1
00:00:00,750 --> 00:00:06,640
[Music]

2
00:00:05,540 --> 00:00:06,880
[Applause]

3
00:00:06,640 --> 00:00:12,480
[Music]

4
00:00:06,880 --> 00:00:12,480
[Applause]

5
00:00:13,900 --> 00:00:15,679
[Music]

6
00:00:15,280 --> 00:00:17,039
Stop.

7
00:00:15,679 --> 00:00:19,840
Stop.

8
00:00:17,039 --> 00:00:21,119
Okay, good. You can do it.

9
00:00:19,840 --> 00:00:22,880
We've forgot to teach you how to stop

10
00:00:21,119 --> 00:00:24,320
clapping yesterday to cause all kinds of

11
00:00:22,880 --> 00:00:26,240
problems.

12
00:00:24,320 --> 00:00:27,920
Yeah. When we tell you to stop clapping

13
00:00:26,240 --> 00:00:29,920
during talks, remember they only get

14
00:00:27,920 --> 00:00:32,160
five minutes. And when you when you're

15
00:00:29,920 --> 00:00:35,200
clapping, they don't get time to give a

16
00:00:32,160 --> 00:00:36,559
talk cuz we don't stop the stopwatch.

17
00:00:35,200 --> 00:00:38,239
The best way to build rapport with an

18
00:00:36,559 --> 00:00:40,239
audience, by the way, is by chastising

19
00:00:38,239 --> 00:00:42,320
them immediately. As soon as you start,

20
00:00:40,239 --> 00:00:46,399
right off the bat.

21
00:00:42,320 --> 00:00:49,600
Uh, stop.

22
00:00:46,399 --> 00:00:51,600
Try again. Stop.

23
00:00:49,600 --> 00:00:53,280
Better.

24
00:00:51,600 --> 00:00:54,719
Uh, the other complaint that we got was

25
00:00:53,280 --> 00:00:56,640
that there weren't nearly enough bad

26
00:00:54,719 --> 00:00:59,359
jokes in between the talks today. The

27
00:00:56,640 --> 00:01:01,440
good news is that we've had two days to

28
00:00:59,359 --> 00:01:03,760
become completely exhausted and turn off

29
00:01:01,440 --> 00:01:06,560
that filter. Also, I've written a whole

30
00:01:03,760 --> 00:01:08,240
bunch of them beforehand just in case um

31
00:01:06,560 --> 00:01:10,720
handing over to my wonderful

32
00:01:08,240 --> 00:01:13,119
co-presenter here. Uh the quality of

33
00:01:10,720 --> 00:01:13,920
jokes decreases during as we swap off

34
00:01:13,119 --> 00:01:15,760
just in case, you know.

35
00:01:13,920 --> 00:01:19,439
I have a spreadsheet that Chris has

36
00:01:15,760 --> 00:01:21,119
given me. Uh there are talk titles,

37
00:01:19,439 --> 00:01:24,840
presenter names, and then there are

38
00:01:21,119 --> 00:01:24,840
jokes in bold.

39
00:01:26,240 --> 00:01:33,119
I have standards.

40
00:01:28,479 --> 00:01:35,119
If you say so, Chris. I I do. Um,

41
00:01:33,119 --> 00:01:37,040
it's time to begin the lightning talks.

42
00:01:35,119 --> 00:01:39,439
Uh, and to begin this block, we have

43
00:01:37,040 --> 00:01:44,799
Nick, who's going to be talking about

44
00:01:39,439 --> 00:01:46,399
blocks. It's true. Good joke. Good joke.

45
00:01:44,799 --> 00:01:48,159
Has it come up? Has it come up? Is it

46
00:01:46,399 --> 00:01:49,840
going to come up? It's come up. Cool.

47
00:01:48,159 --> 00:01:51,439
Hello. This is my talk called The

48
00:01:49,840 --> 00:01:53,680
Children Yearn for the Minds. It's about

49
00:01:51,439 --> 00:01:55,439
importing Ultima 4 maps into Minecraft.

50
00:01:53,680 --> 00:01:57,920
Well, not Minecraft. mind test, although

51
00:01:55,439 --> 00:02:00,719
it's called Luante now. Who cares why?

52
00:01:57,920 --> 00:02:01,439
It's actually a talk about Oh, stop it.

53
00:02:00,719 --> 00:02:02,640
It's actually a talk about

54
00:02:01,439 --> 00:02:04,159
procrastinate. It's all right. I'll fix

55
00:02:02,640 --> 00:02:06,719
that later. Um, it's a talk about

56
00:02:04,159 --> 00:02:08,479
procrastination inspired by Adah's talk

57
00:02:06,719 --> 00:02:10,640
yesterday because I thought, you know,

58
00:02:08,479 --> 00:02:12,319
young developers like tend to

59
00:02:10,640 --> 00:02:14,720
procrastinate and find themselves lost

60
00:02:12,319 --> 00:02:16,720
in this mine of procrastination and and

61
00:02:14,720 --> 00:02:18,080
I just wanted to demonstrate that more

62
00:02:16,720 --> 00:02:21,680
senior developers might do the same

63
00:02:18,080 --> 00:02:23,680
thing but like harder. Anyway, so I was

64
00:02:21,680 --> 00:02:24,959
working on my presentation for and all

65
00:02:23,680 --> 00:02:26,400
these slides are in the wrong order, I

66
00:02:24,959 --> 00:02:29,040
think. I was working on my presentation

67
00:02:26,400 --> 00:02:30,160
for my thing yesterday, and I was

68
00:02:29,040 --> 00:02:31,680
working very hard at it. I had a

69
00:02:30,160 --> 00:02:33,680
proposal in, everyone liked it. It was

70
00:02:31,680 --> 00:02:34,879
really good. I had a to-do list. I had a

71
00:02:33,680 --> 00:02:36,080
bunch of stuff to look up, a bunch of

72
00:02:34,879 --> 00:02:37,440
stuff to write, a bunch of slides to

73
00:02:36,080 --> 00:02:39,760
make, all of that. It was very, very,

74
00:02:37,440 --> 00:02:42,239
very boring. Um, so of course, I did

75
00:02:39,760 --> 00:02:45,200
what anyone sensible would do, which was

76
00:02:42,239 --> 00:02:47,680
to to Well, so I really liked this game

77
00:02:45,200 --> 00:02:50,160
when I was growing up called Ultima 4.

78
00:02:47,680 --> 00:02:51,599
It was great. It ran on the Apple 2. It

79
00:02:50,160 --> 00:02:53,519
was an adventure. There was a whole

80
00:02:51,599 --> 00:02:56,080
world to explore. There was so much to

81
00:02:53,519 --> 00:02:58,000
do and find out about. And then at some

82
00:02:56,080 --> 00:02:59,760
point, oh, that the whole world to find

83
00:02:58,000 --> 00:03:01,200
out about. And you had this picture of

84
00:02:59,760 --> 00:03:02,400
this world, this picture of this place,

85
00:03:01,200 --> 00:03:03,599
even though you'd never seen it because

86
00:03:02,400 --> 00:03:05,040
it was always in this tiny little

87
00:03:03,599 --> 00:03:06,879
window. You had this picture of the

88
00:03:05,040 --> 00:03:08,239
place. You know how places are in games?

89
00:03:06,879 --> 00:03:09,760
You imagine them even though they're not

90
00:03:08,239 --> 00:03:11,680
real. They're very meaningful to you.

91
00:03:09,760 --> 00:03:13,280
And I thought, wouldn't it be cool if I

92
00:03:11,680 --> 00:03:15,599
could still have that? And when

93
00:03:13,280 --> 00:03:17,200
Minecraft came out, I thought I could

94
00:03:15,599 --> 00:03:19,200
put it in Minecraft. Imagine if I could

95
00:03:17,200 --> 00:03:20,720
take this world of blocks and put it in

96
00:03:19,200 --> 00:03:24,959
that world of blocks. Wouldn't that be

97
00:03:20,720 --> 00:03:26,319
awesome? No picture. Um, and so I had to

98
00:03:24,959 --> 00:03:27,680
learn how to read the map. So I learned

99
00:03:26,319 --> 00:03:30,000
how to read the map. And the thing is

100
00:03:27,680 --> 00:03:32,159
about Ultima 4 is so old that nothing

101
00:03:30,000 --> 00:03:34,319
could be complicated. There wasn't time.

102
00:03:32,159 --> 00:03:35,760
So we hadn't made anything hard yet. So

103
00:03:34,319 --> 00:03:37,040
like all the map is just a bunch of

104
00:03:35,760 --> 00:03:39,280
bites in a row. They're not the right

105
00:03:37,040 --> 00:03:40,640
row, but it's a row. And then I went to

106
00:03:39,280 --> 00:03:42,319
write it into Minecraft. And Minecraft's

107
00:03:40,640 --> 00:03:43,920
hard and it's in Java and mine test

108
00:03:42,319 --> 00:03:46,640
isn't and it's open source and it's in

109
00:03:43,920 --> 00:03:49,440
lure, which is okay. That's fine, too.

110
00:03:46,640 --> 00:03:51,519
but also it's got a a documented format

111
00:03:49,440 --> 00:03:53,360
for its files. It's an insane format.

112
00:03:51,519 --> 00:03:55,120
It's completely crazy, but it's

113
00:03:53,360 --> 00:03:56,799
documented or you might call it

114
00:03:55,120 --> 00:03:59,519
confessed.

115
00:03:56,799 --> 00:04:01,599
It is very complicated.

116
00:03:59,519 --> 00:04:03,040
Whoa, thank you. Um, so this is some of

117
00:04:01,599 --> 00:04:04,319
the code that writes it. I promised in

118
00:04:03,040 --> 00:04:07,040
my thing that I would have some actual

119
00:04:04,319 --> 00:04:08,560
Python code in a talk at PyCon. Hey,

120
00:04:07,040 --> 00:04:12,159
there it is. I didn't do syntax

121
00:04:08,560 --> 00:04:13,599
highlighting. Get good. Um uh basically

122
00:04:12,159 --> 00:04:16,000
what you do is you output a whole bunch

123
00:04:13,599 --> 00:04:17,199
of random binary data and a whole bunch

124
00:04:16,000 --> 00:04:18,239
of other data that is meant to do

125
00:04:17,199 --> 00:04:19,840
something but I just put a bunch of

126
00:04:18,239 --> 00:04:22,479
zeros there and saw if it worked and it

127
00:04:19,840 --> 00:04:24,479
some of it worked and then you take that

128
00:04:22,479 --> 00:04:26,720
you compress it and you write it

129
00:04:24,479 --> 00:04:29,600
somewhere around here.

130
00:04:26,720 --> 00:04:32,560
You write it into an SQLite 3 database

131
00:04:29,600 --> 00:04:35,120
as a zlib compressed row which is just

132
00:04:32,560 --> 00:04:36,560
like I say it's a confession. Anyway,

133
00:04:35,120 --> 00:04:39,040
doing that, you could suddenly write

134
00:04:36,560 --> 00:04:41,919
blocks, 16 x6 cubes of crap into the

135
00:04:39,040 --> 00:04:43,680
Luanti maps. This is awesome. So then I

136
00:04:41,919 --> 00:04:44,639
looked at this is the Ultima 4 map. So I

137
00:04:43,680 --> 00:04:45,919
looked at that and I went, wouldn't it

138
00:04:44,639 --> 00:04:47,680
be cool if I could just pick those tiles

139
00:04:45,919 --> 00:04:49,280
up and write them into Luante? There

140
00:04:47,680 --> 00:04:50,720
they are. And I went, they're very

141
00:04:49,280 --> 00:04:52,960
square and ugly looking, but that's

142
00:04:50,720 --> 00:04:55,199
okay. There's also towns. Towns have

143
00:04:52,960 --> 00:04:57,280
their own 32x 32 maps. They're one grid

144
00:04:55,199 --> 00:04:59,440
scale on the big map, but they're 32x 32

145
00:04:57,280 --> 00:05:00,639
on the little maps of each town. So you

146
00:04:59,440 --> 00:05:02,560
have to pick them up, too, and put them

147
00:05:00,639 --> 00:05:06,080
in separately onto the scaled up map

148
00:05:02,560 --> 00:05:07,840
from the big map. There they are. Um,

149
00:05:06,080 --> 00:05:10,160
and then somewhere this clicker is not

150
00:05:07,840 --> 00:05:11,600
helping me. And then I looked at these

151
00:05:10,160 --> 00:05:12,800
things and went square rivers. Square

152
00:05:11,600 --> 00:05:14,160
rivers are ugly. I have to do something

153
00:05:12,800 --> 00:05:15,600
about that. So I worked out a way that

154
00:05:14,160 --> 00:05:18,080
you could make like rounded rivers. And

155
00:05:15,600 --> 00:05:19,600
there it is. And then I put that into

156
00:05:18,080 --> 00:05:21,919
there, too. And now we have little

157
00:05:19,600 --> 00:05:24,000
roundy rivers. That's pretty cool. And

158
00:05:21,919 --> 00:05:25,039
then I had this long list of other

159
00:05:24,000 --> 00:05:26,240
things I need to do. So I need to have

160
00:05:25,039 --> 00:05:28,000
terrain. I need to have mountains coming

161
00:05:26,240 --> 00:05:29,199
up and oceans going down. I need to have

162
00:05:28,000 --> 00:05:30,960
trees in the right places. There's

163
00:05:29,199 --> 00:05:32,479
shrines and dungeons and the moon gates.

164
00:05:30,960 --> 00:05:34,080
And there's all this other complexity

165
00:05:32,479 --> 00:05:35,759
and stuff. I now have a really long

166
00:05:34,080 --> 00:05:37,280
tedious to-do list that I really can't

167
00:05:35,759 --> 00:05:38,960
be bothered doing. And so the good news

168
00:05:37,280 --> 00:05:40,639
is

169
00:05:38,960 --> 00:05:43,639
that's how I finished my talk about

170
00:05:40,639 --> 00:05:43,639
bioinformatics.

171
00:05:44,000 --> 00:05:47,639
Thank you very much.

172
00:05:53,199 --> 00:06:00,000
Thank you so much, Nick. Up next on this

173
00:05:55,680 --> 00:06:02,320
side, we have Ben and Ellena. But first,

174
00:06:00,000 --> 00:06:04,639
Miss Linda is here to tell us a story.

175
00:06:02,320 --> 00:06:07,919
Good morning, Miss Linda.

176
00:06:04,639 --> 00:06:10,160
Good morning, everyone.

177
00:06:07,919 --> 00:06:12,000
Slight content warning. There will be a

178
00:06:10,160 --> 00:06:14,479
mention of mental health and uh

179
00:06:12,000 --> 00:06:16,560
workplace exploitation.

180
00:06:14,479 --> 00:06:18,080
Once upon a time, there was an

181
00:06:16,560 --> 00:06:20,160
oligarchy.

182
00:06:18,080 --> 00:06:22,080
It wasn't your usual kind of oligarchy.

183
00:06:20,160 --> 00:06:24,800
The oligarchs weren't actually in

184
00:06:22,080 --> 00:06:26,800
charge. It was a sneaky kind of

185
00:06:24,800 --> 00:06:28,720
oligarchy. Although, it's kind of weird

186
00:06:26,800 --> 00:06:31,840
to call it sneaky because it was very,

187
00:06:28,720 --> 00:06:33,440
very loud.

188
00:06:31,840 --> 00:06:36,319
They weren't just loud. They just

189
00:06:33,440 --> 00:06:39,039
weren't loud about being in charge. They

190
00:06:36,319 --> 00:06:42,000
were loud in a hey look what's that over

191
00:06:39,039 --> 00:06:43,520
there kind of way. The oligarchs didn't

192
00:06:42,000 --> 00:06:46,240
think of themselves as oligarchs.

193
00:06:43,520 --> 00:06:48,639
Oligarchs very rarely do. Instead they

194
00:06:46,240 --> 00:06:52,319
thought of themselves as benefactors,

195
00:06:48,639 --> 00:06:54,639
saviors of mankind. Though in truth the

196
00:06:52,319 --> 00:06:57,840
man part was more important than the

197
00:06:54,639 --> 00:06:59,840
kind part. It turns out that in order to

198
00:06:57,840 --> 00:07:01,680
save mankind you have to do some pretty

199
00:06:59,840 --> 00:07:04,080
terrible things.

200
00:07:01,680 --> 00:07:06,880
In their own minds, they were creating a

201
00:07:04,080 --> 00:07:08,880
glorious new future. A future in which

202
00:07:06,880 --> 00:07:11,039
no one would need to work, or at least

203
00:07:08,880 --> 00:07:13,840
no one would need to get paid for their

204
00:07:11,039 --> 00:07:17,199
work, which was almost as good. In fact,

205
00:07:13,840 --> 00:07:19,360
for the oligarchs, it was better.

