

288
00:10:13,120 --> 00:10:16,640
still valid uh but i think practically

289
00:10:15,279 --> 00:10:19,200
everybody does if you don't use source

290
00:10:16,640 --> 00:10:20,959
control i would love to hear why

291
00:10:19,200 --> 00:10:23,040
but it's at least assumed that you

292
00:10:20,959 --> 00:10:24,880
should be using source control

293
00:10:23,040 --> 00:10:26,640
do you make daily builds maybe not

294
00:10:24,880 --> 00:10:28,079
everybody does this but it's another one

295
00:10:26,640 --> 00:10:29,920
of those things that

296
00:10:28,079 --> 00:10:31,920
the bar has been raised

297
00:10:29,920 --> 00:10:32,959
nowadays we tend to talk about are you

298
00:10:31,920 --> 00:10:36,640
doing

299
00:10:32,959 --> 00:10:39,279
daily deployments not daily build

300
00:10:36,640 --> 00:10:41,360
but i felt like it's time for something

301
00:10:39,279 --> 00:10:43,440
new so this is why i have created the

302
00:10:41,360 --> 00:10:45,279
jonathan test which are my 12 steps for

303
00:10:43,440 --> 00:10:47,279
better software delivery it's a little

304
00:10:45,279 --> 00:10:49,279
bit different focus than joel's

305
00:10:47,279 --> 00:10:51,360
um but it also is a little bit more

306
00:10:49,279 --> 00:10:52,720
updated you know 20 years newer

307
00:10:51,360 --> 00:10:54,160
as i already mentioned i feel like

308
00:10:52,720 --> 00:10:56,000
everybody should be using source control

309
00:10:54,160 --> 00:10:58,079
that's kind of assumed uh if you're not

310
00:10:56,000 --> 00:11:00,480
get with the game

311
00:10:58,079 --> 00:11:03,600
now maybe you're not doing one uh build

312
00:11:00,480 --> 00:11:05,440
per day but uh it's it's not exciting if

313
00:11:03,600 --> 00:11:06,320
you are so let's raise the bar on that

314
00:11:05,440 --> 00:11:07,760
one

315
00:11:06,320 --> 00:11:09,839
there's a lot of new technology it's

316
00:11:07,760 --> 00:11:12,560
easier to build things faster more

317
00:11:09,839 --> 00:11:14,640
frequently than it was in 2000

318
00:11:12,560 --> 00:11:16,160
um also we have some new science not a

319
00:11:14,640 --> 00:11:17,200
lot there's not a whole lot of science

320
00:11:16,160 --> 00:11:19,440
that's gone into the software

321
00:11:17,200 --> 00:11:22,000
development process but we do have some

322
00:11:19,440 --> 00:11:23,680
so we we do have some

323
00:11:22,000 --> 00:11:25,839
qualitative and quantitative data that

324
00:11:23,680 --> 00:11:28,320
we didn't in 2000 about what works and

325
00:11:25,839 --> 00:11:30,320
what doesn't for example we have we have

326
00:11:28,320 --> 00:11:31,600
evidence now that that code review is

327
00:11:30,320 --> 00:11:34,560
one of the

328
00:11:31,600 --> 00:11:36,000
um sure things to reduce defect count uh

329
00:11:34,560 --> 00:11:37,839
reviewing to this code that didn't we

330
00:11:36,000 --> 00:11:39,519
didn't know that in 2000 maybe people

331
00:11:37,839 --> 00:11:42,880
guessed but now we have some hard

332
00:11:39,519 --> 00:11:45,920
science to back some of that up

333
00:11:42,880 --> 00:11:48,160
so here's my test number one

334
00:11:45,920 --> 00:11:50,720
do devs merge multiple times per day

335
00:11:48,160 --> 00:11:52,079
also known as continuous integration

336
00:11:50,720 --> 00:11:53,920
this is another thing that science has

337
00:11:52,079 --> 00:11:56,160
proven is uh

338
00:11:53,920 --> 00:11:59,760
is beneficial uh in the sense that

339
00:11:56,160 --> 00:12:01,839
companies that do continuous integration

340
00:11:59,760 --> 00:12:03,279
outperform those in terms of delivery

341
00:12:01,839 --> 00:12:06,320
and revenue

342
00:12:03,279 --> 00:12:06,320
versus those that do not

343
00:12:07,440 --> 00:12:11,200
is all code tested thoroughly before

344
00:12:09,600 --> 00:12:13,600
merging into mainline

345
00:12:11,200 --> 00:12:17,040
