1
00:00:11,920 --> 00:00:18,560
hello pike online 2021 welcome back to

2
00:00:15,839 --> 00:00:21,039
the education track and it is my

3
00:00:18,560 --> 00:00:25,760
absolute pleasure to introduce shrey

4
00:00:21,039 --> 00:00:28,000
samaya to give a talk uh on python type

5
00:00:25,760 --> 00:00:33,040
hints and how they can supercharge cs

6
00:00:28,000 --> 00:00:35,280
education um this is very exciting i uh

7
00:00:33,040 --> 00:00:38,000
i want to know this and i i want this to

8
00:00:35,280 --> 00:00:40,000
be better and i'm hoping that you stray

9
00:00:38,000 --> 00:00:42,160
will improve the quality my quality of

10
00:00:40,000 --> 00:00:45,039
life through this talk

11
00:00:42,160 --> 00:00:48,640
it's a big ask i know um

12
00:00:45,039 --> 00:00:49,440
but i i i reckon you can do it um now

13
00:00:48,640 --> 00:00:52,640
uh

14
00:00:49,440 --> 00:00:54,000
trey i have i found a fabulous bio that

15
00:00:52,640 --> 00:00:55,840
i did submit but i didn't find it to

16
00:00:54,000 --> 00:00:59,760
begin with um and he said that i could

17
00:00:55,840 --> 00:01:01,920
make up anything about him um so uh i

18
00:00:59,760 --> 00:01:04,799
would like to start his bio by saying uh

19
00:01:01,920 --> 00:01:08,400
that shrey is joining us um following a

20
00:01:04,799 --> 00:01:12,799
wildly successful uh olympic performance

21
00:01:08,400 --> 00:01:14,560
um uh i was just amazed at how how how

22
00:01:12,799 --> 00:01:15,600
true to form um

23
00:01:14,560 --> 00:01:17,520
you showed real courage and

24
00:01:15,600 --> 00:01:19,840
determination i'm i'm proud of you i'm

25
00:01:17,520 --> 00:01:21,920
proud of you stretch for me yeah um

26
00:01:19,840 --> 00:01:26,000
but also shay is a course administrator

27
00:01:21,920 --> 00:01:27,680
and tutor of comp 1511

28
00:01:26,000 --> 00:01:30,560
uh the programming fundamentals course

29
00:01:27,680 --> 00:01:33,280
uh in unsw's school of computer science

30
00:01:30,560 --> 00:01:35,040
um he's been teaching in some form for

31
00:01:33,280 --> 00:01:37,200
most of his life and started tinkering

32
00:01:35,040 --> 00:01:39,119
with python in early high school uh

33
00:01:37,200 --> 00:01:41,840
trey's also currently co-president of

34
00:01:39,119 --> 00:01:45,600
the cse soc unsw which is the resident

35
00:01:41,840 --> 00:01:47,759
computer society of unsw um

36
00:01:45,600 --> 00:01:50,320
which is one of our universities here in

37
00:01:47,759 --> 00:01:52,960
new south wales um

38
00:01:50,320 --> 00:01:55,600
and he's leading a team of about 150

39
00:01:52,960 --> 00:01:58,479
people to create produce social events

40
00:01:55,600 --> 00:02:01,600
podcasts hackathons career events videos

41
00:01:58,479 --> 00:02:03,680
workshops peer mentoring ctfs assorted

42
00:02:01,600 --> 00:02:05,439
technical projects high school outreach

43
00:02:03,680 --> 00:02:08,080
and more

44
00:02:05,439 --> 00:02:09,920
so many things uh but you do so many

45
00:02:08,080 --> 00:02:11,680
fantastic things as part of it so um

46
00:02:09,920 --> 00:02:12,640
it's a fantastic society

47
00:02:11,680 --> 00:02:15,440
um

48
00:02:12,640 --> 00:02:17,760
and otherwise uh shrey apparently you

49
00:02:15,440 --> 00:02:19,280
love cats music reading and video games

50
00:02:17,760 --> 00:02:21,200
and that seems

51
00:02:19,280 --> 00:02:22,800
great and i would say you also

52
00:02:21,200 --> 00:02:24,160
love python ah

53
00:02:22,800 --> 00:02:27,360
we could have had a cat on the stream

54
00:02:24,160 --> 00:02:28,640
yes that's quite sad ah maybe next time

55
00:02:27,360 --> 00:02:30,640
all right well

56
00:02:28,640 --> 00:02:31,680
i'm keen to hear what you think um shrey

57
00:02:30,640 --> 00:02:33,280
will be

58
00:02:31,680 --> 00:02:36,640
accepting questions assuming i haven't

59
00:02:33,280 --> 00:02:39,360
taken up all of the time in this chat um

60
00:02:36,640 --> 00:02:40,319
so throw your questions in the stream

61
00:02:39,360 --> 00:02:41,519
yard

62
00:02:40,319 --> 00:02:43,680
chat

63
00:02:41,519 --> 00:02:45,920
and i will let you get to it

64
00:02:43,680 --> 00:02:48,160
cool awesome thank you so much nikki um

65
00:02:45,920 --> 00:02:49,200
cool so hello everyone as nicki drews my

66
00:02:48,160 --> 00:02:50,800
name is shrey

67
00:02:49,200 --> 00:02:54,000
just a quick note on this talk this talk

68
00:02:50,800 --> 00:02:55,840
is going through quite a lot um as usual

69
00:02:54,000 --> 00:02:57,360
um i want to talk about python type in

70
00:02:55,840 --> 00:02:59,280
so we're going to go a bit introduction

71
00:02:57,360 --> 00:03:01,120
about what is a type what is types

72
00:02:59,280 --> 00:03:02,959
safety what other type systems what are

73
00:03:01,120 --> 00:03:04,800
they useful for um we'll then go into

74
00:03:02,959 --> 00:03:06,239
the specifics of python type hints and

75
00:03:04,800 --> 00:03:08,239
how we can look at doing those and then

76
00:03:06,239 --> 00:03:09,760
we'll jump from there to

77