206
00:07:17,199 --> 00:07:21,759
The oligarchs set about building a

207
00:07:19,360 --> 00:07:23,680
device. The device would write

208
00:07:21,759 --> 00:07:26,720
masterpieces the like of which the world

209
00:07:23,680 --> 00:07:30,000
had never seen. Or maybe it would be a

210
00:07:26,720 --> 00:07:31,520
friend or even a therapist. Obviously,

211
00:07:30,000 --> 00:07:33,919
it would save the world from climate

212
00:07:31,520 --> 00:07:36,000
change, and it would absolutely rescue

213
00:07:33,919 --> 00:07:39,199
humanity from the need to be creative or

214
00:07:36,000 --> 00:07:42,080
interesting in any way. Honestly, it was

215
00:07:39,199 --> 00:07:43,759
going to be incredible as long as no one

216
00:07:42,080 --> 00:07:46,720
asked the oligarchs exactly what

217
00:07:43,759 --> 00:07:48,319
incredible meant. It was going to be all

218
00:07:46,720 --> 00:07:51,599
things to all people. That much was

219
00:07:48,319 --> 00:07:54,240
clear. Unfortunately, not de defining

220
00:07:51,599 --> 00:07:57,039
what the device was made for made it

221
00:07:54,240 --> 00:07:59,599
kind of tricky to build. The oligarchs

222
00:07:57,039 --> 00:08:01,840
scooped up a ton of funding because

223
00:07:59,599 --> 00:08:03,919
obviously everybody wanted an all things

224
00:08:01,840 --> 00:08:06,720
to all people device and they were

225
00:08:03,919 --> 00:08:08,800
terribly convincing. They built the

226
00:08:06,720 --> 00:08:11,759
device and set it free with great pomp

227
00:08:08,800 --> 00:08:13,680
and ceremony. It could do anything up to

228
00:08:11,759 --> 00:08:15,919
and including flying to the moon.

229
00:08:13,680 --> 00:08:19,360
Indeed, it practically had a PhD in

230
00:08:15,919 --> 00:08:21,280
flying to the moon. Unfortunately, it

231
00:08:19,360 --> 00:08:23,680
flew to the moon in much the same way as

232
00:08:21,280 --> 00:08:25,680
a bucket of sewage would. It didn't get

233
00:08:23,680 --> 00:08:27,599
very high and it made a rather stinky

234
00:08:25,680 --> 00:08:30,560
and uh terrible mess when it hit the

235
00:08:27,599 --> 00:08:33,279
ground.

236
00:08:30,560 --> 00:08:35,680
Undaunted for they were fearless

237
00:08:33,279 --> 00:08:38,080
worldsaving men, they built a second

238
00:08:35,680 --> 00:08:40,159
device feeding it everything they could

239
00:08:38,080 --> 00:08:41,760
find. And they were so good at finding

240
00:08:40,159 --> 00:08:44,399
things that they even found many things

241
00:08:41,760 --> 00:08:46,080
that belong to other people. But because

242
00:08:44,399 --> 00:08:48,560
they were building an all things to all

243
00:08:46,080 --> 00:08:50,959
people device, it was perfectly fine

244
00:08:48,560 --> 00:08:53,279
obviously for them to uh to take the

245
00:08:50,959 --> 00:08:56,399
things without asking because the device

246
00:08:53,279 --> 00:08:58,880
was essential to saving mankind. The

247
00:08:56,399 --> 00:09:01,360
actual device was amazing. It passed all

248
00:08:58,880 --> 00:09:03,600
the tests, knew all of the things. It

249
00:09:01,360 --> 00:09:05,680
was just a teensy bit embarrassing when

250
00:09:03,600 --> 00:09:08,080
it turned out not to be able to solve

251
00:09:05,680 --> 00:09:09,839
all of the problems. It certainly solved

252
00:09:08,080 --> 00:09:11,920
some of the problems, which was pretty

253
00:09:09,839 --> 00:09:14,480
impressive actually, as long as you were

254
00:09:11,920 --> 00:09:17,440
happy with the uh answer to two plus two

255
00:09:14,480 --> 00:09:19,519
being a kind of bug.

256
00:09:17,440 --> 00:09:21,200
But everyone was jolly impressed anyway

257
00:09:19,519 --> 00:09:24,080
because the oligarchs told them to be,

258
00:09:21,200 --> 00:09:26,320
and they were terribly convincing.

259
00:09:24,080 --> 00:09:27,920
Unfortunately, word got out that the

260
00:09:26,320 --> 00:09:29,680
actual device was drinking a huge amount

261
00:09:27,920 --> 00:09:32,160
of water, burning up quite a lot of the

262
00:09:29,680 --> 00:09:34,080
world's atmosphere, uh, and still not

263
00:09:32,160 --> 00:09:36,240
successfully doing all the tricks. But

264
00:09:34,080 --> 00:09:38,399
the oligarchs said that there was no way

265
00:09:36,240 --> 00:09:40,480
to do it to build the actual device. And

266
00:09:38,399 --> 00:09:41,760
obviously we had to do that and there

267
00:09:40,480 --> 00:09:43,440
was plenty more water in atmosphere

268
00:09:41,760 --> 00:09:46,800
where all of that came from anyway. And

269
00:09:43,440 --> 00:09:48,640
they were terribly convincing.

270
00:09:46,800 --> 00:09:50,560
So they scooped up more money, burnt

271
00:09:48,640 --> 00:09:52,640
even more atmosphere, stole even more

272
00:09:50,560 --> 00:09:54,640
things, and built the really actual

273
00:09:52,640 --> 00:09:56,640
device this time, or Lucifer, as they

274
00:09:54,640 --> 00:09:58,720
affectionately named it. And it seemed

275
00:09:56,640 --> 00:10:00,480
truly incredible until the word got out

276
00:09:58,720 --> 00:10:02,399
that Lucifer had been fed poor and

277
00:10:00,480 --> 00:10:04,000
vulnerable people so that they could

278
00:10:02,399 --> 00:10:06,640
wade around in the rather horrifically

279
00:10:04,000 --> 00:10:09,360
grubby insides of the machine and remove

280
00:10:06,640 --> 00:10:10,560
the problematic bits. Of course, they

281
00:10:09,360 --> 00:10:12,160
didn't necessarily make it out of

282
00:10:10,560 --> 00:10:13,600
Lucifer alive or undamaged, but the

283
00:10:12,160 --> 00:10:15,839
oligarchs pointed out that there was

284
00:10:13,600 --> 00:10:17,519
simply no other way to build Lucifer or

285
00:10:15,839 --> 00:10:19,519
his descendants. And obviously, we had

286
00:10:17,519 --> 00:10:22,000
to do that and they were terribly

287
00:10:19,519 --> 00:10:23,760
convincing. Of course, this argument

288
00:10:22,000 --> 00:10:25,600
came as rather a surprise to the people

289
00:10:23,760 --> 00:10:27,279
who were researching ways to build other

290
00:10:25,600 --> 00:10:28,560
devices that didn't eat people, burn up

291
00:10:27,279 --> 00:10:29,920
the atmosphere, drink all the water,

292
00:10:28,560 --> 00:10:31,839
steal all the things, or scoop up all

293
00:10:29,920 --> 00:10:34,000
the money, but simply did one thing

294
00:10:31,839 --> 00:10:36,160
really, really well. But it turned out

295
00:10:34,000 --> 00:10:38,399
nobody really cared about all of that

296
00:10:36,160 --> 00:10:41,560
stuff. And so the devices continued to

297
00:10:38,399 --> 00:10:41,560
be built,

298
00:10:43,440 --> 00:10:49,200
skipping all of the things.

299
00:10:46,240 --> 00:10:54,760
The story doesn't stop there, but

300
00:10:49,200 --> 00:10:54,760
perhaps we could choose our own ending.

301
00:11:00,000 --> 00:11:04,959
Thank you, Miss Linda. I have no idea

302
00:11:02,880 --> 00:11:07,680
what that story could have been about,

303
00:11:04,959 --> 00:11:10,640
but I'm sure it's very important.

304
00:11:07,680 --> 00:11:13,279
Uh, Yakov, you're on deck. Uh, but

305
00:11:10,640 --> 00:11:13,600
first, uh, Benelina.

306
00:11:13,279 --> 00:11:16,600
Hey,

307
00:11:13,600 --> 00:11:16,600
Benelina.

308
00:11:16,959 --> 00:11:19,040
microphone.

309
00:11:18,000 --> 00:11:21,279
Thank you, sir.

310
00:11:19,040 --> 00:11:23,040
Hello, PyCon.

311
00:11:21,279 --> 00:11:23,360
How are we?

312
00:11:23,040 --> 00:11:25,760
Good.

313
00:11:23,360 --> 00:11:28,480
That's great. I'm a uh longtime

314
00:11:25,760 --> 00:11:30,640
listener, first- time talker.

315
00:11:28,480 --> 00:11:33,600
Um, and I'm here to ask you some

316
00:11:30,640 --> 00:11:36,720
questions. Do you like Python?

317
00:11:33,600 --> 00:11:39,760
Do you like the Python community?

318
00:11:36,720 --> 00:11:43,600
Do you like high availability?

319
00:11:39,760 --> 00:11:46,240
Do you like things that are easy to do?

320
00:11:43,600 --> 00:11:47,920
Yeah. Yeah. Thought that. So come be a

321
00:11:46,240 --> 00:11:49,839
part of the Python Australia community.

322
00:11:47,920 --> 00:11:54,079
It's your Python community in Australia

323
00:11:49,839 --> 00:11:56,079
the other 360.25 days of the year. Um we

324
00:11:54,079 --> 00:11:58,240
have a website and a discord. You can

325
00:11:56,079 --> 00:12:01,600
use the website to get to the discord.

326
00:11:58,240 --> 00:12:02,079
Um and we are

327
00:12:01,600 --> 00:12:03,680
so

328
00:12:02,079 --> 00:12:06,000
yeah so I mean you might be asking

329
00:12:03,680 --> 00:12:07,920
yourself what is a Python Australia?

330
00:12:06,000 --> 00:12:10,240
Well you might note there is a Python

331
00:12:07,920 --> 00:12:12,639
Africa, a Python Asia, a Python Europe,

332
00:12:10,240 --> 00:12:14,639
a Python Latin America. There is no

333
00:12:12,639 --> 00:12:16,959
Python Australia. We believe that there

334
00:12:14,639 --> 00:12:19,279
should be at the end of last PyCon AU in

335
00:12:16,959 --> 00:12:21,040
the hallway track uh small handful of us

336
00:12:19,279 --> 00:12:23,279
went all got together and went why don't

337
00:12:21,040 --> 00:12:26,160
we make a Python Australia? So your next

338
00:12:23,279 --> 00:12:30,720
question is probably who is a Python

339
00:12:26,160 --> 00:12:32,560
Australia? Um well we're in the user

340
00:12:30,720 --> 00:12:34,079
groups. So if you're not part of your

341
00:12:32,560 --> 00:12:35,760
local user group or you do not have one,

342
00:12:34,079 --> 00:12:37,360
we recommend you join one. And there's

343
00:12:35,760 --> 00:12:40,000
another thing too that we tend to over

344
00:12:37,360 --> 00:12:41,760
represent on the coding front. Um there

345
00:12:40,000 --> 00:12:44,959
are more core contributors in Australia

346
00:12:41,760 --> 00:12:47,120
than is statistically consistent. Uh we

347
00:12:44,959 --> 00:12:48,560
are very very proud and uh big shout out

348
00:12:47,120 --> 00:12:50,240
to all these people. They are also part

349
00:12:48,560 --> 00:12:51,600
of our community. There are many online

350
00:12:50,240 --> 00:12:53,519
spaces which many Australians

351
00:12:51,600 --> 00:12:54,800
participate in and we're very proud. You

352
00:12:53,519 --> 00:12:56,639
are part of the Python Australia

353
00:12:54,800 --> 00:12:57,920
community. There are all kinds of other

354
00:12:56,639 --> 00:13:00,560
weird and wonderful things we do with

355
00:12:57,920 --> 00:13:03,040
Python. All of you are also welcome as

356
00:13:00,560 --> 00:13:06,240
are all of just you know our day-to-day

357
00:13:03,040 --> 00:13:08,959
professional Python users. So who is

358
00:13:06,240 --> 00:13:11,680
part of Python Australia?

359
00:13:08,959 --> 00:13:14,880
everyone.

360
00:13:11,680 --> 00:13:16,959
Yes, including you. Um, and you know

361
00:13:14,880 --> 00:13:19,040
that thing Chris Chris No was talking

362
00:13:16,959 --> 00:13:23,120
about in consequences this morning. This

363
00:13:19,040 --> 00:13:26,000
is that. Come join us. Uh, we having a

364
00:13:23,120 --> 00:13:27,760
meeting tomorrow in our Discord. You'll

365
00:13:26,000 --> 00:13:29,279
find the details there.

366
00:13:27,760 --> 00:13:31,839
I've got one.

367
00:13:29,279 --> 00:13:35,120
Um, so come join us. Uh, all experience

368
00:13:31,839 --> 00:13:37,680
levels are welcome. Um, we we hope to

369
00:13:35,120 --> 00:13:39,600
represent you well. Um we we have

370
00:13:37,680 --> 00:13:41,360
somewhat come together from all the

371
00:13:39,600 --> 00:13:44,560
different states and other small

372
00:13:41,360 --> 00:13:46,000
community bodies. Um and uh yeah, we

373
00:13:44,560 --> 00:13:48,160
we're hoping to represent you and give

374
00:13:46,000 --> 00:13:51,160
you a space to be connected to each

375
00:13:48,160 --> 00:13:51,160
other.

376
00:13:51,700 --> 00:13:55,259
[Music]

377
00:13:57,440 --> 00:14:01,519
Thank you so much, Ben and Elellanena,

378
00:13:59,199 --> 00:14:04,720
for giving me a way to hang out with all

379
00:14:01,519 --> 00:14:07,360
of you folks after Tuesday. Uh, up next

380
00:14:04,720 --> 00:14:11,839
on this side is Christopher Phillips.

381
00:14:07,360 --> 00:14:12,959
But first, over to Yakov. Over.

382
00:14:11,839 --> 00:14:14,880
I should have expected that. Hello.

383
00:14:12,959 --> 00:14:16,079
Hello. I am a human and you are humans

384
00:14:14,880 --> 00:14:18,240
and I would like to communicate with

385
00:14:16,079 --> 00:14:19,760
you, but this is difficult. As Lily

386
00:14:18,240 --> 00:14:22,079
mentioned in her wonderful talk earlier

387
00:14:19,760 --> 00:14:23,680
today, what is in my head and what is in

388
00:14:22,079 --> 00:14:25,279
your head are two different things and

389
00:14:23,680 --> 00:14:27,680
we need a way to relate them. And so to

390
00:14:25,279 --> 00:14:30,720
do this, we try our best with this

391
00:14:27,680 --> 00:14:33,040
wonderful thing called language. And it

392
00:14:30,720 --> 00:14:34,639
is wonderful. Uh, sometimes it's spoken,

393
00:14:33,040 --> 00:14:36,959
sometimes it's written, but is a great

394
00:14:34,639 --> 00:14:39,440
way to try and encode all of my thoughts

395
00:14:36,959 --> 00:14:41,680
and ideas. Well, most of them if we're

396
00:14:39,440 --> 00:14:43,360
being honest. But these are not just

397
00:14:41,680 --> 00:14:45,920
used to make up words. We use these

398
00:14:43,360 --> 00:14:48,000
letters as all sorts of identifiers

399
00:14:45,920 --> 00:14:50,639
everywhere. Letters are so good. We put

400
00:14:48,000 --> 00:14:51,920
them everywhere. We use them for words.

401
00:14:50,639 --> 00:14:53,519
We use them for license plates. We use

402
00:14:51,920 --> 00:14:55,440
them for registrations. We put them for

403
00:14:53,519 --> 00:14:57,279
We use them for taxi ways. We use them

404
00:14:55,440 --> 00:14:59,040
for key areas in Counterstrike only A

405
00:14:57,279 --> 00:15:01,040
and B cuz there's never more than that.

406
00:14:59,040 --> 00:15:02,639
And so on. We love using letters

407
00:15:01,040 --> 00:15:04,959
everywhere,

408
00:15:02,639 --> 00:15:06,240
but sometimes using letters is hard.

409
00:15:04,959 --> 00:15:08,480
There are some problems that can happen

410
00:15:06,240 --> 00:15:10,160
when you try and encode and decode the

411
00:15:08,480 --> 00:15:12,880
magic of thoughts into some other

412
00:15:10,160 --> 00:15:14,560
medium. And one of them is a bit of a

413
00:15:12,880 --> 00:15:17,199
pet peeve of mine. And it's very much

414
00:15:14,560 --> 00:15:19,519
self-inflicted. And that is if you want

415
00:15:17,199 --> 00:15:23,680
to describe a letter, it has a name. You

416
00:15:19,519 --> 00:15:25,040
just say the name. A B C D. Well, four

417
00:15:23,680 --> 00:15:27,760
in and three of them sound the same

418
00:15:25,040 --> 00:15:29,600
already. This can be a problem

419
00:15:27,760 --> 00:15:31,760
particularly if you're in an environment

420
00:15:29,600 --> 00:15:32,880
where you can't have clear audio. So

421
00:15:31,760 --> 00:15:35,120
you've got a phone call that's a bit

422
00:15:32,880 --> 00:15:36,560
dodgy. If you're using a two-way radio,

423
00:15:35,120 --> 00:15:38,160
if you're in person, if you're wearing a

424
00:15:36,560 --> 00:15:40,000
mask, if you have a thick accent, it can

425
00:15:38,160 --> 00:15:41,680
be very hard when you have a letter to

426
00:15:40,000 --> 00:15:43,279
know exactly which one was said when it

427
00:15:41,680 --> 00:15:46,000
could have been one of two, one of six,

428
00:15:43,279 --> 00:15:48,079
one of seven. So there is a standard for

429
00:15:46,000 --> 00:15:50,240
this. It emerged from civil aviation in

430
00:15:48,079 --> 00:15:52,399
the 1960s. there is a clear word that

431
00:15:50,240 --> 00:15:55,199
can be used instead of a letter because

432
00:15:52,399 --> 00:15:57,279
aviation really really loves to use

433
00:15:55,199 --> 00:15:58,800
letters for everything. Each word

434
00:15:57,279 --> 00:16:01,199
basically has two rules. Number one is

435
00:15:58,800 --> 00:16:03,279
it starts with the letter that it

436
00:16:01,199 --> 00:16:06,079
represents which normal names of letters

437
00:16:03,279 --> 00:16:07,600
do not. Uh and number two is that it

438
00:16:06,079 --> 00:16:09,360
doesn't sound like any of the others.