when i when i took on the the developer

346
00:12:13,600 --> 00:12:19,600
project this was not assumed uh our main

347
00:12:17,040 --> 00:12:22,000
branch was broken frequently and a big

348
00:12:19,600 --> 00:12:24,399
part of the software release process

349
00:12:22,000 --> 00:12:26,639
was testing the main branch to make sure

350
00:12:24,399 --> 00:12:28,399
that it wasn't broken and fixing any

351
00:12:26,639 --> 00:12:29,760
bugs that had crept in in the

352
00:12:28,399 --> 00:12:31,200
intervening months since the last

353
00:12:29,760 --> 00:12:32,720
release

354
00:12:31,200 --> 00:12:34,880
nowadays with things like trunk based

355
00:12:32,720 --> 00:12:36,839
development we can assume that maine is

356
00:12:34,880 --> 00:12:40,079
always ready to be

357
00:12:36,839 --> 00:12:42,079
released continuous deployment

358
00:12:40,079 --> 00:12:44,160
this is your project built and deployed

359
00:12:42,079 --> 00:12:46,079
automatically after every merge to

360
00:12:44,160 --> 00:12:48,000
mainline this was

361
00:12:46,079 --> 00:12:49,440
almost inconceivable in 2000 when the

362
00:12:48,000 --> 00:12:51,440
dual test was written

363
00:12:49,440 --> 00:12:53,920
in fact it was in 2009

364
00:12:51,440 --> 00:12:55,839
when the the famous talk was done called

365
00:12:53,920 --> 00:12:58,240
10 plus deploys per day

366
00:12:55,839 --> 00:13:00,560
uh and

367
00:12:58,240 --> 00:13:01,519
that's what kind of birthed the devops

368
00:13:00,560 --> 00:13:02,959
movement

369
00:13:01,519 --> 00:13:05,200
so nowadays

370
00:13:02,959 --> 00:13:06,639
you know back then we were talking about

371
00:13:05,200 --> 00:13:08,959
is it even possible to deploy that

372
00:13:06,639 --> 00:13:10,639
frequently now we're pretty well

373
00:13:08,959 --> 00:13:13,440
confident that it is possible it's more

374
00:13:10,639 --> 00:13:16,480
a business question now should we

375
00:13:13,440 --> 00:13:16,480
i believe you should

376
00:13:16,959 --> 00:13:20,480
is all code reviewed by another

377
00:13:18,639 --> 00:13:21,839
developer before merge

378
00:13:20,480 --> 00:13:24,320
like i said this is one of the things

379
00:13:21,839 --> 00:13:26,320
that that the little science we have has

380
00:13:24,320 --> 00:13:30,000
demonstrated pretty conclusively that

381
00:13:26,320 --> 00:13:30,000
code review is effective

382
00:13:30,320 --> 00:13:33,120
do you fix bugs before writing new code

383
00:13:32,079 --> 00:13:35,040
this is when i borrowed straight from

384
00:13:33,120 --> 00:13:37,440
the drill test i believe that you should

385
00:13:35,040 --> 00:13:39,519
fix all bugs uh all significant bugs

386
00:13:37,440 --> 00:13:41,120
before writing code maybe if your

387
00:13:39,519 --> 00:13:43,519
alignment is off by two pixels you don't

388
00:13:41,120 --> 00:13:44,720
care enough to ever fix it but if it's

389
00:13:43,519 --> 00:13:46,839
something that affects users you should

390
00:13:44,720 --> 00:13:49,279
fix that before doing new

391
00:13:46,839 --> 00:13:51,760
work do you have a prioritized backlog

392
00:13:49,279 --> 00:13:53,120
of features and bugs now

393
00:13:51,760 --> 00:13:54,240
bugs should be the top of that list in

394
00:13:53,120 --> 00:13:55,600
my view

395
00:13:54,240 --> 00:13:57,440
and then features should be ranked

396
00:13:55,600 --> 00:13:59,360
according to whatever criteria your

397
00:13:57,440 --> 00:14:03,079
project or product manager

398
00:13:59,360 --> 00:14:03,079
has has in mind

399
00:14:03,199 --> 00:14:06,720
do you have a reasonably complete

400
00:14:04,720 --> 00:14:08,720
automated test suite now i say

401
00:14:06,720 --> 00:14:11,600
reasonably complete i don't say 100 or

402
00:14:08,720 --> 00:14:13,760
95 because i think those percentages are

403
00:14:11,600 --> 00:14:14,560