00:03:08,239 --> 00:03:12,159
using them in education how we can use

78
00:03:09,760 --> 00:03:14,319
type ins to force students to design

79
00:03:12,159 --> 00:03:16,879
better think uh as a software engineer

80
00:03:14,319 --> 00:03:19,040
and actually do a full uh front to back

81
00:03:16,879 --> 00:03:20,879
sort of design of that program so let's

82
00:03:19,040 --> 00:03:22,640
jump straight into it um and i'm hoping

83
00:03:20,879 --> 00:03:24,799
that i stick on time otherwise i'm going

84
00:03:22,640 --> 00:03:26,959
to get yelled at so let's go um

85
00:03:24,799 --> 00:03:28,000
disclaimer i am not a genius i'm not

86
00:03:26,959 --> 00:03:29,920
guido

87
00:03:28,000 --> 00:03:31,120
i do not know everything i only know

88
00:03:29,920 --> 00:03:33,280
some things and i'm going to tell you

89
00:03:31,120 --> 00:03:35,440
about those some things i may be wrong

90
00:03:33,280 --> 00:03:38,159
and please please inform me if i am

91
00:03:35,440 --> 00:03:40,560
yes thank you cool so

92
00:03:38,159 --> 00:03:42,319
safety uh what is safety when we talk

93
00:03:40,560 --> 00:03:44,480
about programming languages and running

94
00:03:42,319 --> 00:03:46,000
our code what is safety um i was reading

95
00:03:44,480 --> 00:03:48,640
a stanford article that talks about

96
00:03:46,000 --> 00:03:50,319
safety as a vague term lacking great

97
00:03:48,640 --> 00:03:51,760
definition and i was like great that's

98
00:03:50,319 --> 00:03:54,000
very useful for my python talk i'm going

99
00:03:51,760 --> 00:03:55,200
to slap that right in there um but for

100
00:03:54,000 --> 00:03:57,040
the purpose of this talk we'll say

101
00:03:55,200 --> 00:03:59,599
something on the lines of safety is

102
00:03:57,040 --> 00:04:01,200
about avoiding harmful mistakes um now

103
00:03:59,599 --> 00:04:02,879
when we run our code when we run our

104
00:04:01,200 --> 00:04:05,360
python code we usually run into sort of

105
00:04:02,879 --> 00:04:07,360
two major problems uh our program

106
00:04:05,360 --> 00:04:08,879
crashes and burns it holds it stops

107
00:04:07,360 --> 00:04:12,159
there's some sort of error and says you

108
00:04:08,879 --> 00:04:14,000
can't do this or this failed or we get

109
00:04:12,159 --> 00:04:16,479
we have like a human error undefined

110
00:04:14,000 --> 00:04:18,000
behavior unexpected behavior something

111
00:04:16,479 --> 00:04:21,359
went wrong but the program still runs

112
00:04:18,000 --> 00:04:23,840
fine um and so if only there was a way

113
00:04:21,359 --> 00:04:25,919
that we could avoid these behaviors as

114
00:04:23,840 --> 00:04:27,759
much as we could uh

115
00:04:25,919 --> 00:04:29,680
psych there is and that's called type

116
00:04:27,759 --> 00:04:31,520
safety so

117
00:04:29,680 --> 00:04:35,199
a lot of type languages

118
00:04:31,520 --> 00:04:38,000
such as c rust etc can attempt to

119
00:04:35,199 --> 00:04:40,880
enforce good behavior including safety

120
00:04:38,000 --> 00:04:42,800
by performing static checks when you run

121
00:04:40,880 --> 00:04:44,240
the code so when you run the code it

122
00:04:42,800 --> 00:04:46,880
scans through and checks all right

123
00:04:44,240 --> 00:04:48,400
you've said this variable is an integer

124
00:04:46,880 --> 00:04:50,479
but you're trying to multiply it by a

125
00:04:48,400 --> 00:04:52,320
string i can't multiply an integer by a

126
00:04:50,479 --> 00:04:53,840
string i'm not going to let you run the

127
00:04:52,320 --> 00:04:55,759
code so it doesn't even get to that

128
00:04:53,840 --> 00:04:58,160
running point so this is what we call a

129
00:04:55,759 --> 00:04:59,680
statically checked language

130
00:04:58,160 --> 00:05:01,360
the process of checking the code is

131
00:04:59,680 --> 00:05:03,120
called type checking

132
00:05:01,360 --> 00:05:04,960
and the algorithm that does the checking

133
00:05:03,120 --> 00:05:07,199
is called the type checker so if you

134
00:05:04,960 --> 00:05:08,479
hear me jumble those like language later

135
00:05:07,199 --> 00:05:10,000
on the talk

136
00:05:08,479 --> 00:05:11,120
that's what we're talking about so

137
00:05:10,000 --> 00:05:13,520
essentially

138
00:05:11,120 --> 00:05:16,400
every value has a type and when you run

139
00:05:13,520 --> 00:05:18,800
the code the operations on the value are

140
00:05:16,400 --> 00:05:20,880
checked per the type now python doesn't

141
00:05:18,800 --> 00:05:22,880
do this python does something different

142
00:05:20,880 --> 00:05:25,120
python's what we call it dynamically

143
00:05:22,880 --> 00:05:26,320
oops i already talked about this yikes

144
00:05:25,120 --> 00:05:28,240
um

145
00:05:26,320 --> 00:05:29,919
python does something different python

146
00:05:28,240 --> 00:05:31,199
does what we call dynamically typelink

147
00:05:29,919 --> 00:05:34,000
it's a dynamically typed language and

148
00:05:31,199 --> 00:05:35,360
we'll go over it in a bit um but in

149
00:05:34,000 --> 00:05:37,120
python we don't have these types right

150