439
00:16:07,600 --> 00:16:12,000
They all sound very distinct. So you

440
00:16:09,360 --> 00:16:13,600
have a for alpha, b for bravo, c for

441
00:16:12,000 --> 00:16:15,680
charlie,

442
00:16:13,600 --> 00:16:19,120
u for uniform. I have heard so many

443
00:16:15,680 --> 00:16:21,440
people say you for unicorn.

444
00:16:19,120 --> 00:16:25,199
It's close enough. If it works, it

445
00:16:21,440 --> 00:16:27,040
works. Why not? Exactly. Um, I'd heard

446
00:16:25,199 --> 00:16:29,519
of this. Yeah, cool. Something for like

447
00:16:27,040 --> 00:16:31,680
cops and fireflies, whatever to use. I

448
00:16:29,519 --> 00:16:33,040
didn't really get into it until I was

449
00:16:31,680 --> 00:16:34,399
trying to memorize it when I got into

450
00:16:33,040 --> 00:16:36,079
amateur radio a few years back. I think

451
00:16:34,399 --> 00:16:38,399
I've already shared this picture at PyCon

452
00:16:36,079 --> 00:16:40,000
maybe last year or the year before. This

453
00:16:38,399 --> 00:16:42,480
was me working with the Wavely Amateur

454
00:16:40,000 --> 00:16:45,199
Radio Society on an international radio

455
00:16:42,480 --> 00:16:47,680
communications contest. Yes, they have a

456
00:16:45,199 --> 00:16:49,519
contest for that. And no, it's probably

457
00:16:47,680 --> 00:16:52,079
not as exciting as watching last night's

458
00:16:49,519 --> 00:16:54,720
tram driving contest, which I still need

459
00:16:52,079 --> 00:16:56,880
to catch up on. But this uses shortwave

460
00:16:54,720 --> 00:16:58,639
radio communication. And shortwave radio

461
00:16:56,880 --> 00:17:00,560
relies on bouncing signals off the

462
00:16:58,639 --> 00:17:02,639
Earth's atmosphere. So, as the

463
00:17:00,560 --> 00:17:04,640
atmosphere shifts around, other stations

464
00:17:02,639 --> 00:17:07,120
will fade in, fade out. Can be very hard

465
00:17:04,640 --> 00:17:08,720
to hear them sometimes. By sometimes,

466
00:17:07,120 --> 00:17:09,839
most of the time, in my experience,

467
00:17:08,720 --> 00:17:11,120
particular if they're in another state

468
00:17:09,839 --> 00:17:13,439
or another country, you can reach

469
00:17:11,120 --> 00:17:16,160
overseas. No internet. not exactly the

470
00:17:13,439 --> 00:17:18,400
clearest signal. Or if you're visiting

471
00:17:16,160 --> 00:17:20,559
from interstate or overseas and you're

472
00:17:18,400 --> 00:17:21,839
on a plane into or out of Melbourne, you

473
00:17:20,559 --> 00:17:23,919
definitely wouldn't want them to mix

474
00:17:21,839 --> 00:17:26,160
things up. We've got the taxiway map

475
00:17:23,919 --> 00:17:29,120
here. Uh you wouldn't want to mix up B

476
00:17:26,160 --> 00:17:31,440
and C up the top. We've got M and N over

477
00:17:29,120 --> 00:17:34,799
to the left. Um or for the main runway,

478
00:17:31,440 --> 00:17:36,720
you've got B, C, G, and E, which all

479
00:17:34,799 --> 00:17:38,240
sound very similar. And you really want

480
00:17:36,720 --> 00:17:41,120
to make sure you land up at the right

481
00:17:38,240 --> 00:17:42,559
spot. and ideally not on a runway you

482
00:17:41,120 --> 00:17:45,280
not on a taxi way where there's another

483
00:17:42,559 --> 00:17:47,200
aircraft already standing there. But

484
00:17:45,280 --> 00:17:49,760
it's not just industry use. I've found

485
00:17:47,200 --> 00:17:50,880
uses for this in everyday life. Um as a

486
00:17:49,760 --> 00:17:52,400
practical example, since not

487
00:17:50,880 --> 00:17:54,240
everything's digitized, I was on a phone

488
00:17:52,400 --> 00:17:56,880
call a few years back with my car

489
00:17:54,240 --> 00:17:59,120
insurance and they wanted my 17digit

490
00:17:56,880 --> 00:18:00,720
VIN, the vehicle identification number,

491
00:17:59,120 --> 00:18:02,480
which by the way, did you know that

492
00:18:00,720 --> 00:18:04,720
that's also a standard? It's not just a

493
00:18:02,480 --> 00:18:06,559
random series of numbers and letters.

494
00:18:04,720 --> 00:18:10,080
Um, but it's really not fun when you've

495
00:18:06,559 --> 00:18:11,919
got 17 17 digits, a bad phone line cuz

496
00:18:10,080 --> 00:18:14,799
they've offshored their phone support to

497
00:18:11,919 --> 00:18:16,240
some other country. Um, and so many

498
00:18:14,799 --> 00:18:18,320
letters sound the same. It reminded me

499
00:18:16,240 --> 00:18:20,880
of Windows product activation by phone,

500
00:18:18,320 --> 00:18:22,880
but it was worse. Using phonetics helped

501
00:18:20,880 --> 00:18:24,400
a lot here, though not everyone expects

502
00:18:22,880 --> 00:18:25,679
it. I was once on a phone call with the

503
00:18:24,400 --> 00:18:27,440
police. When I started using phonetics,

504
00:18:25,679 --> 00:18:29,919
I officer just kind of froze and they're

505
00:18:27,440 --> 00:18:31,919
like, "Are you a police officer?" I was

506
00:18:29,919 --> 00:18:34,000
like, "No, you you don't need to be a

507
00:18:31,919 --> 00:18:36,160
police officer to use phonetics." So I

508
00:18:34,000 --> 00:18:38,160
strongly suggest people could go off try

509
00:18:36,160 --> 00:18:40,240
learn this it becomes much easier to

510
00:18:38,160 --> 00:18:43,280
read out letters when you're spelling

511
00:18:40,240 --> 00:18:45,440
things and as well as or instead of if

512
00:18:43,280 --> 00:18:47,039
you want to be cheeky there's the native

513
00:18:45,440 --> 00:18:49,600
phonetic alphabet and then someone came

514
00:18:47,039 --> 00:18:54,000
up with the anti-phenetic alphabet we

515
00:18:49,600 --> 00:18:58,000
have gems like C for Q as in key and the

516
00:18:54,000 --> 00:19:00,480
one I had to Google T as in ulong so you

517
00:18:58,000 --> 00:19:02,880
can guarantee confusion

518
00:19:00,480 --> 00:19:04,400
wherever you go

519
00:19:02,880 --> 00:19:06,000
That's it from me. You can find more of

520
00:19:04,400 --> 00:19:08,300
me online at this URL or offline in the

521
00:19:06,000 --> 00:19:11,120
hallways outside. Thank you.

522
00:19:08,300 --> 00:19:15,840
[Music]

523
00:19:11,120 --> 00:19:19,600
Uh thank you Yankee. Um Morgan, you are

524
00:19:15,840 --> 00:19:23,280
next. Uh but first, uh Christopher is

525
00:19:19,600 --> 00:19:26,280
going to talk to us about um driving

526
00:19:23,280 --> 00:19:26,280
illegally,

527
00:19:26,400 --> 00:19:31,360
something like that. Hi. Uh, my name is

528
00:19:29,919 --> 00:19:33,440
Christopher Phillips and I am a

529
00:19:31,360 --> 00:19:35,120
recovering game developer.

530
00:19:33,440 --> 00:19:36,880
Hi Christopher.

531
00:19:35,120 --> 00:19:38,320
Thanks for sharing.

532
00:19:36,880 --> 00:19:40,320
I'm currently working at the Curtain

533
00:19:38,320 --> 00:19:41,679
Institute of Radio Astronomy maintaining

534
00:19:40,320 --> 00:19:44,000
software at the core of the Merches and

535
00:19:41,679 --> 00:19:45,600
Widefield Array radio telescope. And

536
00:19:44,000 --> 00:19:47,760
this is a short version of a longer talk

537
00:19:45,600 --> 00:19:51,039
I gave somewhere else about a year ago.

538
00:19:47,760 --> 00:19:53,200
Uh, let's rewind a few years,

539
00:19:51,039 --> 00:19:55,919
right? As a teenager, the computer to

540
00:19:53,200 --> 00:19:58,559
have was the Commodore 64.

541
00:19:55,919 --> 00:20:00,000
Yay. Many people fans cool. My friends

542
00:19:58,559 --> 00:20:01,760
and I spent a fair amount of time

543
00:20:00,000 --> 00:20:02,799
playing computer games, uh, but were

544
00:20:01,760 --> 00:20:04,559
just as interested in the idea of

545
00:20:02,799 --> 00:20:06,720
writing them. Many grand designs were

546
00:20:04,559 --> 00:20:08,559
out loud, outlined on pen and paper, a

547
00:20:06,720 --> 00:20:10,160
few code fragments written, but nothing

548
00:20:08,559 --> 00:20:11,840
concrete or publishable. High school

549
00:20:10,160 --> 00:20:16,240
ended, we went our separate ways, and I

550
00:20:11,840 --> 00:20:18,480
dove into an engineering degree at UWA.

551
00:20:16,240 --> 00:20:19,919
Um, now the degree started out giving

552
00:20:18,480 --> 00:20:21,440
you a pretty broad grounding. Even if

553
00:20:19,919 --> 00:20:22,720
you're eventually doing coding or

554
00:20:21,440 --> 00:20:24,000
electronics, you still had to start out

555
00:20:22,720 --> 00:20:25,679
with some mechanical and civil

556
00:20:24,000 --> 00:20:27,039
engineering units. This turned out to be

557
00:20:25,679 --> 00:20:29,919
pretty useful later on. and I'm a big

558
00:20:27,039 --> 00:20:31,440
fan of broad degrees. Um, that's also

559
00:20:29,919 --> 00:20:33,200
where I became a member of the

560
00:20:31,440 --> 00:20:35,520
university computer club, uh, pictured

561
00:20:33,200 --> 00:20:36,880
in the bottom right there. And I spent

562
00:20:35,520 --> 00:20:38,480
just as much time in the club room

563
00:20:36,880 --> 00:20:40,799
tinkering on old machines as I did in my

564
00:20:38,480 --> 00:20:42,320
formal studies, probably more. Um,

565
00:20:40,799 --> 00:20:44,720
particularly continuing to try and teach

566
00:20:42,320 --> 00:20:47,120
my Commodore 64 new tricks despite its

567
00:20:44,720 --> 00:20:50,799
waning commercial value.

568
00:20:47,120 --> 00:20:51,919
Um, so that said, although Commodore 64s

569
00:20:50,799 --> 00:20:53,280
aren't a thing you can sell games for

570
00:20:51,919 --> 00:20:55,200
anymore, after all, I make more than

571
00:20:53,280 --> 00:20:56,720
maybe a few hundred. A constant with

572
00:20:55,200 --> 00:20:58,159
games consoles is they're designed to be

573
00:20:56,720 --> 00:20:59,600
cheap to make with only just enough

574
00:20:58,159 --> 00:21:00,960
grunt that a good developer can make

575
00:20:59,600 --> 00:21:03,280
shinier games than the ones coming out

576
00:21:00,960 --> 00:21:04,880
from competing hardware. So given that

577
00:21:03,280 --> 00:21:06,400
when I finally took a run at applying

578
00:21:04,880 --> 00:21:07,840
for some games positions a few years

579
00:21:06,400 --> 00:21:09,760
later, studios were just as interested

580
00:21:07,840 --> 00:21:11,280
in my tales of Commodore 64 fast loaders

581
00:21:09,760 --> 00:21:13,440
and sprite multiplexes as there were in

582
00:21:11,280 --> 00:21:15,039
my degree and that helped me get a job

583
00:21:13,440 --> 00:21:16,400
at Reflection Interactive in the north

584
00:21:15,039 --> 00:21:18,720
of England helping develop a whole

585
00:21:16,400 --> 00:21:20,559
series of driving games.

586
00:21:18,720 --> 00:21:22,720
Hands up if you've ever played any of

587
00:21:20,559 --> 00:21:24,880
Driver or Driver 2 or Driver 3 or

588
00:21:22,720 --> 00:21:28,159
Stuntman. There are some really old

589
00:21:24,880 --> 00:21:30,640
people in the audience here.

590
00:21:28,159 --> 00:21:32,559
Um, one of my first really interesting

591
00:21:30,640 --> 00:21:34,400
tasks was to replace an older 2D racing

592
00:21:32,559 --> 00:21:36,159
simulation with a full 3D physical model

593
00:21:34,400 --> 00:21:37,919
of how a car accelerates, hand does

594
00:21:36,159 --> 00:21:39,120
handbrake turns and corners. And for

595
00:21:37,919 --> 00:21:41,039
that, my first few mechanical

596
00:21:39,120 --> 00:21:43,919
engineering was a godsend. A lot of

597
00:21:41,039 --> 00:21:46,320
integration of 3D maths and vector

598
00:21:43,919 --> 00:21:48,400
arithmetic and things. Now, the third

599
00:21:46,320 --> 00:21:50,320
game I worked on

600
00:21:48,400 --> 00:21:52,000
um marked the company's transition to a

601
00:21:50,320 --> 00:21:54,799
new hardware platform, the PS2. And we

602
00:21:52,000 --> 00:21:56,400
really wanted to show off how much

603
00:21:54,799 --> 00:21:58,720
detail we could get in models of cities

604
00:21:56,400 --> 00:22:00,320
like and the like. We sent artists out

605
00:21:58,720 --> 00:22:01,840
to take thousands of reference photos

606
00:22:00,320 --> 00:22:02,960
before they modeled dozens of different

607
00:22:01,840 --> 00:22:04,240
districts and all these different

608
00:22:02,960 --> 00:22:06,159
architectural styles and famous

609
00:22:04,240 --> 00:22:07,600
landmarks. Question is, how do you fit

610
00:22:06,159 --> 00:22:10,240
all that into a games console that's

611
00:22:07,600 --> 00:22:12,400
only got 32 megabytes of RAM? Short

612
00:22:10,240 --> 00:22:15,120
version is im custom image compression

613
00:22:12,400 --> 00:22:16,799
codecs, clever data streaming systems,

614
00:22:15,120 --> 00:22:19,520
um, and a few other bits and pieces

615
00:22:16,799 --> 00:22:21,520
along the way. uh cutting a bit closer

616
00:22:19,520 --> 00:22:24,159
to the present day. I recently had a

617
00:22:21,520 --> 00:22:25,919
opportunity to optimize some code for a

618
00:22:24,159 --> 00:22:27,760
some spatial ecologists monitoring plant

619
00:22:25,919 --> 00:22:30,720
health along creek beds near mine sites

620
00:22:27,760 --> 00:22:32,720
in areas of WA. They 10 centimeter

621
00:22:30,720 --> 00:22:34,000
resolution area surveys across dozens of

622
00:22:32,720 --> 00:22:36,240
square kilometers. That's about 3

623
00:22:34,000 --> 00:22:38,240
billion pixels per site. Um with each

624
00:22:36,240 --> 00:22:40,000
pixel having RGB color, a ground

625
00:22:38,240 --> 00:22:41,919
altitude, canopy altitude, vegetation

626
00:22:40,000 --> 00:22:43,679
classification, way too much to fit into

627
00:22:41,919 --> 00:22:46,000
a desktop computer. and their attempts

628
00:22:43,679 --> 00:22:47,520
to run the reports that had been just

629
00:22:46,000 --> 00:22:50,080
fine on small sites were really

630
00:22:47,520 --> 00:22:52,799
struggling on um monitoring entire

631
00:22:50,080 --> 00:22:54,320
multicometers creek beds. Um even if

632
00:22:52,799 --> 00:22:56,960
they threw more hardware, it just ran

633
00:22:54,320 --> 00:22:58,240
for a week and crashed. And to me, I

634
00:22:56,960 --> 00:22:59,679
noticed that's really not that different

635
00:22:58,240 --> 00:23:01,120
to grabbing just enough imagery of a

636
00:22:59,679 --> 00:23:02,720
game disc to render the scenery along a

637
00:23:01,120 --> 00:23:05,520
route taken by a car touring across a

638
00:23:02,720 --> 00:23:08,640
city. So

639
00:23:05,520 --> 00:23:10,240
I improved the usage of numpy, made it a

640
00:23:08,640 --> 00:23:12,320
bit more idatic, a few less loops in

641
00:23:10,240 --> 00:23:13,760
Python, a few more underlining C, but

642
00:23:12,320 --> 00:23:14,880
also used a lot of game development

643
00:23:13,760 --> 00:23:16,720
techniques for optimizing the

644
00:23:14,880 --> 00:23:18,559
calculation, streaming tiled compressed

645
00:23:16,720 --> 00:23:20,880
data to reduce memory footprint, IO

646
00:23:18,559 --> 00:23:22,159
bandwidth, and storage requirements. Uh,

647
00:23:20,880 --> 00:23:23,760
through all this managed to get one of

648
00:23:22,159 --> 00:23:25,280
the reports down from taking a week and

649
00:23:23,760 --> 00:23:26,799
probably not even finishing to taking

650
00:23:25,280 --> 00:23:28,240
about half an hour on a desktop PC in

651
00:23:26,799 --> 00:23:30,240
the office. I was pretty happy with

652
00:23:28,240 --> 00:23:32,080
that.