are almost worthless they're so easy to

404
00:14:13,760 --> 00:14:16,800
game

405
00:14:14,560 --> 00:14:18,800
that you might as well not use those

406
00:14:16,800 --> 00:14:21,040
what's more important is do you have the

407
00:14:18,800 --> 00:14:23,279
confidence when your test suite runs

408
00:14:21,040 --> 00:14:24,880
that you haven't introduced a regression

409
00:14:23,279 --> 00:14:26,160
uh if you don't have that confidence you

410
00:14:24,880 --> 00:14:28,240
should improve it you should prove your

411
00:14:26,160 --> 00:14:31,839
test suite don't focus on percentages

412
00:14:28,240 --> 00:14:31,839
focus on confidence

413
00:14:32,000 --> 00:14:35,120
does everybody have quite working

414
00:14:33,120 --> 00:14:36,880
conditions another one i borrowed from

415
00:14:35,120 --> 00:14:37,920
the joel test i think this is really

416
00:14:36,880 --> 00:14:39,760
important

417
00:14:37,920 --> 00:14:41,360
especially for knowledge work like we're

418
00:14:39,760 --> 00:14:44,399
doing in software development you should

419
00:14:41,360 --> 00:14:45,920
be able to concentrate

420
00:14:44,399 --> 00:14:47,440
you use the best tools money can buy

421
00:14:45,920 --> 00:14:49,120
another one i borrowed

422
00:14:47,440 --> 00:14:50,240
if you need a better ide you should be

423
00:14:49,120 --> 00:14:52,079
able to get one if your boss is not

424
00:14:50,240 --> 00:14:54,639
willing to pay for it you might consider

425
00:14:52,079 --> 00:14:57,600
getting a new boss

426
00:14:54,639 --> 00:15:00,079
get the best ide for your job uh get the

427
00:14:57,600 --> 00:15:01,279
best monitor the best keyboard

428
00:15:00,079 --> 00:15:02,160
nobody wants to fight with a mouse that

429
00:15:01,279 --> 00:15:03,839
doesn't work correctly when you're

430
00:15:02,160 --> 00:15:05,600
trying to code get the best tools you

431
00:15:03,839 --> 00:15:08,240
can buy whether that's

432
00:15:05,600 --> 00:15:10,320
hardware software

433
00:15:08,240 --> 00:15:11,440
the best trick ticket tracking system

434
00:15:10,320 --> 00:15:12,880
the best

435
00:15:11,440 --> 00:15:14,880
meeting software whatever you need to

436
00:15:12,880 --> 00:15:17,519
get your job done effectively

437
00:15:14,880 --> 00:15:20,079
the money spent is it pales in

438
00:15:17,519 --> 00:15:22,639
comparison to the productivity lost

439
00:15:20,079 --> 00:15:24,079
developers are expensive people uh

440
00:15:22,639 --> 00:15:26,800
get the best tools you can buy and save

441
00:15:24,079 --> 00:15:28,560
your time

442
00:15:26,800 --> 00:15:29,920
do you do hallway usability testing this

443
00:15:28,560 --> 00:15:31,759
is one that joel talked about talks

444
00:15:29,920 --> 00:15:32,720
about a lot in his blog it's one i don't

445
00:15:31,759 --> 00:15:33,920
do well

446
00:15:32,720 --> 00:15:36,079
well enough

447
00:15:33,920 --> 00:15:38,000
but there's a lot you can gain just by

448
00:15:36,079 --> 00:15:40,399
having some random person walking down

449
00:15:38,000 --> 00:15:41,920
the hallway test your code and i don't

450
00:15:40,399 --> 00:15:42,959
mean like reading your your code but

451
00:15:41,920 --> 00:15:44,320
like

452
00:15:42,959 --> 00:15:45,600
show them the prototype show them the

453
00:15:44,320 --> 00:15:47,680
software

454
00:15:45,600 --> 00:15:49,440
look at this screen what does it say to

455
00:15:47,680 --> 00:15:51,120
you just simple stuff like that just

456
00:15:49,440 --> 00:15:53,040
getting a second pair of eyes somebody

457
00:15:51,120 --> 00:15:54,800
who hasn't been looking at that all day

458
00:15:53,040 --> 00:15:58,720
to look at your code or to look at your

459
00:15:54,800 --> 00:16:00,959
your product uh can go a long ways

460
00:15:58,720 --> 00:16:03,279
no handoffs this one gets at the heart

461
00:16:00,959 --> 00:16:04,639