00:05:35,360 --> 00:05:38,479
so why not if it sounds so great if we

151
00:05:37,120 --> 00:05:40,400
can avoid these errors if we can if we

152
00:05:38,479 --> 00:05:42,320
can make our code better why don't we do

153
00:05:40,400 --> 00:05:44,639
it and the answer here is convenience

154
00:05:42,320 --> 00:05:47,440
right imagine if you had to specify for

155
00:05:44,639 --> 00:05:49,759
every single variable what type it was

156
00:05:47,440 --> 00:05:51,360
it would be so slow to write your code

157
00:05:49,759 --> 00:05:53,280
especially when you want to use python

158
00:05:51,360 --> 00:05:54,800
for quick quick tools right so a lot of

159
00:05:53,280 --> 00:05:57,280
the time i'll use python for doing some

160
00:05:54,800 --> 00:05:58,479
quick dynamic or changing formats of

161
00:05:57,280 --> 00:06:00,240
stuff i don't really want to make it

162
00:05:58,479 --> 00:06:02,560
work correctly i just want to work very

163
00:06:00,240 --> 00:06:04,400
quickly i want to write it quickly so

164
00:06:02,560 --> 00:06:06,639
we lose some safety but we get a lot of

165
00:06:04,400 --> 00:06:08,560
convenience um

166
00:06:06,639 --> 00:06:10,639
so python instead instead of having

167
00:06:08,560 --> 00:06:14,000
static typing we have what we call

168
00:06:10,639 --> 00:06:16,560
doctyping um or dynamic typing so what

169
00:06:14,000 --> 00:06:18,319
is doctyping and how is that relevant to

170
00:06:16,560 --> 00:06:21,360
what i'm about to tell you so duck

171
00:06:18,319 --> 00:06:24,400
typing is not this one duck typing is

172
00:06:21,360 --> 00:06:27,600
something like this so python it only

173
00:06:24,400 --> 00:06:29,840
checks if you're allowed to do something

174
00:06:27,600 --> 00:06:31,520
it doesn't check if it's the it doesn't

175
00:06:29,840 --> 00:06:33,440
check it at compile time it checks it

176
00:06:31,520 --> 00:06:34,479
when you're running the code so for

177
00:06:33,440 --> 00:06:36,319
example

178
00:06:34,479 --> 00:06:37,680
let's say you have a class duck

179
00:06:36,319 --> 00:06:39,199
my cursor doesn't show let's say you

180
00:06:37,680 --> 00:06:41,280
have a class duck

181
00:06:39,199 --> 00:06:42,800
this duck has a quack method when you

182
00:06:41,280 --> 00:06:44,720
call the quack method

183
00:06:42,800 --> 00:06:46,880
it will print out quack

184
00:06:44,720 --> 00:06:48,479
there's also in this code a say quack

185
00:06:46,880 --> 00:06:49,840
function which takes in a variable

186
00:06:48,479 --> 00:06:52,000
called duck

187
00:06:49,840 --> 00:06:53,039
so when takes in the variable it'll say

188
00:06:52,000 --> 00:06:55,280
uh

189
00:06:53,039 --> 00:06:57,599
of that instance of a class of that doc

190
00:06:55,280 --> 00:06:59,520
tell it to quack so if we have a new

191
00:06:57,599 --> 00:07:02,000
duck called jeff and we say all right

192
00:06:59,520 --> 00:07:03,440
jeff say quack jeff will say quack cool

193
00:07:02,000 --> 00:07:04,960
this this works fine this is as we

194
00:07:03,440 --> 00:07:06,000
expect um

195
00:07:04,960 --> 00:07:07,919
but

196
00:07:06,000 --> 00:07:10,400
what happens if we have a dog class and

197
00:07:07,919 --> 00:07:12,479
this dog class also has a quack method

198
00:07:10,400 --> 00:07:16,160
uh if we tell the dog to quack it'll

199
00:07:12,479 --> 00:07:17,599
quack and so duck typing essentially is

200
00:07:16,160 --> 00:07:20,160
if it quacks like a duck and it walks

201
00:07:17,599 --> 00:07:22,000
like a duck then it must be a dock so if

202
00:07:20,160 --> 00:07:24,000
it's able to do the thing that you ask

203
00:07:22,000 --> 00:07:26,880
it to do then that's fine that's okay

204
00:07:24,000 --> 00:07:29,120
and that's all we care about um

205
00:07:26,880 --> 00:07:30,479
so you might be wondering oh my gosh why

206
00:07:29,120 --> 00:07:31,919
are we talking about type systems and

207
00:07:30,479 --> 00:07:34,000
type theory i thought i was here a lot

208
00:07:31,919 --> 00:07:35,120
of education and yes as i mentioned

209
00:07:34,000 --> 00:07:37,280
we'll get we'll get there we'll get

210
00:07:35,120 --> 00:07:39,840
there i promise uh assuming i get kicked

211
00:07:37,280 --> 00:07:41,039
out but we'll get there um

212
00:07:39,840 --> 00:07:43,199
so

213
00:07:41,039 --> 00:07:45,039
i was going through some old hsc code

214
00:07:43,199 --> 00:07:46,400
that i wrote so uh in new south wales

215
00:07:45,039 --> 00:07:47,840
there's a system called the hsc where

216
00:07:46,400 --> 00:07:50,560
students will do a series of subjects

217
00:07:47,840 --> 00:07:53,039
towards their final grade um and as part

218
00:07:50,560 --> 00:07:55,039
of the software design hsd subject which

219
00:07:53,039 --> 00:07:57,360
i did in elective you have to write a

220
00:07:55,039 --> 00:07:59,120
massive major work so i was going

221
00:07:57,360 --> 00:08:00,720
through a lot of my code for this talk

222
00:07:59,120 --> 00:08:02,720
to see if there's anything i could use

223
00:08:00,720 --> 00:08:05,039