653
00:23:30,240 --> 00:23:33,520
Um, I know all that without moving away

654
00:23:32,080 --> 00:23:35,600
from Python. didn't have to write a sing

655
00:23:33,520 --> 00:23:39,120
rewrite a single line in Python or Rust

656
00:23:35,600 --> 00:23:41,760
in C rust. Python all the way. Um,

657
00:23:39,120 --> 00:23:43,360
moving on, my current position started

658
00:23:41,760 --> 00:23:44,720
out with lunch with another old unique

659
00:23:43,360 --> 00:23:46,480
contact who organized with his team

660
00:23:44,720 --> 00:23:47,679
leader. Um, I wasn't that keen on

661
00:23:46,480 --> 00:23:49,039
commuting, but he was like, "No, just

662
00:23:47,679 --> 00:23:50,960
come and have a chat with Greg, see what

663
00:23:49,039 --> 00:23:52,480
we've got in common." I shared my

664
00:23:50,960 --> 00:23:54,720
stories of compressing and streaming

665
00:23:52,480 --> 00:23:56,240
vast maps of data, how video games had

666
00:23:54,720 --> 00:23:57,280
been useful prep for spatial ecology,

667
00:23:56,240 --> 00:23:59,440
and how I've been messing around with

668
00:23:57,280 --> 00:24:01,440
for transforms for various things on a

669
00:23:59,440 --> 00:24:03,200
range of platforms over the years. In

670
00:24:01,440 --> 00:24:05,280
return, Greg waxed lyrical about a novel

671
00:24:03,200 --> 00:24:06,799
radio telescope featuring thousands of

672
00:24:05,280 --> 00:24:09,039
little radio antennas scattered across

673
00:24:06,799 --> 00:24:10,559
the desert that steered by changing the

674
00:24:09,039 --> 00:24:13,440
delays with which they listen signals

675
00:24:10,559 --> 00:24:14,720
rather than moving big dishes around. At

676
00:24:13,440 --> 00:24:16,720
the core of the telescope were some

677
00:24:14,720 --> 00:24:18,559
monolithic C programs thousands of lines

678
00:24:16,720 --> 00:24:19,600
long running on a little server a few

679
00:24:18,559 --> 00:24:22,640
kilometers down the road from the

680
00:24:19,600 --> 00:24:24,960
receiver. Um, one colle they collected

681
00:24:22,640 --> 00:24:25,760
heaps of data um, correlations and

682
00:24:24,960 --> 00:24:29,279
things.

683
00:24:25,760 --> 00:24:31,200
Thank you so much, Christopher. Stop.

684
00:24:29,279 --> 00:24:33,520
Well done everyone. Thank you so much

685
00:24:31,200 --> 00:24:35,200
Christopher for taking us on a whole

686
00:24:33,520 --> 00:24:38,559
bunch of different topics. Up next on

687
00:24:35,200 --> 00:24:40,480
this side is Matt Sangia. But first

688
00:24:38,559 --> 00:24:43,679
taking us from running rivers to runtime

689
00:24:40,480 --> 00:24:45,760
semantics is Morgan Bartholomew.

690
00:24:43,679 --> 00:24:47,760
Hello everyone. Uh today I will be

691
00:24:45,760 --> 00:24:49,120
talking about binary operation runtime

692
00:24:47,760 --> 00:24:52,400
semantics and their type system

693
00:24:49,120 --> 00:24:54,080
implications. So I hope to explain what

694
00:24:52,400 --> 00:24:55,440
is happening in this expression by the

695
00:24:54,080 --> 00:24:58,799
end of this talk and what it means for

696
00:24:55,440 --> 00:25:01,679
type checkers. So if we look at a simple

697
00:24:58,799 --> 00:25:04,320
case of a object that supports a binary

698
00:25:01,679 --> 00:25:06,640
operation in this example plus um we

699
00:25:04,320 --> 00:25:08,960
just have a add dunder fairly simple to

700
00:25:06,640 --> 00:25:11,279
understand. We would then move on to the

701
00:25:08,960 --> 00:25:15,400
right hand dunder which operates when

702
00:25:11,279 --> 00:25:15,400
the value is on the right hand side.

703
00:25:15,760 --> 00:25:20,320
We also have the inline dunder which is

704
00:25:18,400 --> 00:25:23,039
called when you use the plus equals

705
00:25:20,320 --> 00:25:24,799
augmented assignment operator.

706
00:25:23,039 --> 00:25:27,919
The way that these work in combination

707
00:25:24,799 --> 00:25:30,559
is that one dunder will return the

708
00:25:27,919 --> 00:25:33,120
notimplemented singleton which will then

709
00:25:30,559 --> 00:25:35,679
be which will then pass the execution on

710
00:25:33,120 --> 00:25:37,840
to the right hand side objects right-

711
00:25:35,679 --> 00:25:40,080
hand version of that dunder. So in this

712
00:25:37,840 --> 00:25:41,840
case it would call the add dunder which

713
00:25:40,080 --> 00:25:44,240
will then return not implemented and

714
00:25:41,840 --> 00:25:47,520
then it will call writes rd dunder and

715
00:25:44,240 --> 00:25:49,200
return three.

716
00:25:47,520 --> 00:25:51,120
The next complexity in this system is

717
00:25:49,200 --> 00:25:52,960
the right hand subtype rule. So what

718
00:25:51,120 --> 00:25:56,000
this means is when the right hand side

719
00:25:52,960 --> 00:25:58,799
is a subtype of the type of the left

720
00:25:56,000 --> 00:26:01,279
hand side value then the interpreter

721
00:25:58,799 --> 00:26:06,880
will invoke the right hand dunder first

722
00:26:01,279 --> 00:26:09,279
before moving on to the normal dunder.

723
00:26:06,880 --> 00:26:12,240
These all combine when using a augmented

724
00:26:09,279 --> 00:26:14,320
assignment operator. So in this case we

725
00:26:12,240 --> 00:26:17,279
will see the I add being invoked first.

726
00:26:14,320 --> 00:26:20,480
It superseds the other operators. then

727
00:26:17,279 --> 00:26:23,480
the add operator, then the R add

728
00:26:20,480 --> 00:26:23,480
operator.

729
00:26:23,919 --> 00:26:27,520
A reimplementation of this in Python

730
00:26:25,600 --> 00:26:31,520
would look something like this. We start

731
00:26:27,520 --> 00:26:33,360
by checking the IAD. If it exists, we

732
00:26:31,520 --> 00:26:35,840
call it. If it returns not implemented,

733
00:26:33,360 --> 00:26:38,960
we move on. We then go through checking

734
00:26:35,840 --> 00:26:40,159
the RAD. We check the um subtype rule.

735
00:26:38,960 --> 00:26:42,480
We would then move on to checking the

736
00:26:40,159 --> 00:26:44,480
ad. And then finally, the R ad if it was

737
00:26:42,480 --> 00:26:45,919
not invoked previously. If none of these

738
00:26:44,480 --> 00:26:48,240
exist, then we'd raise an attribute

739
00:26:45,919 --> 00:26:48,240
error.

740
00:26:48,640 --> 00:26:54,240
So what do these rules what implications

741
00:26:52,320 --> 00:26:56,720
they have in the type system and type

742
00:26:54,240 --> 00:26:58,880
checkers?

743
00:26:56,720 --> 00:27:01,039
So currently if you annotated the return

744
00:26:58,880 --> 00:27:02,880
type as not implemented

745
00:27:01,039 --> 00:27:04,799
all of the type checkers will not

746
00:27:02,880 --> 00:27:06,320
understand these semantics and will say

747
00:27:04,799 --> 00:27:08,240
that this operator returns not

748
00:27:06,320 --> 00:27:11,600
implemented type when in reality this

749
00:27:08,240 --> 00:27:14,559
will raise an attribute error.

750
00:27:11,600 --> 00:27:16,159
This results in some unsoundness when

751
00:27:14,559 --> 00:27:17,679
this value will actually return a

752
00:27:16,159 --> 00:27:19,520
string, but the type checker thinks it

753
00:27:17,679 --> 00:27:22,520
returns an instance of not implemented

754
00:27:19,520 --> 00:27:22,520
type.

755
00:27:24,240 --> 00:27:29,279
In this case, at runtime, it's returning

756
00:27:27,120 --> 00:27:30,720
a string, but the type checker thinks

757
00:27:29,279 --> 00:27:32,640
it's returning an int. Oh, and by the

758
00:27:30,720 --> 00:27:35,640
way, I'm using ty for all of these

759
00:27:32,640 --> 00:27:35,640
examples.

760
00:27:36,000 --> 00:27:42,480
Um here when we have a

761
00:27:39,200 --> 00:27:44,159
right-hand operator that returns an

762
00:27:42,480 --> 00:27:47,600
incompatible type with the left hand

763
00:27:44,159 --> 00:27:50,080
operator, it's very important that we

764
00:27:47,600 --> 00:27:54,080
understand what the possible value will

765
00:27:50,080 --> 00:27:57,440
be. In this case, ty and it's um agrees

766
00:27:54,080 --> 00:27:59,120
with my here that it would return int

767
00:27:57,440 --> 00:28:02,320
which is wrong. it'll actually return a

768
00:27:59,120 --> 00:28:03,520
string. And this is because these type

769
00:28:02,320 --> 00:28:06,159
checkers have implemented the right-h

770
00:28:03,520 --> 00:28:08,320
hand rule incorrectly. And it assumes

771
00:28:06,159 --> 00:28:10,640
that because the left hand side is a

772
00:28:08,320 --> 00:28:13,200
left and it the right hand side is also

773
00:28:10,640 --> 00:28:15,039
a left, it does not take into

774
00:28:13,200 --> 00:28:16,640
consideration the possibility that the

775
00:28:15,039 --> 00:28:18,399
right hand is actually a subtype of the

776
00:28:16,640 --> 00:28:19,840
left hand. So it will say it returns an

777
00:28:18,399 --> 00:28:21,279
int when it will actually invoke the

778
00:28:19,840 --> 00:28:24,520
right hand operator and not the left

779
00:28:21,279 --> 00:28:24,520
hand operator.

780
00:28:25,600 --> 00:28:33,039
In this case, we have the inverse where

781
00:28:29,600 --> 00:28:36,640
the right hand side is

782
00:28:33,039 --> 00:28:39,039
being um wait what?

783
00:28:36,640 --> 00:28:43,039
Oh yeah, the other way around. The left

784
00:28:39,039 --> 00:28:45,279
hand side is um a subtype of the left

785
00:28:43,039 --> 00:28:47,039
hand type. So the type checker here

786
00:28:45,279 --> 00:28:49,200
thinks that the right hand side is a

787
00:28:47,039 --> 00:28:50,880
subtype of the left hand side value, but

788
00:28:49,200 --> 00:28:53,760
actually the value on the left is a

789
00:28:50,880 --> 00:28:55,520
subtype of the left class.

790
00:28:53,760 --> 00:28:58,159
And the right hand side is not a subtype

791
00:28:55,520 --> 00:28:59,600
of this imposter class. So the type

792
00:28:58,159 --> 00:29:01,919
checker is incorrectly applying the

793
00:28:59,600 --> 00:29:03,520
right hand side subtype rule when it

794
00:29:01,919 --> 00:29:04,880
shouldn't be. And it says it returns a

795
00:29:03,520 --> 00:29:07,360
string when it will actually return an

796
00:29:04,880 --> 00:29:08,880
integer.

797
00:29:07,360 --> 00:29:11,679
So how can we fix this? How can we

798
00:29:08,880 --> 00:29:14,480
update the type checkers? What possible

799
00:29:11,679 --> 00:29:15,760
rules could we imply to or what rules

800
00:29:14,480 --> 00:29:18,760
could we add to get the correct

801
00:29:15,760 --> 00:29:18,760
behaviors?

802
00:29:19,679 --> 00:29:26,880
We can add validations on the subtypes

803
00:29:24,399 --> 00:29:29,919
for the corresponding right- hand side

804
00:29:26,880 --> 00:29:31,840
operators when they have an incompatible

805
00:29:29,919 --> 00:29:35,360
return type with the super types left

806
00:29:31,840 --> 00:29:38,559
hand side operator. We will say that

807
00:29:35,360 --> 00:29:40,799
this okay has to maintain compatibility.

808
00:29:38,559 --> 00:29:44,840
More examples.

809
00:29:40,799 --> 00:29:44,840
More examples. Thank you.

810
00:29:45,039 --> 00:29:48,640
Ah,

811
00:29:46,880 --> 00:29:51,679
thanks. Thanks, Morgan. That is one of

812
00:29:48,640 --> 00:29:55,679
my favorite types of talk.

813
00:29:51,679 --> 00:29:58,559
Uh, Brianna, you are on deck. Um, when I

814
00:29:55,679 --> 00:30:00,000
wasuling today's lightning talks, um, I

815
00:29:58,559 --> 00:30:02,080
had great difficulty trying to figure

816
00:30:00,000 --> 00:30:04,399
out what to, uh, follow up a talk about

817
00:30:02,080 --> 00:30:06,480
binary operations with. Anyway, here's

818
00:30:04,399 --> 00:30:09,520
Matt. Please make them welcome.

819
00:30:06,480 --> 00:30:11,440
[Music]

820
00:30:09,520 --> 00:30:13,760
Thank you, Chris.

821
00:30:11,440 --> 00:30:17,200
Um, can you all hear me?

822
00:30:13,760 --> 00:30:19,679
Yep. Cool. Uh so uh I suspect a lot of

823
00:30:17,200 --> 00:30:21,600
you have heard about or used uh Jupyter

824
00:30:19,679 --> 00:30:24,960
notebooks before. If you haven't, uh

825
00:30:21,600 --> 00:30:26,640
Jupyter Notebook is a way to um

826
00:30:24,960 --> 00:30:28,320
basically via a web page create a

827
00:30:26,640 --> 00:30:30,159
notebook of information that is

828
00:30:28,320 --> 00:30:33,039
interactive and has runnable Python

829
00:30:30,159 --> 00:30:36,240
code. Um Jupyter Notebooks and Jupyter

830
00:30:33,039 --> 00:30:38,399
Lab and um IPython have been around for

831
00:30:36,240 --> 00:30:40,559
quite some time now. uh I think that

832
00:30:38,399 --> 00:30:42,480
they've built a bit of a legacy and have

833
00:30:40,559 --> 00:30:44,080
got some historical design decisions

834
00:30:42,480 --> 00:30:47,039
that I think that we could do a little

835
00:30:44,080 --> 00:30:48,640
bit better with these days with you know

836
00:30:47,039 --> 00:30:49,679
now that we've learned from some of the

837
00:30:48,640 --> 00:30:52,640
stuff that's been that's been done

838
00:30:49,679 --> 00:30:54,799
there. Um and a couple years ago I

839
00:30:52,640 --> 00:30:58,640
learned about a new up and coming

840
00:30:54,799 --> 00:31:00,399
notebook framework called MIMO. Um and

841
00:30:58,640 --> 00:31:02,559
so I'm going to tell you a little bit

842
00:31:00,399 --> 00:31:04,480
about it. So let's have a look at this.

843
00:31:02,559 --> 00:31:07,520
This is an IPython notebook for now.

844
00:31:04,480 --> 00:31:09,039
Sorry, a a Jupyter notebook. So we've

845
00:31:07,520 --> 00:31:13,279
got a notebook. It's running in our

846
00:31:09,039 --> 00:31:16,720
browser. Um if I go to display uh the

847
00:31:13,279 --> 00:31:19,120
contents of this notebook. Um so we're

848
00:31:16,720 --> 00:31:20,559
looking at the the file that contains

849
00:31:19,120 --> 00:31:22,799
the Python notebook. If I have a look at

850
00:31:20,559 --> 00:31:25,360
that. Oh, this is what happens when I do

851
00:31:22,799 --> 00:31:28,559
live demos, right? Um so the issue here

852
00:31:25,360 --> 00:31:31,200
is that I have run a cell and the that

853
00:31:28,559 --> 00:31:34,080
cell depended on the output of another

854
00:31:31,200 --> 00:31:36,080
cell. And so I have to run the notebook

855
00:31:34,080 --> 00:31:38,799
in a particular order, potentially

856
00:31:36,080 --> 00:31:41,039
sequentially from top to bottom, but not

857
00:31:38,799 --> 00:31:43,600
necessarily. Uh, in this case, I might

858
00:31:41,039 --> 00:31:46,000
want to put all of my imports uh for a

859
00:31:43,600 --> 00:31:49,120
notebook down the bottom out of the way

860
00:31:46,000 --> 00:31:52,159
and be able to use those further up, but

861
00:31:49,120 --> 00:31:53,679
I can't do that easily um with Jupiter

862
00:31:52,159 --> 00:31:55,760
because it's not going to know to go and

863
00:31:53,679 --> 00:31:57,120
run that. Um, so I'm going to go to the

864
00:31:55,760 --> 00:31:59,679
cell below. I'm going to pull in my

865
00:31:57,120 --> 00:32:01,519
imports, and then I'm going to fail

866
00:31:59,679 --> 00:32:04,000
miserably. Let's open that one. run that

867
00:32:01,519 --> 00:32:05,600
one as well. There we go. So, we're

868
00:32:04,000 --> 00:32:07,200
looking at the contents of this of this

869
00:32:05,600 --> 00:32:10,159
notebook file. And we can see it's a

870
00:32:07,200 --> 00:32:12,240
whole lot of JSON. Um, it's not super

871
00:32:10,159 --> 00:32:13,360
easy to maintain or work with. And if I

872
00:32:12,240 --> 00:32:15,360
wanted to just run this as a regular

873
00:32:13,360 --> 00:32:16,399
Python script, I'm going to have a few

874
00:32:15,360 --> 00:32:19,039
challenges. I'm going to have to go and

875
00:32:16,399 --> 00:32:21,519
convert it. It's a bit of a pain. Um, so

876
00:32:19,039 --> 00:32:24,240
let's have a look at the way this. So,

877
00:32:21,519 --> 00:32:26,240
if I open up my MIMO notebook,

878
00:32:24,240 --> 00:32:28,159
um, if I have a look at the contents of

879
00:32:26,240 --> 00:32:32,240
this notebook,

880
00:32:28,159 --> 00:32:35,760
I need to go and enable this cell.