of the devops movement do developers

462
00:16:03,279 --> 00:16:06,480
control their workflow from start to

463
00:16:04,639 --> 00:16:08,800
finish without handoffs

464
00:16:06,480 --> 00:16:11,040
it could be to qa with operations or to

465
00:16:08,800 --> 00:16:13,519
any other team if you have handoffs

466
00:16:11,040 --> 00:16:17,839
you're not doing devops devops the whole

467
00:16:13,519 --> 00:16:17,839
idea of devops is to kill those silos

468
00:16:19,360 --> 00:16:22,959
can your developers work effectively

469
00:16:21,120 --> 00:16:26,000
from an airplane or anywhere else where

470
00:16:22,959 --> 00:16:26,000
there's no wi-fi connection

471
00:16:26,079 --> 00:16:29,440
this doesn't necessarily mean that you

472
00:16:28,079 --> 00:16:30,959
should be able to build an entire

473
00:16:29,440 --> 00:16:33,600
working application and run it from your

474
00:16:30,959 --> 00:16:35,199
laptop but you should be able to do a

475
00:16:33,600 --> 00:16:36,959
meaningful amount of work without an

476
00:16:35,199 --> 00:16:38,800
internet connection

477
00:16:36,959 --> 00:16:41,600
and this is just in my view it's just

478
00:16:38,800 --> 00:16:43,759
common sense um especially if you have

479
00:16:41,600 --> 00:16:45,839
developers who are going to conferences

480
00:16:43,759 --> 00:16:48,880
or uh they have a long commute through

481
00:16:45,839 --> 00:16:50,399
it on on a train or something like that

482
00:16:48,880 --> 00:16:52,959
or they like to work from a cafe

483
00:16:50,399 --> 00:16:54,959
somewhere it should be possible to work

484
00:16:52,959 --> 00:16:56,399
when you're offline and that's not just

485
00:16:54,959 --> 00:16:58,399
because wi-fi doesn't exist everywhere

486
00:16:56,399 --> 00:17:00,480
else sometimes github goes down

487
00:16:58,399 --> 00:17:02,720
or sometimes your production server goes

488
00:17:00,480 --> 00:17:04,959
down or sometimes anything can happen

489
00:17:02,720 --> 00:17:06,640
that prevents your entire network from

490
00:17:04,959 --> 00:17:08,640
being online and available

491
00:17:06,640 --> 00:17:12,480
you should not [ __ ] your developers

492
00:17:08,640 --> 00:17:12,480
in those instances let them keep working

493
00:17:15,760 --> 00:17:20,480
so with my 12 step test in mind let's

494
00:17:18,480 --> 00:17:21,839
let's jump back to when i started uh

495
00:17:20,480 --> 00:17:25,839
with double check

496
00:17:21,839 --> 00:17:25,839
uh what did things look like back then

497
00:17:26,240 --> 00:17:29,760
do develop do contributors

498
00:17:28,079 --> 00:17:31,039
merge their changes into mainline

499
00:17:29,760 --> 00:17:32,480
multiple times per day no we were not

500
00:17:31,039 --> 00:17:34,320
doing this we did have subversion we

501
00:17:32,480 --> 00:17:36,000
were using version control

502
00:17:34,320 --> 00:17:37,600
but there was no rule about continuous

503
00:17:36,000 --> 00:17:39,440
integration whatsoever

504
00:17:37,600 --> 00:17:41,280
code was not thoroughly tested in fact

505
00:17:39,440 --> 00:17:43,360
as i explained a minute ago we just

506
00:17:41,280 --> 00:17:45,360
assumed that mainland was broken at all

507
00:17:43,360 --> 00:17:48,799
times until we were ready to do a

508
00:17:45,360 --> 00:17:51,280
release then we would test mainline

509
00:17:48,799 --> 00:17:53,679
our project was not built or deployed at

510
00:17:51,280 --> 00:17:55,360
all uh manually or otherwise until we

511
00:17:53,679 --> 00:17:56,400
were ready to do a release so we failed

512
00:17:55,360 --> 00:17:58,000
there

513
00:17:56,400 --> 00:18:01,280
is code reviewed by other developers no

514
00:17:58,000 --> 00:18:02,960
we had no code review policy in place

515
00:18:01,280 --> 00:18:04,559
we did not fix bugs before writing new

516
00:18:02,960 --> 00:18:07,120
code we fixed bugs when they became

517