as an example of code that was

224
00:08:02,720 --> 00:08:07,440
questionable and hard to understand and

225
00:08:05,039 --> 00:08:09,599
i came across this function um this

226
00:08:07,440 --> 00:08:11,599
function is called get body it takes in

227
00:08:09,599 --> 00:08:15,360
a message it checks if the message is

228
00:08:11,599 --> 00:08:18,720
multi-part uh then it returns itself

229
00:08:15,360 --> 00:08:21,840
or a get payload so what is what is

230
00:08:18,720 --> 00:08:24,319
message uh is message some sort of class

231
00:08:21,840 --> 00:08:26,639
is is it a function or a type from

232
00:08:24,319 --> 00:08:28,400
somewhere else like uh what is is

233
00:08:26,639 --> 00:08:30,800
multi-part i've never seen it before at

234
00:08:28,400 --> 00:08:32,800
all um so

235
00:08:30,800 --> 00:08:34,719
looking at my code it's very difficult

236
00:08:32,800 --> 00:08:36,640
to reason what it's doing and how it

237
00:08:34,719 --> 00:08:38,080
works i don't know what messages i don't

238
00:08:36,640 --> 00:08:39,839
know what it's supposed to return there

239
00:08:38,080 --> 00:08:42,159
are so many things here that is very

240
00:08:39,839 --> 00:08:43,680
difficult to understand and extracts the

241
00:08:42,159 --> 00:08:45,040
body from the email

242
00:08:43,680 --> 00:08:46,560
doesn't help at all

243
00:08:45,040 --> 00:08:49,440
until i read this comment i had no idea

244
00:08:46,560 --> 00:08:51,680
this is about emails um so

245
00:08:49,440 --> 00:08:54,320
i wonder what that's looking at

246
00:08:51,680 --> 00:08:56,240
so if i were to call get body of hello

247
00:08:54,320 --> 00:08:58,480
uh i don't really know what will happen

248
00:08:56,240 --> 00:09:01,040
here and so oh no my error message

249
00:08:58,480 --> 00:09:01,920
didn't appear oh no okay well if i did

250
00:09:01,040 --> 00:09:04,240
this

251
00:09:01,920 --> 00:09:06,720
there's a big error saying uh you cannot

252
00:09:04,240 --> 00:09:09,120
run is multipart on hello um so my

253
00:09:06,720 --> 00:09:10,880
python code just stops and breaks

254
00:09:09,120 --> 00:09:13,600
which is not good we want this stuff to

255
00:09:10,880 --> 00:09:15,519
be checked before we run the code um so

256
00:09:13,600 --> 00:09:17,279
the way we can do this is using type ins

257
00:09:15,519 --> 00:09:19,680
um and so we can tell our function

258
00:09:17,279 --> 00:09:21,839
message is of type blah and we can say

259
00:09:19,680 --> 00:09:24,399
please only do this if this is correct

260
00:09:21,839 --> 00:09:26,080
right so if i say message is a string

261
00:09:24,399 --> 00:09:28,240
this will run fine if i say message

262
00:09:26,080 --> 00:09:30,160
should be an integer or a number and i

263
00:09:28,240 --> 00:09:32,320
say get body of hello my program will

264
00:09:30,160 --> 00:09:34,160
error saying hey you've made a mistake

265
00:09:32,320 --> 00:09:36,560
you've passed in a string where i

266
00:09:34,160 --> 00:09:38,959
expected a number um so let's go over

267
00:09:36,560 --> 00:09:41,279
how to do some of that in python

268
00:09:38,959 --> 00:09:42,720
so in python type ins are written as

269
00:09:41,279 --> 00:09:45,040
annotations

270
00:09:42,720 --> 00:09:47,040
our type ins are completely ignored when

271
00:09:45,040 --> 00:09:48,160
you run the program uh which is you

272
00:09:47,040 --> 00:09:51,120
might be thinking well what the heck's

273
00:09:48,160 --> 00:09:52,959
the point then um typings are ignored

274
00:09:51,120 --> 00:09:54,959
for quite a good reason

275
00:09:52,959 --> 00:09:56,880
which is that you shouldn't be forced to

276
00:09:54,959 --> 00:09:58,320
write python code that has types right

277
00:09:56,880 --> 00:10:00,399
you should that shouldn't be mandatory

278
00:09:58,320 --> 00:10:02,160
so there are actual tools out there

279
00:10:00,399 --> 00:10:03,040
where you run you run the tool on your

280
00:10:02,160 --> 00:10:04,640
code

281
00:10:03,040 --> 00:10:07,680
a good example is something called my pi

282
00:10:04,640 --> 00:10:09,680
which is a very common python supported

283
00:10:07,680 --> 00:10:11,519
type checking program

284
00:10:09,680 --> 00:10:13,120
and you run this over your code it will

285
00:10:11,519 --> 00:10:14,560
then look at the annotations and it will

286
00:10:13,120 --> 00:10:16,640
check that you've done the right things

287
00:10:14,560 --> 00:10:17,920
when you just run python 3 in your code

288
00:10:16,640 --> 00:10:19,760
it won't actually check your types you

289
00:10:17,920 --> 00:10:21,920
can do something completely wrong um so

290
00:10:19,760 --> 00:10:22,880
as python we start as annotations and so

291
00:10:21,920 --> 00:10:24,959
the syntax for this looks something

292
00:10:22,880 --> 00:10:27,200
along the lines of this so

293
00:10:24,959 --> 00:10:29,760
for example i have a variable called age

294
00:10:27,200 --> 00:10:33,680
age i've said is an int or an integer or

295
00:10:29,760 --> 00:10:35,920
a number um and that has a value of 20.