881
00:32:32,240 --> 00:32:39,440
Um this notebook

882
00:32:35,760 --> 00:32:43,440
uh this notebook is pure Python. So it's

883
00:32:39,440 --> 00:32:45,600
imported marino marrimo sorry and each

884
00:32:43,440 --> 00:32:49,200
cell in the notebook is an anonymous

885
00:32:45,600 --> 00:32:50,880
function. So we can

886
00:32:49,200 --> 00:32:52,240
go and copy the code out of each cell

887
00:32:50,880 --> 00:32:53,840
and put it into a regular Python script

888
00:32:52,240 --> 00:32:56,159
and it will just work. You can also use

889
00:32:53,840 --> 00:32:59,279
MIMO to run the notebook as a regular

890
00:32:56,159 --> 00:33:01,440
Python script and it's I think rather

891
00:32:59,279 --> 00:33:03,360
easy to read. So if I go a little bit

892
00:33:01,440 --> 00:33:05,200
further down, one of the other benefits

893
00:33:03,360 --> 00:33:07,600
of Mimo is that it's built in with a

894
00:33:05,200 --> 00:33:09,840
whole bunch of stuff that didn't come

895
00:33:07,600 --> 00:33:12,000
with Jupiter notebooks but has, you

896
00:33:09,840 --> 00:33:14,240
know, is available as add-ons, but um

897
00:33:12,000 --> 00:33:16,320
MIMO has a lot more stuff that they make

898
00:33:14,240 --> 00:33:19,200
natively available. One of those things

899
00:33:16,320 --> 00:33:21,039
is duck DB. So um one of the things that

900
00:33:19,200 --> 00:33:25,600
I can do with duct DB is I can say I

901
00:33:21,039 --> 00:33:28,320
would like to open uh a JSON file and

902
00:33:25,600 --> 00:33:31,279
put it into a pandas dataf frame and I

903
00:33:28,320 --> 00:33:33,279
can view that data frame here in a

904
00:33:31,279 --> 00:33:34,960
pretty table with a whole bunch of like

905
00:33:33,279 --> 00:33:37,039
frequency distribution stuff. And then

906
00:33:34,960 --> 00:33:39,600
what I've done to view the table is I

907
00:33:37,039 --> 00:33:41,120
have used raw SQL. I haven't had to do

908
00:33:39,600 --> 00:33:42,880
anything else in this cell. This is an

909
00:33:41,120 --> 00:33:44,799
SQL

910
00:33:42,880 --> 00:33:48,000
an SQL cell and I've been able to use

911
00:33:44,799 --> 00:33:50,559
raw SQL to display the contents of this

912
00:33:48,000 --> 00:33:52,480
spreadsheet this um table. Likewise, I

913
00:33:50,559 --> 00:33:54,960
can use other raw SQL where I can go and

914
00:33:52,480 --> 00:33:56,480
filter it. Um I could change this to

915
00:33:54,960 --> 00:33:58,000
Europe

916
00:33:56,480 --> 00:34:00,240
and it'll go and change the contents

917
00:33:58,000 --> 00:34:02,960
there. Um, and the other thing that I

918
00:34:00,240 --> 00:34:04,720
can do is I can have, and this is

919
00:34:02,960 --> 00:34:06,080
available again in Jupiter as well, but

920
00:34:04,720 --> 00:34:09,760
um, I feel like it was a little easier

921
00:34:06,080 --> 00:34:12,480
for me to get this working in Mimo. Um,

922
00:34:09,760 --> 00:34:14,639
was that I can add a widget here, and I

923
00:34:12,480 --> 00:34:16,079
can say, "Show me cars that are that

924
00:34:14,639 --> 00:34:18,960
have at least this many miles per

925
00:34:16,079 --> 00:34:20,639
gallon." Um, and we can also see that

926
00:34:18,960 --> 00:34:22,879
way down the bottom here, here are my

927
00:34:20,639 --> 00:34:26,000
imports for Maro and Pandas. And the way

928
00:34:22,879 --> 00:34:27,440
this works is that each, if I go back up

929
00:34:26,000 --> 00:34:29,040
to the top here, I've got 40 seconds.

930
00:34:27,440 --> 00:34:30,639
This is fine. If I go back back up the

931
00:34:29,040 --> 00:34:35,040
top, we can see that the cell at the

932
00:34:30,639 --> 00:34:38,639
bottom returns my two imports.

933
00:34:35,040 --> 00:34:41,040
And if I go further up to the top,

934
00:34:38,639 --> 00:34:43,359
uh where's it gone? Here, we can see

935
00:34:41,040 --> 00:34:46,560
that it takes in

936
00:34:43,359 --> 00:34:49,280
a lot of those imports as arguments so

937
00:34:46,560 --> 00:34:50,800
that it can do internal dependency

938
00:34:49,280 --> 00:34:52,639
ordering without me having to think

939
00:34:50,800 --> 00:34:54,399
about it. And if I were to run this

940
00:34:52,639 --> 00:34:56,079
cell, it would make sure that all the

941
00:34:54,399 --> 00:34:57,119
the prior cells have already been run.

942
00:34:56,079 --> 00:34:58,640
There's a whole bunch of other stuff

943
00:34:57,119 --> 00:35:00,560
that you can do with MIMO. I highly

944
00:34:58,640 --> 00:35:04,359
recommend you check it out.app and

945
00:35:00,560 --> 00:35:04,359
marrimo.io are there.

946
00:35:08,160 --> 00:35:12,320
Thank you so much, Matt. Um, while

947
00:35:10,560 --> 00:35:13,760
lightning talks are going on, we're

948
00:35:12,320 --> 00:35:15,440
setting up the next person, but all I

949
00:35:13,760 --> 00:35:16,640
can just hear is ooze every 10 seconds.

950
00:35:15,440 --> 00:35:20,480
I'm really excited to go back and watch

951
00:35:16,640 --> 00:35:22,640
that one. Um, up on this side next is

952
00:35:20,480 --> 00:35:25,119
Evan. But in the meantime, it's not

953
00:35:22,640 --> 00:35:28,720
lightning talks unless there's some

954
00:35:25,119 --> 00:35:33,160
controversy. So, Brianna Argo is here to

955
00:35:28,720 --> 00:35:33,160
talk to us about some more about fish.

956
00:35:34,160 --> 00:35:38,119
So, I thought it would be

957
00:35:38,160 --> 00:35:42,079
I thought given that it's my first pyon,

958
00:35:40,240 --> 00:35:43,760
the way to make a good impression at the

959
00:35:42,079 --> 00:35:46,480
start would be to challenge something

960
00:35:43,760 --> 00:35:49,480
someone said yesterday.

961
00:35:46,480 --> 00:35:49,480
Stop.

962
00:35:49,520 --> 00:35:54,960
So why listen to me? To establish some

963
00:35:52,000 --> 00:35:56,560
credibility, uh I'm currently a masters

964
00:35:54,960 --> 00:35:59,040
of health data science student, but

965
00:35:56,560 --> 00:36:01,040
previously I majored in ecology and

966
00:35:59,040 --> 00:36:02,880
conservation biology. And as part of

967
00:36:01,040 --> 00:36:05,040
that, I did an honors project and I've

968
00:36:02,880 --> 00:36:06,560
copied some slides directly that had

969
00:36:05,040 --> 00:36:08,720
nothing to do with taxonomy or

970
00:36:06,560 --> 00:36:11,200
classification. But as you can see,

971
00:36:08,720 --> 00:36:13,359
there's a picture of a fish on this

972
00:36:11,200 --> 00:36:15,920
slide.

973
00:36:13,359 --> 00:36:17,680
And if we look at the next one, it's got

974
00:36:15,920 --> 00:36:19,280
some equations and there's some partial

975
00:36:17,680 --> 00:36:22,190
derivatives. So clearly I'm a very

976
00:36:19,280 --> 00:36:24,720
technical and competent person.

977
00:36:22,190 --> 00:36:26,720
[Applause]

978
00:36:24,720 --> 00:36:28,640
So let's get on to the content of

979
00:36:26,720 --> 00:36:31,040
phoggenetic trees. These can also be

980
00:36:28,640 --> 00:36:32,960
called clatograms. And what you can see

981
00:36:31,040 --> 00:36:35,440
is the relationships of ancestry here.

982
00:36:32,960 --> 00:36:37,680
So you've got some common ancestor

983
00:36:35,440 --> 00:36:41,119
branches down and then you can see the

984
00:36:37,680 --> 00:36:44,720
subdivisions of the filer there. In a

985
00:36:41,119 --> 00:36:46,240
monopillettic tree, you have the recent

986
00:36:44,720 --> 00:36:48,320
common ancestor and all of the

987
00:36:46,240 --> 00:36:51,520
descendants a group together into the

988
00:36:48,320 --> 00:36:54,560
one monopilletic group. Another example

989
00:36:51,520 --> 00:36:58,560
of that is shown in this instance.

990
00:36:54,560 --> 00:37:01,440
But controversially perhaps other types

991
00:36:58,560 --> 00:37:03,760
of groupings are possible. So we can see

992
00:37:01,440 --> 00:37:07,040
here we have a polyilletic grouping

993
00:37:03,760 --> 00:37:09,599
where you do not have um everything

994
00:37:07,040 --> 00:37:11,040
flowing from the one common ancestor. we

995
00:37:09,599 --> 00:37:13,599
have different most recent common

996
00:37:11,040 --> 00:37:15,040
ancestors involved.

997
00:37:13,599 --> 00:37:18,640
So the takeaway here in terms of

998
00:37:15,040 --> 00:37:21,520
interpreting a follow a phogenetic tree

999
00:37:18,640 --> 00:37:23,200
is that in a monopilotic group you do

1000
00:37:21,520 --> 00:37:25,760
need to share the common ancestors and

1001
00:37:23,200 --> 00:37:27,200
so that rule about if these members are

1002
00:37:25,760 --> 00:37:30,000
in the group then we know that this

1003
00:37:27,200 --> 00:37:31,680
other member such as humans for instance

1004
00:37:30,000 --> 00:37:34,079
has to be in the group as well would

1005
00:37:31,680 --> 00:37:37,119
hold. If you don't have a monopilletic

1006
00:37:34,079 --> 00:37:38,800
group that's not necessarily the case.

1007
00:37:37,119 --> 00:37:42,880
I thought for my talk I would bring up

1008
00:37:38,800 --> 00:37:44,560
the definition of fish. And

1009
00:37:42,880 --> 00:37:47,119
so this is the first screenshot I found.

1010
00:37:44,560 --> 00:37:48,560
I just went to Mariam Webster. And here

1011
00:37:47,119 --> 00:37:50,560
it actually presented me with somewhat

1012
00:37:48,560 --> 00:37:53,280
of a problem. Uh you might see I've put

1013
00:37:50,560 --> 00:37:54,720
a box around starfish. And the problem

1014
00:37:53,280 --> 00:37:56,160
here is it shows me as a bit of a

1015
00:37:54,720 --> 00:37:58,000
hypocrite because I looked at that and I

1016
00:37:56,160 --> 00:38:00,000
went that's not starfish. We should be

1017
00:37:58,000 --> 00:38:02,079
calling them sea stars because they're

1018
00:38:00,000 --> 00:38:05,079
not fish. And I know that because of

1019
00:38:02,079 --> 00:38:05,079
phogenetics.

1020
00:38:05,200 --> 00:38:12,400
Stop.

1021
00:38:07,119 --> 00:38:15,280
If we if we look at the next uh proposed

1022
00:38:12,400 --> 00:38:16,800
definition um we can see that it's

1023
00:38:15,280 --> 00:38:19,119
describing some of the characteristics.

1024
00:38:16,800 --> 00:38:23,520
So it's telling us about the morphology

1025
00:38:19,119 --> 00:38:27,680
and we have that included in so I prefer

1026
00:38:23,520 --> 00:38:29,839
the second version sits happier with me.

1027
00:38:27,680 --> 00:38:31,119
But my morphology based definitions can

1028
00:38:29,839 --> 00:38:33,520
be quite limiting and have their own

1029
00:38:31,119 --> 00:38:36,320
issues as well. So if I say for instance

1030
00:38:33,520 --> 00:38:38,960
that a mammal is something that produces

1031
00:38:36,320 --> 00:38:41,440
milk and it has fur, then plenty of

1032
00:38:38,960 --> 00:38:43,280
mammals could be categorized under that.

1033
00:38:41,440 --> 00:38:45,390
But you might also have a bad drawing of

1034
00:38:43,280 --> 00:38:48,000
a coconut.

1035
00:38:45,390 --> 00:38:50,000
[Music]

1036
00:38:48,000 --> 00:38:51,680
But if we want to use this phoggenetic

1037
00:38:50,000 --> 00:38:53,520
based approach, then we might want to

1038
00:38:51,680 --> 00:38:54,800
consider what are our nodes. And usually

1039
00:38:53,520 --> 00:38:56,720
when you zoom in at the bottom, you

1040
00:38:54,800 --> 00:38:58,560
might have different species. So it's

1041
00:38:56,720 --> 00:39:01,280
important that we can consistently

1042
00:38:58,560 --> 00:39:03,440
consider what is a species. If we look

1043
00:39:01,280 --> 00:39:05,680
at the process of something becoming two

1044
00:39:03,440 --> 00:39:08,160
different species, we start with the one

1045
00:39:05,680 --> 00:39:10,320
population, they become more different

1046
00:39:08,160 --> 00:39:12,079
over time due to limited gene flow.

1047
00:39:10,320 --> 00:39:14,560
Typically there, this is some geographic

1048
00:39:12,079 --> 00:39:16,480
barrier. Um, and then because they're

1049
00:39:14,560 --> 00:39:18,000
not having that gene flow to keep them

1050
00:39:16,480 --> 00:39:20,400
more similar with each other,

1051
00:39:18,000 --> 00:39:21,760
differences increase until eventually

1052
00:39:20,400 --> 00:39:23,920
they get to the point where we've got

1053
00:39:21,760 --> 00:39:25,920
these two subpopuls that have split off.

1054
00:39:23,920 --> 00:39:27,680
They're so different members of them

1055
00:39:25,920 --> 00:39:29,599
can't create fertile offspring with each

1056
00:39:27,680 --> 00:39:31,520
other. And according to the biological

1057
00:39:29,599 --> 00:39:34,160
species concept, that's when they're

1058
00:39:31,520 --> 00:39:36,320
different species.

1059
00:39:34,160 --> 00:39:38,880
But there are many issues with this

1060
00:39:36,320 --> 00:39:41,680
definition as well. So for instance, we

1061
00:39:38,880 --> 00:39:44,240
have ring species. Imagine a group of

1062
00:39:41,680 --> 00:39:46,400
lizards all around this canyon and

1063
00:39:44,240 --> 00:39:48,160
pair-wise they have that compatibility,

1064
00:39:46,400 --> 00:39:49,599
but the ones far across from each other

1065
00:39:48,160 --> 00:39:52,320
don't. Would you consider them the same

1066
00:39:49,599 --> 00:39:55,119
species or not? Another thing we might

1067
00:39:52,320 --> 00:39:57,440
consider is what if they're an asexual

1068
00:39:55,119 --> 00:39:59,680
type of lizard and so we don't have that

1069
00:39:57,440 --> 00:40:02,480
reproduction criteria to look at. You

1070
00:39:59,680 --> 00:40:04,160
could think even differently. What about

1071
00:40:02,480 --> 00:40:07,440
if we have viruses and you've got

1072
00:40:04,160 --> 00:40:10,240
horizontal gene transfer? So there is no

1073
00:40:07,440 --> 00:40:12,079
perfect objective generalizable rule set

1074
00:40:10,240 --> 00:40:14,160
for classifying life. My righteous

1075
00:40:12,079 --> 00:40:18,040
indignation that I had yesterday might

1076
00:40:14,160 --> 00:40:18,040
have been a little bit mistaken.