00:18:04,559 --> 00:18:08,559
quote priority

518
00:18:07,120 --> 00:18:10,480
do you have a prioritized backlog of

519
00:18:08,559 --> 00:18:12,400
features as i recall when i started here

520
00:18:10,480 --> 00:18:14,160
we did have a backlog but it was a text

521
00:18:12,400 --> 00:18:16,240
file in subversion that the developer

522
00:18:14,160 --> 00:18:18,160
had been keeping so he would just as

523
00:18:16,240 --> 00:18:20,559
request came in he would add a an item

524
00:18:18,160 --> 00:18:22,240
to the to his text file and when he was

525
00:18:20,559 --> 00:18:24,000
done he would remove it and that's fine

526
00:18:22,240 --> 00:18:25,280
if it works um

527
00:18:24,000 --> 00:18:27,280
the problem of course is that there's no

528
00:18:25,280 --> 00:18:30,160
visibility to the rest of the company uh

529
00:18:27,280 --> 00:18:31,440
you know customer support can't see that

530
00:18:30,160 --> 00:18:33,039
it didn't matter for him because he was

531
00:18:31,440 --> 00:18:34,000
a solo operation he was technical

532
00:18:33,039 --> 00:18:35,440
support

533
00:18:34,000 --> 00:18:37,039
once you have more than

534
00:18:35,440 --> 00:18:38,640
two people uh you probably want

535
00:18:37,039 --> 00:18:40,799
something a little more centralized than

536
00:18:38,640 --> 00:18:42,799
a text file of course if it works it

537
00:18:40,799 --> 00:18:43,840
works

538
00:18:42,799 --> 00:18:45,919
do you have a reasonably complete

539
00:18:43,840 --> 00:18:47,760
automated test suite uh no we i think

540
00:18:45,919 --> 00:18:50,160
there were some automated tests but most

541
00:18:47,760 --> 00:18:51,600
of them didn't work correctly um so it

542
00:18:50,160 --> 00:18:53,280
certainly wasn't being used and it was

543
00:18:51,600 --> 00:18:55,039
not a reasonable complete test suite

544
00:18:53,280 --> 00:18:56,320
whatsoever

545
00:18:55,039 --> 00:18:58,720
does everyone have quiet condition

546
00:18:56,320 --> 00:19:01,360
working conditions yes uh we did uh we

547
00:18:58,720 --> 00:19:05,360
were a small office uh we uh but it was

548
00:19:01,360 --> 00:19:06,799
relatively quiet so that one uh is a win

549
00:19:05,360 --> 00:19:08,480
to use the best tools money can buy

550
00:19:06,799 --> 00:19:10,640
honestly i don't remember the answer on

551
00:19:08,480 --> 00:19:12,559
this one i'm assuming so i don't

552
00:19:10,640 --> 00:19:14,320
remember any complaints about the tools

553
00:19:12,559 --> 00:19:17,120
we had at our disposal

554
00:19:14,320 --> 00:19:19,039
so i'm going to give them a pass

555
00:19:17,120 --> 00:19:21,600
hallway usability testing no we didn't

556
00:19:19,039 --> 00:19:23,280
really do that

557
00:19:21,600 --> 00:19:25,120
the developers control their work from

558
00:19:23,280 --> 00:19:27,200
start to finish without handoffs

559
00:19:25,120 --> 00:19:28,000
i mean in a sense i suppose so because

560
00:19:27,200 --> 00:19:30,559
uh

561
00:19:28,000 --> 00:19:32,720
there was just one developer or two

562
00:19:30,559 --> 00:19:34,320
once i started um but it certainly

563
00:19:32,720 --> 00:19:36,160
wasn't like a mindset that developers

564
00:19:34,320 --> 00:19:39,840
should be able to do all this it was it

565
00:19:36,160 --> 00:19:39,840
was by accident if anything

566
00:19:40,160 --> 00:19:43,679
can your developers work productively

567
00:19:41,520 --> 00:19:45,600
from an airplane uh you know i don't

568
00:19:43,679 --> 00:19:48,640
remember the details uh this time in

569
00:19:45,600 --> 00:19:50,480
2006 uh probably not i'm sure it was not

570
00:19:48,640 --> 00:19:52,240
intentional that if it was possible to

571
00:19:50,480 --> 00:19:55,280
do meaningful work from an airplane it

572
00:19:52,240 --> 00:19:58,000
wasn't was by accident

573
00:19:55,280 --> 00:20:00,640
final score 2.5 out of 12.