296
00:10:33,680 --> 00:10:38,959
so quite standard you say your variable

297
00:10:35,920 --> 00:10:42,079
name the type and then the value

298
00:10:38,959 --> 00:10:45,279
groceries groceries is a dictionary of

299
00:10:42,079 --> 00:10:46,880
uh string keys to int values

300
00:10:45,279 --> 00:10:49,200
and you can see i've initialized it with

301
00:10:46,880 --> 00:10:50,399
apples goes to five and chocolates goes

302
00:10:49,200 --> 00:10:51,760
to three

303
00:10:50,399 --> 00:10:55,040
and so there are a bunch of what we call

304
00:10:51,760 --> 00:10:57,440
basic types here um so integer bull

305
00:10:55,040 --> 00:10:58,880
float string bytes list blah blah blah

306
00:10:57,440 --> 00:11:03,040
blah

307
00:10:58,880 --> 00:11:05,360
none is also a type that we can use um

308
00:11:03,040 --> 00:11:07,360
so a quick aside

309
00:11:05,360 --> 00:11:10,959
um oh that's not the quickest i wanted

310
00:11:07,360 --> 00:11:12,560
uh here back here we have a dict

311
00:11:10,959 --> 00:11:14,399
in earlier versions of python so

312
00:11:12,560 --> 00:11:17,000
previous to 3.9

313
00:11:14,399 --> 00:11:19,200
you actually have to use typing.dick so

314
00:11:17,000 --> 00:11:21,120
typing.capitaldedict and so you often

315
00:11:19,200 --> 00:11:22,720
see sometimes that sometimes people use

316
00:11:21,120 --> 00:11:24,720
capital dedic sometimes people use

317
00:11:22,720 --> 00:11:27,680
lowercase dict it's encouraged that if

318
00:11:24,720 --> 00:11:29,519
you're using above python 3.9 you'll be

319
00:11:27,680 --> 00:11:31,040
using the lowercase d dict because i

320
00:11:29,519 --> 00:11:32,800
think it just looks a lot better like

321
00:11:31,040 --> 00:11:34,399
why would you use typing into

322
00:11:32,800 --> 00:11:36,399
typing.typing.list

323
00:11:34,399 --> 00:11:38,880
when you can just write dict list int

324
00:11:36,399 --> 00:11:40,160
right it's a lot easier to read um so

325
00:11:38,880 --> 00:11:41,839
just want to note that if you do come

326
00:11:40,160 --> 00:11:43,519
across this in the wild and you see a

327
00:11:41,839 --> 00:11:45,200
uppercase d dict

328
00:11:43,519 --> 00:11:47,920
or an uppercase l list

329
00:11:45,200 --> 00:11:49,600
that is presumably using old uh typing

330
00:11:47,920 --> 00:11:51,839
instances so

331
00:11:49,600 --> 00:11:53,760
we want to know that so

332
00:11:51,839 --> 00:11:55,440
if i was to run this code i specified

333
00:11:53,760 --> 00:11:58,079
all right agent int

334
00:11:55,440 --> 00:11:59,680
age is equal to string 20.