1077
00:40:18,400 --> 00:40:23,119
And I think the takeaway here is to

1078
00:40:20,720 --> 00:40:24,960
consider what makes sense for your users

1079
00:40:23,119 --> 00:40:26,640
and what makes sense for the purpose and

1080
00:40:24,960 --> 00:40:29,640
context you're applying it for. Thank

1081
00:40:26,640 --> 00:40:29,640
you.

1082
00:40:34,320 --> 00:40:38,880
Oh, thank you for that. I I love these

1083
00:40:36,400 --> 00:40:40,960
response talks. Um, if you want to give

1084
00:40:38,880 --> 00:40:42,560
a response talk to anything today,

1085
00:40:40,960 --> 00:40:44,400
you've got a whole year to think about

1086
00:40:42,560 --> 00:40:46,079
it, which is going to make you look

1087
00:40:44,400 --> 00:40:48,079
extra deranged and like you're holding a

1088
00:40:46,079 --> 00:40:50,160
grudge and such, which we really like.

1089
00:40:48,079 --> 00:40:52,079
So, do be thinking about your response

1090
00:40:50,160 --> 00:40:55,119
lightning talks for the uh for the next

1091
00:40:52,079 --> 00:40:57,760
12-ish months. Uh, Ash, you are on deck.

1092
00:40:55,119 --> 00:41:00,079
Um,

1093
00:40:57,760 --> 00:41:03,359
Evan's not your parents, but he is going

1094
00:41:00,079 --> 00:41:05,920
to tell you how to eat.

1095
00:41:03,359 --> 00:41:08,400
Hello. Hello. Can you hear me? Okay,

1096
00:41:05,920 --> 00:41:10,160
great. Great. Awesome. Hello. Now, I I'm

1097
00:41:08,400 --> 00:41:11,839
not a fruit connoisseur by any means,

1098
00:41:10,160 --> 00:41:14,400
but I've noticed that no one knows how

1099
00:41:11,839 --> 00:41:16,800
to eat fruit. So, here we go. U first

1100
00:41:14,400 --> 00:41:18,880
we'll start Eezy with oranges, the most

1101
00:41:16,800 --> 00:41:20,800
illustrious of all fruit. Cut them in

1102
00:41:18,880 --> 00:41:22,480
half and then into thirds. And this

1103
00:41:20,800 --> 00:41:25,680
allows you to easily eat off the

1104
00:41:22,480 --> 00:41:27,760
individual segments by bending the skin.

1105
00:41:25,680 --> 00:41:30,160
Mandarins kind of like an orange, but

1106
00:41:27,760 --> 00:41:32,960
the inner segments are already uh pulled

1107
00:41:30,160 --> 00:41:34,319
apart. Peel off a ribbon of skin with

1108
00:41:32,960 --> 00:41:35,920
your fingers and then keep the

1109
00:41:34,319 --> 00:41:37,839
hemisphere to hold the skin and the

1110
00:41:35,920 --> 00:41:39,920
seeds.

1111
00:41:37,839 --> 00:41:42,079
You have the apple classic. You take

1112
00:41:39,920 --> 00:41:44,319
bites out without peeling and once you

1113
00:41:42,079 --> 00:41:46,880
get to the core, don't be wasteful and

1114
00:41:44,319 --> 00:41:48,319
stop. Finish it off. Nothing bad can

1115
00:41:46,880 --> 00:41:50,640
happen to you since they keep the doctor

1116
00:41:48,319 --> 00:41:53,119
away.

1117
00:41:50,640 --> 00:41:56,400
Pears are basically messily green apples

1118
00:41:53,119 --> 00:41:59,200
best cut into slices. Prickly pears are

1119
00:41:56,400 --> 00:42:01,119
not pears, but the fruits of cacti. They

1120
00:41:59,200 --> 00:42:02,960
are covered in spikes and they're best

1121
00:42:01,119 --> 00:42:05,119
eaten by quickly torching them over

1122
00:42:02,960 --> 00:42:08,640
fire, washing and then filleting their

1123
00:42:05,119 --> 00:42:11,839
skin to reveal their juicy innards.

1124
00:42:08,640 --> 00:42:14,480
Tomatoes are not vegetables. They have

1125
00:42:11,839 --> 00:42:16,560
seeds inside and thus they are fruit.

1126
00:42:14,480 --> 00:42:20,079
Cut into segments and top off with your

1127
00:42:16,560 --> 00:42:22,240
favorite salt. Honestly, delicious.

1128
00:42:20,079 --> 00:42:24,160
Lemons are popular ingredients, but they

1129
00:42:22,240 --> 00:42:26,480
need more love than that. Have you tried

1130
00:42:24,160 --> 00:42:28,800
eating them raw like an orange? Very

1131
00:42:26,480 --> 00:42:31,760
nice.

1132
00:42:28,800 --> 00:42:34,640
Strawberries, they are not berries or

1133
00:42:31,760 --> 00:42:37,119
fruit. They are the thing that holds the

1134
00:42:34,640 --> 00:42:39,040
seeds, which aren't seeds, but the

1135
00:42:37,119 --> 00:42:40,880
fruits itself.

1136
00:42:39,040 --> 00:42:43,839
Anyways, just cut off the green part and

1137
00:42:40,880 --> 00:42:46,000
eat them whole.

1138
00:42:43,839 --> 00:42:49,040
Avocados

1139
00:42:46,000 --> 00:42:51,040
have a seed inside and thus are a fruit.

1140
00:42:49,040 --> 00:42:53,599
First, cut along the shortest axis to

1141
00:42:51,040 --> 00:42:56,800
minimize surface area.

1142
00:42:53,599 --> 00:43:00,079
It will last longer. Then spoon out the

1143
00:42:56,800 --> 00:43:03,520
flesh, throw away the flesh, and keep

1144
00:43:00,079 --> 00:43:06,400
the seed to admire its beauty.

1145
00:43:03,520 --> 00:43:09,280
This is why they're so expensive.

1146
00:43:06,400 --> 00:43:11,359
Banana, the ultimate weapon. Most people

1147
00:43:09,280 --> 00:43:13,200
would eat a banana like this, peeling

1148
00:43:11,359 --> 00:43:15,760
from the stem. But this is a grave

1149
00:43:13,200 --> 00:43:18,319
mistake. Even primordial primates will

1150
00:43:15,760 --> 00:43:20,640
get this right. Pinch from the tip and

1151
00:43:18,319 --> 00:43:22,960
eat as you do.

1152
00:43:20,640 --> 00:43:25,920
Figs. Another one that's not actually a

1153
00:43:22,960 --> 00:43:28,720
fruit, but a sconium, which is an inside

1154
00:43:25,920 --> 00:43:34,319
out group of lots of teeny tiny flowers

1155
00:43:28,720 --> 00:43:36,960
that all bear not seeds, but fruit,

1156
00:43:34,319 --> 00:43:38,400
like an inside out strawberry.

1157
00:43:36,960 --> 00:43:41,920
There's no leaves on these, so just eat

1158
00:43:38,400 --> 00:43:44,480
them whole. A kiwi, often mistaken as

1159
00:43:41,920 --> 00:43:47,680
the bird due to resemblance. Normally

1160
00:43:44,480 --> 00:43:50,720
cut in half

1161
00:43:47,680 --> 00:43:53,680
and eaten with eaten with a spoon, but

1162
00:43:50,720 --> 00:43:57,119
best eaten like an apple with a skin on,

1163
00:43:53,680 --> 00:43:59,520
letting you eat it wherever you are.

1164
00:43:57,119 --> 00:44:02,160
Mango. Most people cut off the cheeks to

1165
00:43:59,520 --> 00:44:03,920
dice, but have you tried peeling off its

1166
00:44:02,160 --> 00:44:07,760
skin?

1167
00:44:03,920 --> 00:44:09,119
Now you can eat it like an apple.

1168
00:44:07,760 --> 00:44:11,920
Blueberries. Did you know that

1169
00:44:09,119 --> 00:44:13,599
blueberries can be white on the inside?

1170
00:44:11,920 --> 00:44:18,520
So when you're eating them, make sure to

1171
00:44:13,599 --> 00:44:18,520
cut each one in half to double check

1172
00:44:18,560 --> 00:44:23,520
before devouring them whole.

1173
00:44:21,359 --> 00:44:26,640
Grapes, the food of the gods, most

1174
00:44:23,520 --> 00:44:29,920
oftenly mistaken. Uh most often

1175
00:44:26,640 --> 00:44:31,599
mistakenly eaten whole and one by one,

1176
00:44:29,920 --> 00:44:34,079
but should actually have their skin

1177
00:44:31,599 --> 00:44:37,359
peeled before consumption. Try it. You

1178
00:44:34,079 --> 00:44:40,319
really don't need the fiber.

1179
00:44:37,359 --> 00:44:42,880
Blackberries. Oh, wait a minute. Yeah,

1180
00:44:40,319 --> 00:44:45,280
they they they look the same to me. And

1181
00:44:42,880 --> 00:44:47,599
if you think about it, no one eats

1182
00:44:45,280 --> 00:44:50,960
grapes whole with the vine. So, make

1183
00:44:47,599 --> 00:44:53,760
sure you pluck each one individually

1184
00:44:50,960 --> 00:44:55,680
to savor them as much as you can.

1185
00:44:53,760 --> 00:44:57,040
Pineapple. My favorites. They're

1186
00:44:55,680 --> 00:44:58,560
actually made through a fusion of

1187
00:44:57,040 --> 00:45:00,720
flowers that are out to kill you with

1188
00:44:58,560 --> 00:45:02,240
their spikiness. Although there's some

1189
00:45:00,720 --> 00:45:03,680
cutting methods that are superior,

1190
00:45:02,240 --> 00:45:05,920
they're all difficult, so just hack it

1191
00:45:03,680 --> 00:45:08,319
with a peeler.

1192
00:45:05,920 --> 00:45:10,240
Chilies are my arch nemesis. You can see

1193
00:45:08,319 --> 00:45:12,560
they have seeds inside and thus they're

1194
00:45:10,240 --> 00:45:15,440
a fruit. To eat, don't cut them like

1195
00:45:12,560 --> 00:45:17,680
this. Cut them head to tail. Remove all

1196
00:45:15,440 --> 00:45:19,680
of the seeds. And since it no longer has

1197
00:45:17,680 --> 00:45:23,000
seeds, it is no longer a fruit and can

1198
00:45:19,680 --> 00:45:23,000
go in the bin.

1199
00:45:24,640 --> 00:45:30,480
So, this talk for some might have been

1200
00:45:27,839 --> 00:45:32,319
funny or infuriating, but why should we

1201
00:45:30,480 --> 00:45:35,280
shame people for their preferences or

1202
00:45:32,319 --> 00:45:37,520
upbringings? Eat fruit however you want

1203
00:45:35,280 --> 00:45:39,410
and apply that principle to the rest of

1204
00:45:37,520 --> 00:45:40,670
your life. Thank you.

1205
00:45:39,410 --> 00:45:42,280
[Applause]

1206
00:45:40,670 --> 00:45:43,400
[Music]

1207
00:45:42,280 --> 00:45:46,550
[Applause]

1208
00:45:43,400 --> 00:45:46,550
[Music]

1209
00:45:48,160 --> 00:45:51,960
Thank you. I think

1210
00:45:54,560 --> 00:45:58,319
yeah

1211
00:45:56,720 --> 00:46:01,599
up next on this side while I process

1212
00:45:58,319 --> 00:46:04,079
what just happened is Alex Verico. But

1213
00:46:01,599 --> 00:46:06,240
first, uh, we've had a talk from Brianna

1214
00:46:04,079 --> 00:46:08,560
about jeans. We had a talk from Evan

1215
00:46:06,240 --> 00:46:10,720
effectively about jeans. So, we had to

1216
00:46:08,560 --> 00:46:12,810
tie it up with a third talk about jeans.

1217
00:46:10,720 --> 00:46:13,000
Please make Alex feel welcome.

1218
00:46:12,810 --> 00:46:15,520
[Music]

1219
00:46:13,000 --> 00:46:17,920
[Applause]

1220
00:46:15,520 --> 00:46:20,720
Hi. Uh, I'm here to talk about pants or

1221
00:46:17,920 --> 00:46:22,160
as Daisy alluded, jeans. So, who here is

1222
00:46:20,720 --> 00:46:25,359
wearing jeans? Can I see a show of

1223
00:46:22,160 --> 00:46:28,480
hands? A lot of you. Uh, how many of you

1224
00:46:25,359 --> 00:46:30,000
have a favorite pair of jeans?

1225
00:46:28,480 --> 00:46:31,839
Do you know why it's your favorite pair

1226
00:46:30,000 --> 00:46:33,760
of jeans? I don't need an answer now,

1227
00:46:31,839 --> 00:46:35,040
but come see me after the talk. I would

1228
00:46:33,760 --> 00:46:38,880
actually genuinely love to hear about

1229
00:46:35,040 --> 00:46:41,760
it. So, last year, this pair of pants

1230
00:46:38,880 --> 00:46:44,160
was voted the best pair of jeans in the

1231
00:46:41,760 --> 00:46:45,599
world, or rather the best faded pair of

1232
00:46:44,160 --> 00:46:48,079
jeans in the world. Look at those

1233
00:46:45,599 --> 00:46:51,280
whiskers, those combs, those stacks.

1234
00:46:48,079 --> 00:46:53,760
Very impressive stuff. But Ash, you ask,

1235
00:46:51,280 --> 00:46:56,560
how do we judge the best pair of jeans

1236
00:46:53,760 --> 00:46:58,400
in the world? Well, enter this

1237
00:46:56,560 --> 00:47:01,920
happylooking Swedish man on the right.

1238
00:46:58,400 --> 00:47:05,760
His name is Rudy Kra and he runs the

1239
00:47:01,920 --> 00:47:08,240
genes museum with over 14,000 samples

1240
00:47:05,760 --> 00:47:10,720
that he's collected over the past 50 or

1241
00:47:08,240 --> 00:47:13,359
60 years or so. And he's the final judge

1242
00:47:10,720 --> 00:47:17,280
in a jeans fading competition called the

1243
00:47:13,359 --> 00:47:20,079
Indigo Invitational. Now, the Indigo

1244
00:47:17,280 --> 00:47:22,319
Invitational goes for 365 days and some

1245
00:47:20,079 --> 00:47:24,480
people have very specific techniques and

1246
00:47:22,319 --> 00:47:26,960
strategies to their jeans fading. The

1247
00:47:24,480 --> 00:47:29,119
most dedicated will go as long as they

1248
00:47:26,960 --> 00:47:30,800
can without washing their jeans, aiming

1249
00:47:29,119 --> 00:47:32,960
to produce higher contrast wear

1250
00:47:30,800 --> 00:47:35,760
patterns. Some even going the whole year

1251
00:47:32,960 --> 00:47:38,560
without washing. Here's a pair the day

1252
00:47:35,760 --> 00:47:40,000
that the competition started last year.

1253
00:47:38,560 --> 00:47:41,520
And this is what they looked like at the

1254
00:47:40,000 --> 00:47:43,760
end.

1255
00:47:41,520 --> 00:47:45,599
Yeah. Uh a little gross. These uh these

1256
00:47:43,760 --> 00:47:47,839
jeans look like they've had a a pretty

1257
00:47:45,599 --> 00:47:51,839
rough life. Um this is absurd, right?

1258
00:47:47,839 --> 00:47:54,000
Surely only the most dedicated, most

1259
00:47:51,839 --> 00:47:56,079
hygienically challenged people would

1260
00:47:54,000 --> 00:47:58,560
ever consider taking part in this

1261
00:47:56,079 --> 00:48:00,400
competition. So naturally, I decided to

1262
00:47:58,560 --> 00:48:02,640
enter.

1263
00:48:00,400 --> 00:48:05,040
This is the pair that I chose this year.

1264
00:48:02,640 --> 00:48:08,319
I'm wearing them right now. It's a 14.5

1265
00:48:05,040 --> 00:48:09,760
oz double black jeans from Nudie. Uh

1266
00:48:08,319 --> 00:48:11,119
they worked for me in a style that I

1267
00:48:09,760 --> 00:48:12,640
liked. They're roomy enough for me to

1268
00:48:11,119 --> 00:48:13,839
climb in. Um I think they're pretty

1269
00:48:12,640 --> 00:48:17,680
neat. They've got a little orange

1270
00:48:13,839 --> 00:48:19,920
salvage ID. uh sulfur rope dyed warp and

1271
00:48:17,680 --> 00:48:22,319
a colorast weft. I don't have time to

1272
00:48:19,920 --> 00:48:23,839
explain that. Ask me after the the talk.

1273
00:48:22,319 --> 00:48:27,599
Um and I think they should produce some

1274
00:48:23,839 --> 00:48:29,920
interesting outcomes. So why enter this

1275
00:48:27,599 --> 00:48:31,760
competition? Well, in a world of fast

1276
00:48:29,920 --> 00:48:34,319
fashion, constant trend cycles, and

1277
00:48:31,760 --> 00:48:36,480
predistressed items, personalization of

1278
00:48:34,319 --> 00:48:38,720
an object through consistent dedicated

1279
00:48:36,480 --> 00:48:40,880
wear becomes a little bit of an oddity.

1280
00:48:38,720 --> 00:48:42,720
And if you're really odd like me, you're

1281
00:48:40,880 --> 00:48:44,880
also probably interested in the way that

1282
00:48:42,720 --> 00:48:47,440
objects build memory and meaning through

1283
00:48:44,880 --> 00:48:49,359
their use. Your jeans become a record of

1284
00:48:47,440 --> 00:48:51,359
all the experiences that you have while

1285
00:48:49,359 --> 00:48:53,839
wearing them. Things like where you put

1286
00:48:51,359 --> 00:48:55,440
your wallet, that weird coffee stain,

1287
00:48:53,839 --> 00:48:56,960
the tear from falling off your bike that

1288
00:48:55,440 --> 00:49:00,559
you patched up with scraps from another

1289
00:48:56,960 --> 00:49:02,240
pair, a vessel of memory and nostalgia.