574
00:19:58,000 --> 00:20:00,640
not so great

575
00:20:01,919 --> 00:20:05,679
i stayed a double check for nine years i

576
00:20:03,440 --> 00:20:08,720
left in 2015 in september so almost

577
00:20:05,679 --> 00:20:10,559
exactly nine years uh to the day later

578
00:20:08,720 --> 00:20:13,679
uh is when i left uh let's see how we

579
00:20:10,559 --> 00:20:15,039
scored uh by the time i left

580
00:20:13,679 --> 00:20:17,679
all of these highlighted in green are

581
00:20:15,039 --> 00:20:19,600
things that had improved uh almost

582
00:20:17,679 --> 00:20:20,880
everything improved the only things it

583
00:20:19,600 --> 00:20:23,760
did not

584
00:20:20,880 --> 00:20:25,520
were of course the ability to uh deploy

585
00:20:23,760 --> 00:20:27,679
automatically after every merge we were

586
00:20:25,520 --> 00:20:28,799
not doing continuous deployment

587
00:20:27,679 --> 00:20:30,240
um

588
00:20:28,799 --> 00:20:32,159
and we did not we still did not do

589
00:20:30,240 --> 00:20:33,919
usability testing that's one i fail on

590
00:20:32,159 --> 00:20:34,960
frequently i mean i've done it at times

591
00:20:33,919 --> 00:20:37,039
but it's not something that's ever

592
00:20:34,960 --> 00:20:38,320
become a habit for me

593
00:20:37,039 --> 00:20:40,559
and could we work productively from an

594
00:20:38,320 --> 00:20:42,000
airplane uh to some extent we we had

595
00:20:40,559 --> 00:20:43,600
gotten better about that but there were

596
00:20:42,000 --> 00:20:46,480
still parts of our software that just

597
00:20:43,600 --> 00:20:48,720
didn't work without a a network in fact

598
00:20:46,480 --> 00:20:51,840
i remember writing a pearl module to

599
00:20:48,720 --> 00:20:53,919
allow us to do dns tests do dns based

600
00:20:51,840 --> 00:20:56,080
tests without a network so we could

601
00:20:53,919 --> 00:20:58,000
cache dns services so there was effort

602
00:20:56,080 --> 00:21:00,320
being done there but we hadn't reached

603
00:20:58,000 --> 00:21:02,720
uh our goal

604
00:21:00,320 --> 00:21:04,640
final score 9.5

605
00:21:02,720 --> 00:21:06,640
so we gained seven points over a nine

606
00:21:04,640 --> 00:21:08,960
year period um but i guess that's

607
00:21:06,640 --> 00:21:10,559
something right uh keep in mind i was i

608
00:21:08,960 --> 00:21:12,159
was uh

609
00:21:10,559 --> 00:21:13,440
learning as i went here

610
00:21:12,159 --> 00:21:15,120
um

611
00:21:13,440 --> 00:21:16,640
when i this was my first programming job

612
00:21:15,120 --> 00:21:18,400
it lasted nine years

613
00:21:16,640 --> 00:21:20,480
uh we had two or three other developers

614
00:21:18,400 --> 00:21:22,960
join come and go over that time period

615
00:21:20,480 --> 00:21:25,600
so not bad

616
00:21:22,960 --> 00:21:27,280
let's jump ahead to 2020 last year uh in

617
00:21:25,600 --> 00:21:28,880
january i started with a small startup

618
00:21:27,280 --> 00:21:31,360
called lana they're based in madrid

619
00:21:28,880 --> 00:21:32,400
spain uh they do uh their fintech

620
00:21:31,360 --> 00:21:35,039
startup

621
00:21:32,400 --> 00:21:37,280
when i started in january um

622
00:21:35,039 --> 00:21:41,120
only the cto was able to do deployments

623
00:21:37,280 --> 00:21:42,240
uh so developers were at his mercy to uh

624
00:21:41,120 --> 00:21:44,720
when they wanted something deployed they

625
00:21:42,240 --> 00:21:47,120
had to ask the cto and if he had time

626
00:21:44,720 --> 00:21:48,400
and agreed that it was uh something that

627
00:21:47,120 --> 00:21:50,240
should be deployed or and didn't have

628
00:21:48,400 --> 00:21:51,360
any external dependencies and so on he

629
00:21:50,240 --> 00:21:52,720
would deploy it within a day or two

630
00:21:51,360 --> 00:21:54,720
probably

631
00:21:52,720 --> 00:21:57,440
um

632
00:21:54,720 --> 00:21:59,200
the the only of these 12 items that that

633
00:21:57,440 --> 00:22:00,400
lana had was quite working conditions

634
00:21:59,200 --> 00:22:02,320
and that was mainly because it was

635
00:22:00,400 --> 00:22:05,120
mostly a remote company so most people

636