335
00:11:58,079 --> 00:12:01,680
this should probably error right if i

336
00:11:59,680 --> 00:12:04,800
run this with a type checker tool like

337
00:12:01,680 --> 00:12:06,800
my pi it says incompatible types you

338
00:12:04,800 --> 00:12:08,720
said age is an int but now you're saying

339
00:12:06,800 --> 00:12:10,240
it's a string and all of a sudden we can

340
00:12:08,720 --> 00:12:12,399
really see the value in this sort of

341
00:12:10,240 --> 00:12:14,240
type checking um if i make a silly

342
00:12:12,399 --> 00:12:16,240
mistake later on the line in my code and

343
00:12:14,240 --> 00:12:18,160
i make so many silly mistakes trust me i

344
00:12:16,240 --> 00:12:20,320
can statically check this before i run

345
00:12:18,160 --> 00:12:22,079
my code before i run my code in a server

346
00:12:20,320 --> 00:12:23,440
or i ship my code out to people i can

347
00:12:22,079 --> 00:12:25,440
check everything is okay and there's no

348
00:12:23,440 --> 00:12:27,360
silly mistakes rather than just running

349
00:12:25,440 --> 00:12:29,360
it or writing a billion tests now tests

350
00:12:27,360 --> 00:12:30,959
are important in their own right but um

351
00:12:29,360 --> 00:12:32,560
this gives you some sort of static

352
00:12:30,959 --> 00:12:34,959
assurance which is a common buzzword

353
00:12:32,560 --> 00:12:36,240
that you'll hear thrown around um cool

354
00:12:34,959 --> 00:12:37,839
so we've talked about variable

355
00:12:36,240 --> 00:12:40,399
annotations let's talk about function

356
00:12:37,839 --> 00:12:41,680
annotations um so function annotations

357
00:12:40,399 --> 00:12:43,920
have a similar sort of syntax where you

358
00:12:41,680 --> 00:12:45,920
have your function name and each of the

359
00:12:43,920 --> 00:12:47,920
arguments each of the arguments are

360
00:12:45,920 --> 00:12:50,639
annotated as variables so for example

361
00:12:47,920 --> 00:12:52,639
our nums variable or nums argument is a

362
00:12:50,639 --> 00:12:55,120
list of integers right this is really

363
00:12:52,639 --> 00:12:56,639
easy to read um your target is also an

364
00:12:55,120 --> 00:12:58,720
integer

365
00:12:56,639 --> 00:13:00,079
now the next part is a bit tricky

366
00:12:58,720 --> 00:13:01,600
traditionally with python functions

367
00:13:00,079 --> 00:13:03,680
you'd write uh we'll go back to our

368
00:13:01,600 --> 00:13:05,519
example over here

369
00:13:03,680 --> 00:13:07,760
over here all you need to do is write

370
00:13:05,519 --> 00:13:10,240
def your function name and in brackets

371
00:13:07,760 --> 00:13:11,200
your parameters followed by a

372
00:13:10,240 --> 00:13:14,320
colon

373
00:13:11,200 --> 00:13:14,320
when you're using typeints

374
00:13:14,800 --> 00:13:18,560
you want to write your function name

375
00:13:16,720 --> 00:13:20,959
your arguments which are all typed

376
00:13:18,560 --> 00:13:23,360
followed by an arrow to the return type

377
00:13:20,959 --> 00:13:26,240
so in this case our two sum function

378
00:13:23,360 --> 00:13:28,160
should uh find which two integers and so

379
00:13:26,240 --> 00:13:30,160
it should return a list of int so this

380
00:13:28,160 --> 00:13:32,240
is the syntax you'll see it returns with

381
00:13:30,160 --> 00:13:33,760
the arrow a list of int followed by the

382
00:13:32,240 --> 00:13:35,200
colon

383
00:13:33,760 --> 00:13:36,880
in the case where your function returns

384
00:13:35,200 --> 00:13:39,600
nothing which is often the case for some

385
00:13:36,880 --> 00:13:41,120
python programs um you can just say it

386
00:13:39,600 --> 00:13:42,000
returns none

387
00:13:41,120 --> 00:13:44,560
um

388
00:13:42,000 --> 00:13:45,920
which is a really useful pattern to have

389
00:13:44,560 --> 00:13:47,279
sometimes people will

390
00:13:45,920 --> 00:13:49,279
won't add the

391
00:13:47,279 --> 00:13:50,720
returns none part which can make type

392
00:13:49,279 --> 00:13:53,360
checkers confused some type checkers

393
00:13:50,720 --> 00:13:55,760