1290
00:49:00,559 --> 00:49:04,400
Jeans also carry social meaning.

1291
00:49:02,240 --> 00:49:06,240
Originally a symbol of the working class

1292
00:49:04,400 --> 00:49:08,240
and then progressing into rebellion and

1293
00:49:06,240 --> 00:49:10,160
counterculture. While they do not exist

1294
00:49:08,240 --> 00:49:11,839
solely in these social domains anymore,

1295
00:49:10,160 --> 00:49:13,760
these meanings are firmly embedded into

1296
00:49:11,839 --> 00:49:15,280
their aesthetic and can imply certain

1297
00:49:13,760 --> 00:49:17,119
levels of social status and group

1298
00:49:15,280 --> 00:49:19,119
belonging. Yes, I did just reference

1299
00:49:17,119 --> 00:49:21,200
semiotic sociology and phenomenology in

1300
00:49:19,119 --> 00:49:23,280
a lightning talk. I would I know you

1301
00:49:21,200 --> 00:49:25,280
would love to hear me talk about coard

1302
00:49:23,280 --> 00:49:29,359
and marks, but sadly I don't quite have

1303
00:49:25,280 --> 00:49:30,880
time for that. So, uh uh competition

1304
00:49:29,359 --> 00:49:32,400
started two weeks ago. What's happened

1305
00:49:30,880 --> 00:49:35,040
since I started? So, I wore them to

1306
00:49:32,400 --> 00:49:37,440
PyCon. I'm giving a talk in these pants.

1307
00:49:35,040 --> 00:49:39,040
Um, I wore them when my partner was

1308
00:49:37,440 --> 00:49:40,960
taken in for a surgery that they had

1309
00:49:39,040 --> 00:49:42,800
been fighting for years to get. And I

1310
00:49:40,960 --> 00:49:44,720
was wearing them when I learned that my

1311
00:49:42,800 --> 00:49:46,800
grandfather had passed away. So, in just

1312
00:49:44,720 --> 00:49:48,559
2 weeks, these jeans have shared quite a

1313
00:49:46,800 --> 00:49:50,160
lot of experiences.

1314
00:49:48,559 --> 00:49:52,240
Hopefully, the rest of the year goes a

1315
00:49:50,160 --> 00:49:53,920
little bit smoother, especially because

1316
00:49:52,240 --> 00:49:57,440
now I kind of need to provide an update

1317
00:49:53,920 --> 00:50:00,680
at next year's PyCon.

1318
00:49:57,440 --> 00:50:00,680
Thank you.

1319
00:50:03,839 --> 00:50:09,839
Well, they say you are what you wear and

1320
00:50:06,160 --> 00:50:11,200
I think for you doubly so. Um, if

1321
00:50:09,839 --> 00:50:12,559
everything goes well and you're at next

1322
00:50:11,200 --> 00:50:15,680
year's Lightning Talks, I think you'll

1323
00:50:12,559 --> 00:50:18,160
probably still be wearing them. So, uh,

1324
00:50:15,680 --> 00:50:23,040
say hi to Ash now.

1325
00:50:18,160 --> 00:50:24,960
Um, hey, uh, Catrix, you're on deck. Um,

1326
00:50:23,040 --> 00:50:28,000
Alex is going to tell us something about

1327
00:50:24,960 --> 00:50:29,760
being done.

1328
00:50:28,000 --> 00:50:31,920
Good day, everyone. All right, quick

1329
00:50:29,760 --> 00:50:33,680
disclaimer before I start. I am not here

1330
00:50:31,920 --> 00:50:35,839
on behalf of my company. All stories

1331
00:50:33,680 --> 00:50:38,240
that I share are my own and my opinions

1332
00:50:35,839 --> 00:50:40,319
are also my own, not my companies. It's

1333
00:50:38,240 --> 00:50:42,160
a QR code on screen right now that links

1334
00:50:40,319 --> 00:50:45,200
to notes about this talk. I'll also show

1335
00:50:42,160 --> 00:50:46,960
it later. I love the motto done is

1336
00:50:45,200 --> 00:50:49,040
better than perfect. And today I'm going

1337
00:50:46,960 --> 00:50:51,359
to give an example of when that's a bad

1338
00:50:49,040 --> 00:50:54,240
idea and also some practical ways to

1339
00:50:51,359 --> 00:50:56,319
apply it when it's a good idea. My house

1340
00:50:54,240 --> 00:50:58,559
is majority solar powered, which is

1341
00:50:56,319 --> 00:51:00,720
great because no power bills, but I do

1342
00:50:58,559 --> 00:51:02,800
work in IT, which means lots of devices

1343
00:51:00,720 --> 00:51:04,800
using lots of power. And so I'm

1344
00:51:02,800 --> 00:51:06,400
constantly watching my solar system. How

1345
00:51:04,800 --> 00:51:08,079
much power is being produced? How

1346
00:51:06,400 --> 00:51:11,040
charged are the batteries? How much

1347
00:51:08,079 --> 00:51:13,680
power am I using? And so on. A while

1348
00:51:11,040 --> 00:51:15,520
back, I got sick of always going outside

1349
00:51:13,680 --> 00:51:17,920
to check on everything and decided to

1350
00:51:15,520 --> 00:51:20,960
try and get remote monitoring set up. A

1351
00:51:17,920 --> 00:51:23,200
quick Google search um later, turns out

1352
00:51:20,960 --> 00:51:25,119
my system has a built-in Wi-Fi dongle

1353
00:51:23,200 --> 00:51:26,480
and an app to view everything about your

1354
00:51:25,119 --> 00:51:28,800
system that doesn't require a

1355
00:51:26,480 --> 00:51:30,400
subscription. Once I registered an

1356
00:51:28,800 --> 00:51:32,319
account in the app, it asked me to enter

1357
00:51:30,400 --> 00:51:33,920
the serial number of my system. Then

1358
00:51:32,319 --> 00:51:36,800
immediately started showing all the

1359
00:51:33,920 --> 00:51:38,319
information I wanted. Great job done. No

1360
00:51:36,800 --> 00:51:40,720
more going out in the rain at night to

1361
00:51:38,319 --> 00:51:42,640
check on the batteries. A few weeks went

1362
00:51:40,720 --> 00:51:45,200
by and this monitoring was working

1363
00:51:42,640 --> 00:51:47,119
great. But then I started to wonder, do

1364
00:51:45,200 --> 00:51:48,800
I trust the manufacturer of my solar

1365
00:51:47,119 --> 00:51:50,240
system to have done a good job of

1366
00:51:48,800 --> 00:51:52,319
building this monitoring system that

1367
00:51:50,240 --> 00:51:53,839
they are offering for free? Surely

1368
00:51:52,319 --> 00:51:55,440
they're trying to keep the cost as low

1369
00:51:53,839 --> 00:51:58,880
as possible since they're not charging

1370
00:51:55,440 --> 00:52:00,800
for it. So naturally, I started digging.

1371
00:51:58,880 --> 00:52:02,319
I'm not going to go into depth about how

1372
00:52:00,800 --> 00:52:04,400
I discovered these issues since this

1373
00:52:02,319 --> 00:52:06,880
isn't a cyber security talk, but let me

1374
00:52:04,400 --> 00:52:08,400
summarize the main issues here. No

1375
00:52:06,880 --> 00:52:10,240
HTTPS.

1376
00:52:08,400 --> 00:52:11,839
both the connections from the system to

1377
00:52:10,240 --> 00:52:15,040
the server and from the app to the

1378
00:52:11,839 --> 00:52:17,200
server where unencrypted HTTP a perfect

1379
00:52:15,040 --> 00:52:19,920
example of done is better than perfect

1380
00:52:17,200 --> 00:52:22,079
being a bad thing.

1381
00:52:19,920 --> 00:52:24,480
No verification that you own the system

1382
00:52:22,079 --> 00:52:27,119
you are trying to control. Just put in a

1383
00:52:24,480 --> 00:52:29,040
system ID, you can control it. This

1384
00:52:27,119 --> 00:52:30,960
means you could easily disable

1385
00:52:29,040 --> 00:52:33,760
overcharge protection for someone else's

1386
00:52:30,960 --> 00:52:36,240
batteries and start a fire. This on its

1387
00:52:33,760 --> 00:52:41,119
own wouldn't be terrible if it wasn't

1388
00:52:36,240 --> 00:52:43,760
for the final issue. sequential IDs.

1389
00:52:41,119 --> 00:52:46,079
So I immediately disabled the Wi-Fi

1390
00:52:43,760 --> 00:52:47,599
dongle and started looking for other

1391
00:52:46,079 --> 00:52:50,319
options.

1392
00:52:47,599 --> 00:52:52,559
I found an amazing GitHub repository

1393
00:52:50,319 --> 00:52:54,640
where someone else had already reverse

1394
00:52:52,559 --> 00:52:57,200
engineered the USB communication

1395
00:52:54,640 --> 00:52:59,839
protocol for the system. And so my focus

1396
00:52:57,200 --> 00:53:01,839
turned to building a basic program to

1397
00:52:59,839 --> 00:53:04,079
record, store, and display all the data

1398
00:53:01,839 --> 00:53:05,839
I needed. I split this into three

1399
00:53:04,079 --> 00:53:09,119
distinct parts to make it easier to work

1400
00:53:05,839 --> 00:53:11,359
on and marginally more secure. Part one,

1401
00:53:09,119 --> 00:53:13,599
the monitor connects to the system by

1402
00:53:11,359 --> 00:53:15,599
USB, queries all the parameters I want,

1403
00:53:13,599 --> 00:53:18,559
stores them in a SQLite database with

1404
00:53:15,599 --> 00:53:20,960
the timestamp. Pretty simple. Part two,

1405
00:53:18,559 --> 00:53:23,119
solar viewer for the web serves a basic

1406
00:53:20,960 --> 00:53:25,359
web page using Flask, which renders data

1407
00:53:23,119 --> 00:53:27,280
from the same database. By this point,

1408
00:53:25,359 --> 00:53:29,200
I've accomplished part of my goal. I'm

1409
00:53:27,280 --> 00:53:30,880
recording data and I can view it on my

1410
00:53:29,200 --> 00:53:33,040
local network. But what about remote

1411
00:53:30,880 --> 00:53:34,720
monitoring? Because I've adopted the

1412
00:53:33,040 --> 00:53:36,880
done is better than perfect motto for

1413
00:53:34,720 --> 00:53:38,800
this project. I haven't built any kind

1414
00:53:36,880 --> 00:53:41,359
of authentication system. I haven't set

1415
00:53:38,800 --> 00:53:43,280
up HTTPS and I don't want to expose my

1416
00:53:41,359 --> 00:53:45,599
hack together code to the internet. So

1417
00:53:43,280 --> 00:53:46,880
what now? I want to introduce a concept

1418
00:53:45,599 --> 00:53:49,280
some of you will already be familiar

1419
00:53:46,880 --> 00:53:51,200
with. Risk outsourcing. This basically

1420
00:53:49,280 --> 00:53:53,119
just means palm off the risk to someone

1421
00:53:51,200 --> 00:53:54,800
else. So here's how I put that into

1422
00:53:53,119 --> 00:53:57,680
practice. The third part of my

1423
00:53:54,800 --> 00:53:59,440
monitoring setup, Discord. A basic

1424
00:53:57,680 --> 00:54:01,440
Discord bot that connects to the same

1425
00:53:59,440 --> 00:54:04,000
database, renders the same data as the

1426
00:54:01,440 --> 00:54:05,760
web viewer, post in a Discord server. A

1427
00:54:04,000 --> 00:54:07,599
perfect example of how you can embrace

1428
00:54:05,760 --> 00:54:10,160
done is better than perfect without

1429
00:54:07,599 --> 00:54:11,839
compromising on security by outsourcing

1430
00:54:10,160 --> 00:54:14,400
risky components to well-established

1431
00:54:11,839 --> 00:54:16,800
products. So, how can you adopt the

1432
00:54:14,400 --> 00:54:19,119
motto of done is better than perfect?

1433
00:54:16,800 --> 00:54:20,400
Just start coding. If you have an idea,

1434
00:54:19,119 --> 00:54:22,000
you don't have to plan it out well

1435
00:54:20,400 --> 00:54:24,079
enough for it to scale to thousands of

1436
00:54:22,000 --> 00:54:26,000
users. Build it in a way that works for

1437
00:54:24,079 --> 00:54:28,160
you. Then come back to it and enhance it

1438
00:54:26,000 --> 00:54:31,359
if you want to or don't. Done is better

1439
00:54:28,160 --> 00:54:33,200
than perfect. Second, outsource risk.

1440
00:54:31,359 --> 00:54:35,599
For example, many projects can use

1441
00:54:33,200 --> 00:54:37,359
Discord or a Telegram bot or many other

1442
00:54:35,599 --> 00:54:39,680
platforms to provide secure remote

1443
00:54:37,359 --> 00:54:42,160
access with very little effort or risk.

1444
00:54:39,680 --> 00:54:44,079
Third, share your projects. Sharing your

1445
00:54:42,160 --> 00:54:45,680
projects is a great side effect. Over

1446
00:54:44,079 --> 00:54:47,920
time, you will start to automatically

1447
00:54:45,680 --> 00:54:49,680
adopt good practices such as using

1448
00:54:47,920 --> 00:54:51,119
config files instead of hard coding,

1449
00:54:49,680 --> 00:54:53,200
maintain a requirements file, and

1450
00:54:51,119 --> 00:54:55,119
writing down the setup instructions. Cuz

1451
00:54:53,200 --> 00:54:57,040
believe me, it seems obvious how to set

1452
00:54:55,119 --> 00:54:59,119
it up today, but it won't be obvious two

1453
00:54:57,040 --> 00:55:01,040
years and a dozen projects later.

1454
00:54:59,119 --> 00:55:02,720
Fourth, be transparent when sharing your

1455
00:55:01,040 --> 00:55:04,400
work. Add a disclaimer if you don't

1456
00:55:02,720 --> 00:55:05,760
think it's secure or a note if you think

1457
00:55:04,400 --> 00:55:07,760
something could be improved but don't

1458
00:55:05,760 --> 00:55:09,920
have time to do it. And finally,

1459
00:55:07,760 --> 00:55:11,599
encourage contributions. Your hack

1460
00:55:09,920 --> 00:55:13,440
together weekend project probably works

1461
00:55:11,599 --> 00:55:15,280
great for you and only you and that's

1462
00:55:13,440 --> 00:55:17,040
okay. Remember done is better than

1463
00:55:15,280 --> 00:55:19,280
perfect. But maybe someone else will

1464
00:55:17,040 --> 00:55:21,119
make it work for them. And remember,

1465
00:55:19,280 --> 00:55:25,559
review their change in the same context.

1466
00:55:21,119 --> 00:55:25,559
D is better than perfect. Thank

1467
00:55:30,480 --> 00:55:33,839
There we go. Well done.

1468
00:55:32,160 --> 00:55:34,400
Did we just cut him off before he was

1469
00:55:33,839 --> 00:55:34,960
done?

1470
00:55:34,400 --> 00:55:36,559
No.

1471
00:55:34,960 --> 00:55:40,000
Oh, hey.

1472
00:55:36,559 --> 00:55:45,760
No, I was I was waiting for the zero.

1473
00:55:40,000 --> 00:55:48,640
So, thank you so much, Alex Verico. Um,

1474
00:55:45,760 --> 00:55:51,520
unfortunately up on this side is I did

1475
00:55:48,640 --> 00:55:52,720
it again. Sorry, Peter. Unfortunately,

1476
00:55:51,520 --> 00:55:54,799
up on this side is not a lot of

1477
00:55:52,720 --> 00:55:56,240
lightning talk. Fortunately, however, it

1478
00:55:54,799 --> 00:55:57,839
is our lovely conference director, Peter

1479
00:55:56,240 --> 00:56:00,079
Hall.

1480
00:55:57,839 --> 00:56:04,559
Um, but in the meantime, we've just had

1481
00:56:00,079 --> 00:56:07,599
Alex Vero and now we have another AV

1482
00:56:04,559 --> 00:56:09,119
person to talk. Please direct your

1483
00:56:07,599 --> 00:56:11,119
complaints about that joke to

1484
00:56:09,119 --> 00:56:13,200
Christopher Nogabau.

1485
00:56:11,119 --> 00:56:14,559
Classic.

1486
00:56:13,200 --> 00:56:16,480
That was literally one of the one that

1487
00:56:14,559 --> 00:56:18,240
was the one that he was the most keen on

1488
00:56:16,480 --> 00:56:19,680
out of the spreadsheet jokes. All of the

1489
00:56:18,240 --> 00:56:20,960
all the all of the jokes that I've made

1490
00:56:19,680 --> 00:56:24,000
on stage here that you've liked, those

1491
00:56:20,960 --> 00:56:28,000
are mine. Um

1492
00:56:24,000 --> 00:56:32,079
um anyway, Chris has been uh a core part

1493
00:56:28,000 --> 00:56:34,240
of our AV team since 2020. um multiple

1494
00:56:32,079 --> 00:56:36,319
conference organizers including myself

1495
00:56:34,240 --> 00:56:37,760
have uh trusted him with basically

1496
00:56:36,319 --> 00:56:41,040
everything that make this conference

1497
00:56:37,760 --> 00:56:44,319
run. And based on what I've seen from

1498
00:56:41,040 --> 00:56:46,799
the submission, I'm about to learn that

1499
00:56:44,319 --> 00:56:50,559
whether or not I should have had that

1500
00:56:46,799 --> 00:56:52,000
much trust. We'll find out.