00:22:02,320 --> 00:22:05,120
work from their homes

637
00:22:05,679 --> 00:22:09,679
final score won when i started in

638
00:22:07,520 --> 00:22:11,200
january

639
00:22:09,679 --> 00:22:13,440
less than 12 months later december when

640
00:22:11,200 --> 00:22:15,760
my contract ended

641
00:22:13,440 --> 00:22:17,440
these are the the changes i had been

642
00:22:15,760 --> 00:22:19,440
able to implement um of course i didn't

643
00:22:17,440 --> 00:22:21,360
do it alone but i was a driving force

644
00:22:19,440 --> 00:22:24,720
behind most of this

645
00:22:21,360 --> 00:22:26,159
you see it's a it's a big change um

646
00:22:24,720 --> 00:22:28,480
almost everything the only things that

647
00:22:26,159 --> 00:22:29,360
weren't uh weren't done

648
00:22:28,480 --> 00:22:30,720
were

649
00:22:29,360 --> 00:22:31,840
completely automated test suites we're

650
00:22:30,720 --> 00:22:32,880
still doing a fair amount of manual

651
00:22:31,840 --> 00:22:34,080
testing

652
00:22:32,880 --> 00:22:35,360
um

653
00:22:34,080 --> 00:22:36,559
and you know you never completely get

654
00:22:35,360 --> 00:22:38,320
rid of manual testing but we were doing

655
00:22:36,559 --> 00:22:40,960
more than we would have wanted

656
00:22:38,320 --> 00:22:43,440
we did not uh have the best tools money

657
00:22:40,960 --> 00:22:44,320
could buy um i had some developers ask

658
00:22:43,440 --> 00:22:46,799
for

659
00:22:44,320 --> 00:22:48,640
an ide and they were they were denied i

660
00:22:46,799 --> 00:22:50,000
disagreed with that but i didn't control

661
00:22:48,640 --> 00:22:51,200
the budget so i didn't get to make that

662
00:22:50,000 --> 00:22:52,799
call

663
00:22:51,200 --> 00:22:55,360
we still didn't do much hallway

664
00:22:52,799 --> 00:22:58,000
usability testing

665
00:22:55,360 --> 00:22:59,600
and we didn't have a no handoffs culture

666
00:22:58,000 --> 00:23:02,640
yet so we were still only part way

667
00:22:59,600 --> 00:23:03,840
through the devops uh sort of mindset

668
00:23:02,640 --> 00:23:05,120
transition

669
00:23:03,840 --> 00:23:07,200
in the sense that we did have a manual

670
00:23:05,120 --> 00:23:09,280
qa and some other developers were

671
00:23:07,200 --> 00:23:11,200
handing things off to our manual qa to

672
00:23:09,280 --> 00:23:13,520
do work uh to do testing before

673
00:23:11,200 --> 00:23:14,880
deployment

674
00:23:13,520 --> 00:23:18,640
but there's still pretty good progress

675
00:23:14,880 --> 00:23:20,640
uh i would say um final score of nine

676
00:23:18,640 --> 00:23:23,280
after 12 months a plus eight over over

677
00:23:20,640 --> 00:23:27,280
12 months that's pretty um

678
00:23:23,280 --> 00:23:27,280
pretty encouraging in my view

679
00:23:27,440 --> 00:23:31,520
i would like to invite you to join me on

680
00:23:28,880 --> 00:23:33,360
this uh goal to improving your jonathan

681
00:23:31,520 --> 00:23:35,280
test score uh you don't have to use my

682
00:23:33,360 --> 00:23:36,880
test use whatever test you want but the

683
00:23:35,280 --> 00:23:39,039
the idea is

684
00:23:36,880 --> 00:23:40,799
can you improve on these important areas

685
00:23:39,039 --> 00:23:43,039
if you want to see the jonathan test you

686
00:23:40,799 --> 00:23:44,080
can go to this url jhal.i o

687
00:23:43,039 --> 00:23:46,559
test

688
00:23:44,080 --> 00:23:49,360
score for yourself

689
00:23:46,559 --> 00:23:50,720
i'd also like to invite you to my free

690
00:23:49,360 --> 00:23:53,279
10 day email

691
00:23:50,720 --> 00:23:56,080
uh course uh called the lean cd boot

692
00:23:53,279 --> 00:23:57,360
camp it's it's really designed for

693
00:23:56,080 --> 00:23:58,799
teams that aren't doing continuous

694
00:23:57,360 --> 00:24:00,640
deployment yet so if you're not doing

695
00:23:58,799 --> 00:24:02,799
continuous deployment this breaks it

696
00:24:00,640 --> 00:24:05,279
down into 10 simple steps that you can

697
00:24:02,799 --> 00:24:06,799
apply and in spoiler alert you don't

698
00:24:05,279 --> 00:24:08,240
have to do automated testing before

699
00:24:06,799 --> 00:24:09,520