will uh will assume that you mean none

394
00:13:53,360 --> 00:13:58,560
some type because won't um you can also

395
00:13:55,760 --> 00:14:00,160
do parameters with multiple options so

396
00:13:58,560 --> 00:14:03,760
for example you say my argument is a

397
00:14:00,160 --> 00:14:06,160
string or it's an integer or it's none

398
00:14:03,760 --> 00:14:08,240
um now notably this syntax for the the

399
00:14:06,160 --> 00:14:10,560
union or the bar is only supported from

400
00:14:08,240 --> 00:14:12,639
python 3.10 so i did bait you a little

401
00:14:10,560 --> 00:14:15,120
bit there um but you can use what's

402
00:14:12,639 --> 00:14:17,199
called a union uh if you google python

403
00:14:15,120 --> 00:14:18,880
type union uh you can do the same sort

404
00:14:17,199 --> 00:14:20,480
of thing i just like the syntax a lot

405
00:14:18,880 --> 00:14:22,560
better it's a lot easier to read

406
00:14:20,480 --> 00:14:24,000
um cool so

407
00:14:22,560 --> 00:14:25,440
i did some more research on this

408
00:14:24,000 --> 00:14:27,920
function that we're talking back back

409
00:14:25,440 --> 00:14:30,399
about before um it turns out that my

410
00:14:27,920 --> 00:14:31,920
message is an email message um and it

411
00:14:30,399 --> 00:14:33,199
goes through the email and it checks if

412
00:14:31,920 --> 00:14:35,600
it's a if it's a message that has

413
00:14:33,199 --> 00:14:37,440
multiple parts and it will get me a list

414
00:14:35,600 --> 00:14:40,320
of all those parts so

415
00:14:37,440 --> 00:14:41,839
i added some types to this thing um

416
00:14:40,320 --> 00:14:44,639
i have to import email because these

417
00:14:41,839 --> 00:14:45,760
types are their classes so you can also

418
00:14:44,639 --> 00:14:47,360
add

419
00:14:45,760 --> 00:14:49,519
type annotations that are classes so i

420
00:14:47,360 --> 00:14:51,440
can say my message is an email message

421
00:14:49,519 --> 00:14:53,440
email message class so i have to import

422
00:14:51,440 --> 00:14:54,959
email there so i can access to that so

423
00:14:53,440 --> 00:14:57,120
now when i run get body of test it

424
00:14:54,959 --> 00:14:58,639
errors and says all right argument one

425
00:14:57,120 --> 00:15:00,720
to get body is an incompatible type it

426
00:14:58,639 --> 00:15:01,600
has string i expected an email message

427
00:15:00,720 --> 00:15:04,079
um

428
00:15:01,600 --> 00:15:06,079
please try again uh so

429
00:15:04,079 --> 00:15:07,839
this is really really useful now looking

430
00:15:06,079 --> 00:15:09,600
at this code i can clearly see all right

431
00:15:07,839 --> 00:15:11,600
i know what it's expecting and i know

432
00:15:09,600 --> 00:15:13,440
what it's uh what i what i'm gonna get

433
00:15:11,600 --> 00:15:14,880
out of it and so

434
00:15:13,440 --> 00:15:16,560
without having to run the code i can

435
00:15:14,880 --> 00:15:18,079
already do some manipulations on this

436
00:15:16,560 --> 00:15:19,920
right if i was just reading this code i

437
00:15:18,079 --> 00:15:21,839
can say all right well get body returns

438
00:15:19,920 --> 00:15:23,360
a list i want to then have another

439
00:15:21,839 --> 00:15:25,279
function that will process the body of

440
00:15:23,360 --> 00:15:26,959
emails and it will go through each item

441
00:15:25,279 --> 00:15:28,320
in the list i don't have to waste time

442
00:15:26,959 --> 00:15:30,800
trying to figure out what it's returning

443
00:15:28,320 --> 00:15:32,320
or what it's actually giving etc etc so

444
00:15:30,800 --> 00:15:34,240
you save so much time and energy and

445
00:15:32,320 --> 00:15:35,920
cognitive load you just have to read

446
00:15:34,240 --> 00:15:38,240
this sort of one line to understand how

447
00:15:35,920 --> 00:15:40,480
the function works um which is really

448
00:15:38,240 --> 00:15:41,920
super useful so

449
00:15:40,480 --> 00:15:44,880
let's hop back over to the education

450
00:15:41,920 --> 00:15:47,279
part of things um type ins allow us to

451
00:15:44,880 --> 00:15:49,360