1501
00:56:50,559 --> 00:56:53,680
Hey everyone, there is audio in this

1502
00:56:52,000 --> 00:56:56,640
just to let you guys at the back know.

1503
00:56:53,680 --> 00:56:59,040
Um, so this uh titled is titled, "Is

1504
00:56:56,640 --> 00:57:03,520
this secret data encoded into the PyCon

1505
00:56:59,040 --> 00:57:06,000
AU 2025 logo?" Let's have a look. Well,

1506
00:57:03,520 --> 00:57:07,359
at the start of each of the conferences

1507
00:57:06,000 --> 00:57:08,880
for the last few years, we've had

1508
00:57:07,359 --> 00:57:11,280
something called a bumper. And it's a

1509
00:57:08,880 --> 00:57:13,200
kind of it wakes people up. And it also

1510
00:57:11,280 --> 00:57:15,839
says, "Oh, we're about to start." So,

1511
00:57:13,200 --> 00:57:18,319
some of them we've had is this one. Now,

1512
00:57:15,839 --> 00:57:19,920
I get to I don't draw the logos, but I

1513
00:57:18,319 --> 00:57:23,040
animate them. I do the sound effects

1514
00:57:19,920 --> 00:57:25,280
that go with them. And this year we had

1515
00:57:23,040 --> 00:57:28,880
this beautiful logo by Robert uh

1516
00:57:25,280 --> 00:57:30,559
Shulton. Uh big clap for that wonderful

1517
00:57:28,880 --> 00:57:33,040
human

1518
00:57:30,559 --> 00:57:35,599
who made this great logo. And I got

1519
00:57:33,040 --> 00:57:38,559
inspired by this little bit up here. Uh

1520
00:57:35,599 --> 00:57:40,720
which to me reminded me of a 1980s tape

1521
00:57:38,559 --> 00:57:42,720
loader. What is a tape loader? Well,

1522
00:57:40,720 --> 00:57:45,280
back in the 1980s, we had computers like

1523
00:57:42,720 --> 00:57:47,680
these uh which loaded from uh tape. This

1524
00:57:45,280 --> 00:57:49,119
is a Sinclair ZedX Spectrum. Yes. You

1525
00:57:47,680 --> 00:57:51,040
put your tape in the tape machine, you

1526
00:57:49,119 --> 00:57:53,040
push play, and often while it was

1527
00:57:51,040 --> 00:57:55,440
loading, it would do things like this,

1528
00:57:53,040 --> 00:57:56,799
which has audio. There we go. So, you

1529
00:57:55,440 --> 00:57:59,040
can see there the black and white's

1530
00:57:56,799 --> 00:58:01,119
loading with audio as it goes. And you

1531
00:57:59,040 --> 00:58:02,880
hear the data. And there we go. You see

1532
00:58:01,119 --> 00:58:05,119
it in the logo as well. So, that's where

1533
00:58:02,880 --> 00:58:07,760
the inspiration for that came. But if we

1534
00:58:05,119 --> 00:58:09,280
take that audio and we put it into an

1535
00:58:07,760 --> 00:58:10,799
audio editor program, you can do this

1536
00:58:09,280 --> 00:58:12,880
from your you rip it off YouTube when

1537
00:58:10,799 --> 00:58:14,720
you get home and try this. You'll notice

1538
00:58:12,880 --> 00:58:16,480
that this weird thing happens. If you

1539
00:58:14,720 --> 00:58:18,640
look with phase analysis, the difference

1540
00:58:16,480 --> 00:58:20,319
between left and right. So at some point

1541
00:58:18,640 --> 00:58:22,480
in this, the left and right speaker

1542
00:58:20,319 --> 00:58:24,640
actually flips. So you can cancel the

1543
00:58:22,480 --> 00:58:26,240
audio out. And if we combine the two

1544
00:58:24,640 --> 00:58:28,720
together, you'll see this little block

1545
00:58:26,240 --> 00:58:31,119
of data suddenly appears. Oh, what's

1546
00:58:28,720 --> 00:58:32,640
this?

1547
00:58:31,119 --> 00:58:34,079
If you've ever used a spectrum, you'll

1548
00:58:32,640 --> 00:58:36,079
know what that is. That's the sound of a

1549
00:58:34,079 --> 00:58:37,599
tape loading. Is it? Well, let's find

1550
00:58:36,079 --> 00:58:39,440
out. So we're going to delete the

1551
00:58:37,599 --> 00:58:41,359
waveform off the start and the end of

1552
00:58:39,440 --> 00:58:44,240
this logo bit of audio here. And if we

1553
00:58:41,359 --> 00:58:45,680
look at closely, that looks awfully like

1554
00:58:44,240 --> 00:58:46,960
frequency shift king. For those of you

1555
00:58:45,680 --> 00:58:48,880
who are into radio, you'll know what

1556
00:58:46,960 --> 00:58:51,359
that is. Um, so what we're going to do

1557
00:58:48,880 --> 00:58:53,280
is save it as a creative voice file. No

1558
00:58:51,359 --> 00:58:54,799
one's used these since 1990, but that's

1559
00:58:53,280 --> 00:58:57,040
okay. We're going to use it here. So,

1560
00:58:54,799 --> 00:58:59,520
we're going to save this as logo.voc.

1561
00:58:57,040 --> 00:59:01,280
Now, that is just the audio from our uh

1562
00:58:59,520 --> 00:59:03,200
introduction there. And we're going to

1563
00:59:01,280 --> 00:59:04,720
load it into a program called DOSBox or

1564
00:59:03,200 --> 00:59:06,960
an emulator. We're going to use a little

1565
00:59:04,720 --> 00:59:09,200
program here called Taper, which is for

1566
00:59:06,960 --> 00:59:11,920
reconstructing virtual tape files from

1567
00:59:09,200 --> 00:59:14,000
audio. going to take our logo audio.

1568
00:59:11,920 --> 00:59:16,160
We're going to pull it into taper and

1569
00:59:14,000 --> 00:59:18,240
we're going to open that uh creative

1570
00:59:16,160 --> 00:59:20,640
voice file. And in particular, there's

1571
00:59:18,240 --> 00:59:22,640
this pilot detection, which is a little

1572
00:59:20,640 --> 00:59:24,000
at the start that tells the spectrum,

1573
00:59:22,640 --> 00:59:25,839
oh, are we running at the right speed?

1574
00:59:24,000 --> 00:59:27,280
And it does a whole load of calibration.

1575
00:59:25,839 --> 00:59:29,200
Turns out you can make that really

1576
00:59:27,280 --> 00:59:30,799
short, which is what we've done here.

1577
00:59:29,200 --> 00:59:32,799
And so instead of giving it 10 seconds,

1578
00:59:30,799 --> 00:59:34,880
we've given it just a half a fraction of

1579
00:59:32,799 --> 00:59:37,040
a second. You see, pyons, something's

1580
00:59:34,880 --> 00:59:39,119
happening here. What is happening here?

1581
00:59:37,040 --> 00:59:42,000
Let's find out. We're going to save this

1582
00:59:39,119 --> 00:59:44,240
as a tap file, which is a virtual tape

1583
00:59:42,000 --> 00:59:46,400
file. So, this is the virtual tape of

1584
00:59:44,240 --> 00:59:47,920
the audio from our logo that we've been

1585
00:59:46,400 --> 00:59:49,920
watching for the last three days. Every

1586
00:59:47,920 --> 00:59:51,760
time something starts, says we don't

1587
00:59:49,920 --> 00:59:53,440
fully understand this format. No, it's

1588
00:59:51,760 --> 00:59:55,119
got some weird conference audio in it,

1589
00:59:53,440 --> 00:59:56,400
too. So, it'll say, "Do you want us to

1590
00:59:55,119 --> 00:59:57,599
save it anyway?" You say, "Yeah, don't

1591
00:59:56,400 --> 00:59:59,040
worry about the other bits of audio in

1592
00:59:57,599 --> 01:00:00,960
there. Just save it and see what

1593
00:59:59,040 --> 01:00:03,599
happens." So, it's now going to save

1594
01:00:00,960 --> 01:00:05,920
that as a tap file. What do we do with a

1595
01:00:03,599 --> 01:00:08,000
tap file? Well, we load it into an

1596
01:00:05,920 --> 01:00:10,880
emulator, the free Unix Spectrum

1597
01:00:08,000 --> 01:00:12,480
emulator that we're running in Windows.

1598
01:00:10,880 --> 01:00:14,000
So, you can see here, we're going to go

1599
01:00:12,480 --> 01:00:17,040
tape. We're going to open that virtual

1600
01:00:14,000 --> 01:00:19,520
tape file. And now, with that tape file

1601
01:00:17,040 --> 01:00:21,040
open, we're going to reboot our machine.

1602
01:00:19,520 --> 01:00:22,720
So, you can see it's a hard reset.

1603
01:00:21,040 --> 01:00:24,400
Nothing in RAM here. So, this is an

1604
01:00:22,720 --> 01:00:27,119
empty spectrum. We need to look at the

1605
01:00:24,400 --> 01:00:28,799
keyboard to know to press J for load and

1606
01:00:27,119 --> 01:00:30,960
then control and P twice for the

1607
01:00:28,799 --> 01:00:33,760
inverted commas to get the two load

1608
01:00:30,960 --> 01:00:37,920
things. And then we press play on our

1609
01:00:33,760 --> 01:00:39,280
tape. And our tape goes play.

1610
01:00:37,920 --> 01:00:41,200
That's inspired. That's what inspired

1611
01:00:39,280 --> 01:00:45,119
the logo.

1612
01:00:41,200 --> 01:00:46,480
Hey, what do we found? Bit of data.

1613
01:00:45,119 --> 01:00:48,559
That's the data you've been listening to

1614
01:00:46,480 --> 01:00:51,040
for 3 days. And what do you know? We

1615
01:00:48,559 --> 01:00:52,460
press run and we have a piece of

1616
01:00:51,040 --> 01:00:59,769
software.

1617
01:00:52,460 --> 01:00:59,769
[Applause]

1618
01:01:00,480 --> 01:01:05,760
And because we love open source, if you

1619
01:01:03,599 --> 01:01:07,119
push any key at the end of this, it will

1620
01:01:05,760 --> 01:01:08,640
actually give you the source code and

1621
01:01:07,119 --> 01:01:11,040
you can start playing around with it.

1622
01:01:08,640 --> 01:01:13,280
It's written in basic. It was about 140

1623
01:01:11,040 --> 01:01:14,720
bytes worth of data. Um, there was a few

1624
01:01:13,280 --> 01:01:16,960
things with this. To get it short enough

1625
01:01:14,720 --> 01:01:18,799
to fit it in the logo was a bit of fun.

1626
01:01:16,960 --> 01:01:21,680
Um, also to make it so that it wouldn't

1627
01:01:18,799 --> 01:01:23,280
get affected by MP3 or AAC encoding.

1628
01:01:21,680 --> 01:01:24,559
I've tried encoding it with heaps of

1629
01:01:23,280 --> 01:01:26,240
different encoders. I did a whole lot of

1630
01:01:24,559 --> 01:01:28,799
filtering on the logo soundtrack to make

1631
01:01:26,240 --> 01:01:30,240
it so it didn't cut out. Um, and also

1632
01:01:28,799 --> 01:01:31,599
had to make sure the venue could play it

1633
01:01:30,240 --> 01:01:33,440
in stereo, which they couldn't. This

1634
01:01:31,599 --> 01:01:35,680
whole room is mono, but we figured out a

1635
01:01:33,440 --> 01:01:37,839
way of not having the cutout. Um, and

1636
01:01:35,680 --> 01:01:39,440
then finally, um, thank you to the

1637
01:01:37,839 --> 01:01:41,599
organizers for letting me do a lightning

1638
01:01:39,440 --> 01:01:43,520
talk. And if I'm just changing the code

1639
01:01:41,599 --> 01:01:46,079
down here, cuz we can cuz it's all open

1640
01:01:43,520 --> 01:01:49,920
and you can do this too if you load this

1641
01:01:46,079 --> 01:01:53,020
off YouTube later. I'd like to thank you

1642
01:01:49,920 --> 01:01:55,540
for letting me do a lightning talk.

1643
01:01:53,020 --> 01:01:55,760
[Applause]

1644
01:01:55,540 --> 01:01:58,839
[Music]

1645
01:01:55,760 --> 01:01:58,839
[Applause]

1646
01:01:58,880 --> 01:02:04,079
So, uh, for those of you on the live

1647
01:02:02,000 --> 01:02:07,520
stream or who buy an online ticket for

1648
01:02:04,079 --> 01:02:09,680
next year, um, please actually stay and

1649
01:02:07,520 --> 01:02:12,480
watch the talks and don't spend your

1650
01:02:09,680 --> 01:02:14,480
time trying to hack apart the audio. Uh,

1651
01:02:12,480 --> 01:02:17,680
we really do want you to watch the

1652
01:02:14,480 --> 01:02:21,079
talks. Um, that's the end of the

1653
01:02:17,680 --> 01:02:21,079
lightning talks.

1654
01:02:21,200 --> 01:02:25,839
I'm I'm supposed to like vamp now, but

1655
01:02:23,200 --> 01:02:28,960
I'm still just like struck by I'm I'm

1656
01:02:25,839 --> 01:02:29,680
impressed and also betrayed.

1657
01:02:28,960 --> 01:02:31,280
It

1658
01:02:29,680 --> 01:02:33,119
and it's one of the best Curly Boys I

1659
01:02:31,280 --> 01:02:34,799
think we've seen in the uh however many

1660
01:02:33,119 --> 01:02:37,040
years it's been since Curly Boy's been a

1661
01:02:34,799 --> 01:02:39,680
thing. So, uh thanks for that as well.

1662
01:02:37,040 --> 01:02:42,000
Um while we're doing thanks, thank you

1663
01:02:39,680 --> 01:02:44,319
to all the other AV people for doing the

1664
01:02:42,000 --> 01:02:47,720
amazing job you do do making this run

1665
01:02:44,319 --> 01:02:47,720
actually smoothly.

1666
01:02:48,799 --> 01:02:52,880
Seriously.

1667
01:02:50,559 --> 01:02:55,440
Uh

1668
01:02:52,880 --> 01:02:58,160
uh thank you. Thank thank you to Beno

1669
01:02:55,440 --> 01:03:01,200
who has been our uh stopwatch holder for

1670
01:02:58,160 --> 01:03:04,720
the last two days. Uh we normally have

1671
01:03:01,200 --> 01:03:06,640
him on Q the stopwatch holder. Uh we

1672
01:03:04,720 --> 01:03:08,559
have him on deck to uh to do

1673
01:03:06,640 --> 01:03:10,640
interpretive dances when the AV fails,

1674
01:03:08,559 --> 01:03:12,079
but the AV people did their job. So uh

1675
01:03:10,640 --> 01:03:16,559
you didn't get to see any of that this

1676
01:03:12,079 --> 01:03:19,119
time. Um genuinely when we when we book

1677
01:03:16,559 --> 01:03:20,400
venues for PyCon we tell the venue AV

1678
01:03:19,119 --> 01:03:21,680
people about lightning talks and their

1679
01:03:20,400 --> 01:03:23,920
reaction is almost always consistent

1680
01:03:21,680 --> 01:03:25,119
that you want to do what

1681
01:03:23,920 --> 01:03:27,119
and it still works.

1682
01:03:25,119 --> 01:03:29,839
Yeah it's still like massive testament

1683
01:03:27,119 --> 01:03:32,880
to the venue AV team and especially to

1684
01:03:29,839 --> 01:03:35,920
uh the PON AV team led by Nex video.

1685
01:03:32,880 --> 01:03:39,760
Oh uh thanks you thank you to all 12

1686
01:03:35,920 --> 01:03:41,359
sorry 12 13 speakers uh from today's

1687
01:03:39,760 --> 01:03:43,960
lightning talks. you've been fantastic

1688
01:03:41,359 --> 01:03:47,070
as always.

1689
01:03:43,960 --> 01:03:47,070
[Music]

1690
01:03:47,680 --> 01:03:51,359
Um, thank you Daisy for being the

1691
01:03:49,839 --> 01:03:54,760
wonderful co-host for this year's

1692
01:03:51,359 --> 01:03:54,760
Lightning Talks.

1693
01:03:55,280 --> 01:03:58,240
And thank you so much to Chris for

1694
01:03:56,960 --> 01:03:59,039
coming back and hosting Lightning Talks

1695
01:03:58,240 --> 01:04:03,280
with me once again.

1696
01:03:59,039 --> 01:04:04,880
Hey, it's me. Um, that is the end of

1697
01:04:03,280 --> 01:04:07,359
Lightning Talks. Please give a huge

1698
01:04:04,880 --> 01:04:12,760
round of applause to our fearless

1699
01:04:07,359 --> 01:04:12,760
conference director, Peter Hall. Woo!

1700
01:04:13,400 --> 01:04:22,400
[Applause]

1701
01:04:15,000 --> 01:04:22,400
[Music]

1702
01:04:22,640 --> 01:04:25,799
I'll be