you're doing continuous deployment and

700
00:24:08,240 --> 00:24:10,320
that's kind of the point of the lien

701
00:24:09,520 --> 00:24:11,919
here

702
00:24:10,320 --> 00:24:13,679
get yours get your continuous deployment

703
00:24:11,919 --> 00:24:15,919
in place first and then you can start

704
00:24:13,679 --> 00:24:17,440
add the steps for automation later if

705
00:24:15,919 --> 00:24:19,279
this is interesting to you i encourage

706
00:24:17,440 --> 00:24:20,960
you to go to leanciditybootcamp.com

707
00:24:19,279 --> 00:24:22,720
it's completely free sign up for the 10

708
00:24:20,960 --> 00:24:24,159
day course

709
00:24:22,720 --> 00:24:25,600
if you need any help

710
00:24:24,159 --> 00:24:27,279
with that with any of the things i've

711
00:24:25,600 --> 00:24:28,880
talked about i'm always happy to talk to

712
00:24:27,279 --> 00:24:30,240
you you can go to my website you can

713
00:24:28,880 --> 00:24:32,080
contact me

714
00:24:30,240 --> 00:24:34,960
by email follow me on linkedin or

715
00:24:32,080 --> 00:24:36,240
twitter at tinydevops

716
00:24:34,960 --> 00:24:38,559
and if there's time i'd love to take

717
00:24:36,240 --> 00:24:40,880
some questions uh so let me show you one

718
00:24:38,559 --> 00:24:42,880
last slide though if you follow this qr

719
00:24:40,880 --> 00:24:45,279
code or the url at the bottom there that

720
00:24:42,880 --> 00:24:47,200
will take you to this slide deck and

721
00:24:45,279 --> 00:24:50,000
this list of resources so you can find

722
00:24:47,200 --> 00:24:51,600
everything that i've talked about today

723
00:24:50,000 --> 00:24:54,240
with that what questions can i answer

724
00:24:51,600 --> 00:24:54,240
for the audience

725
00:24:56,960 --> 00:25:02,000
so uh we have a couple of questions that

726
00:24:59,520 --> 00:25:04,400
have been posted in the venulis

727
00:25:02,000 --> 00:25:06,320
now i'm not sure whether or not jonathan

728
00:25:04,400 --> 00:25:07,039
hall is actually here in the chat with

729
00:25:06,320 --> 00:25:08,559
us

730
00:25:07,039 --> 00:25:11,360
that we can have a look at

731
00:25:08,559 --> 00:25:14,480
uh hopefully we can

732
00:25:11,360 --> 00:25:16,480
oh someone's got a 404 on codes oh boo

733
00:25:14,480 --> 00:25:18,799
never mind um

734
00:25:16,480 --> 00:25:20,960
so if you've got some questions uh throw

735
00:25:18,799 --> 00:25:23,679
them into the chat oh and i will approve

736
00:25:20,960 --> 00:25:26,960
them as fast as i can when i can see

737
00:25:23,679 --> 00:25:30,159
them and um we'll just have a chat

738
00:25:26,960 --> 00:25:32,080
once we've uh so feel free to chat um in

739
00:25:30,159 --> 00:25:35,120
the chat that'll be great

740
00:25:32,080 --> 00:25:37,760
um we can also do that in the uh

741
00:25:35,120 --> 00:25:39,679
in the hallway as well after this

742
00:25:37,760 --> 00:25:43,200
thing end ends

743
00:25:39,679 --> 00:25:45,360
and once that's finished um we when we

744
00:25:43,200 --> 00:25:47,600
come back we'll have a break uh the next

745
00:25:45,360 --> 00:25:50,240
one that we have the next talk we have a

746
00:25:47,600 --> 00:25:53,120
break until 10 15 so you've got about 15

747
00:25:50,240 --> 00:25:55,039
20 minutes grab some water um

748
00:25:53,120 --> 00:25:56,880
take a break get up walk around we don't

749
00:25:55,039 --> 00:25:58,960
want to just be standing here looking at

750
00:25:56,880 --> 00:26:00,080
screens the whole time

751
00:25:58,960 --> 00:26:02,480
um

752
00:26:00,080 --> 00:26:04,880
but do come back at 10 15 we have a talk

753
00:26:02,480 --> 00:26:08,080
by yakov called making swiss cheese of

754
00:26:04,880 --> 00:26:10,320
ticket sales and yakov grab it gave a

755
00:26:08,080 --> 00:26:11,600
tremendous talk last year so i'm really

756
00:26:10,320 --> 00:26:13,360
looking forward to see what he's come up

757
00:26:11,600 --> 00:26:14,720
with for this year i'm sure it'd be

758
00:26:13,360 --> 00:26:17,200
entertaining

759
00:26:14,720 --> 00:26:20,200
we'll see you again in about 15-20

760
00:26:17,200 --> 00:26:20,200
minutes