design our program first um when i was

452
00:15:47,279 --> 00:15:51,360
talking to some hsc students who were

453
00:15:49,360 --> 00:15:52,720
doing their software major works

454
00:15:51,360 --> 00:15:55,040
i was talking a lot about designing

455
00:15:52,720 --> 00:15:56,880
their program first don't just rush into

456
00:15:55,040 --> 00:15:58,639
things and code it all up design how

457
00:15:56,880 --> 00:16:01,120
your how your program how part of your

458
00:15:58,639 --> 00:16:02,480
program talk to each other and

459
00:16:01,120 --> 00:16:04,399
they would have had a lot of trouble

460
00:16:02,480 --> 00:16:05,360
coming up with how to do that um and so

461
00:16:04,399 --> 00:16:07,040
this is how i thought of this talk

462
00:16:05,360 --> 00:16:09,199
because type ins allow us to do this

463
00:16:07,040 --> 00:16:11,040
they allow us to specify all right my

464
00:16:09,199 --> 00:16:12,480
function a is going to return a message

465
00:16:11,040 --> 00:16:14,160
that looks like this

466
00:16:12,480 --> 00:16:15,279
function b should also expect a message

467
00:16:14,160 --> 00:16:17,040
that looks like this and return a

468
00:16:15,279 --> 00:16:18,079
different type of message and so all you

469
00:16:17,040 --> 00:16:20,079
need to do is write out the function

470
00:16:18,079 --> 00:16:22,480
prototypes and then you can see how your

471
00:16:20,079 --> 00:16:24,880
program interacts so

472
00:16:22,480 --> 00:16:26,399
for example um i was doing a networks

473
00:16:24,880 --> 00:16:28,320
assignment for my university my

474
00:16:26,399 --> 00:16:31,680
university course um

475
00:16:28,320 --> 00:16:33,360
and so i wanted a way to have a user as

476
00:16:31,680 --> 00:16:35,040
a dictionary right because i didn't want

477
00:16:33,360 --> 00:16:37,519
to do classes classes are too confusing

478
00:16:35,040 --> 00:16:39,360
for me sometimes um so i just wanted a

479
00:16:37,519 --> 00:16:40,800
dictionary but i wanted the dictionary

480
00:16:39,360 --> 00:16:42,399
to have a special type i wanted to be

481
00:16:40,800 --> 00:16:44,000
able to say all right this function

482
00:16:42,399 --> 00:16:46,160
should expect a user this function is a

483
00:16:44,000 --> 00:16:47,839
return user and so there's a thing

484
00:16:46,160 --> 00:16:48,880
called a typed dictionary which you can

485
00:16:47,839 --> 00:16:51,279
use

486
00:16:48,880 --> 00:16:53,360
which will allow you to make a uh

487
00:16:51,279 --> 00:16:55,120
essentially a type which is essentially

488
00:16:53,360 --> 00:16:56,800
a dictionary and it's a bit confusing

489
00:16:55,120 --> 00:16:59,440
but essentially

490
00:16:56,800 --> 00:17:00,480
my user dictionary has a password which

491
00:16:59,440 --> 00:17:02,160
is a string

492
00:17:00,480 --> 00:17:04,160
uh in a number of inbound login attempts

493
00:17:02,160 --> 00:17:05,839
which is an integer at the last login

494
00:17:04,160 --> 00:17:07,600
which is a date or time

495
00:17:05,839 --> 00:17:09,760
the last login can also be none and you

496
00:17:07,600 --> 00:17:11,520
can see the same pattern goes further on

497
00:17:09,760 --> 00:17:13,439
so when i want to make a new user all i

498
00:17:11,520 --> 00:17:15,199
have to do is say all right well john

499
00:17:13,439 --> 00:17:17,199
john is of type user remember we have

500
00:17:15,199 --> 00:17:19,360
the columns say this is of type blah so

501
00:17:17,199 --> 00:17:21,360
this is a type user

502
00:17:19,360 --> 00:17:23,199
and this is all the values for when we

503
00:17:21,360 --> 00:17:25,120
first make our new new

504
00:17:23,199 --> 00:17:26,799
user now this is really useful now

505
00:17:25,120 --> 00:17:28,400
because when i'm designing my program i

506
00:17:26,799 --> 00:17:30,160
can say all right well i want a function

507
00:17:28,400 --> 00:17:32,080
that gets all the logged in users for my

508
00:17:30,160 --> 00:17:33,679
program so let me just write out the

509
00:17:32,080 --> 00:17:35,760
prototype for this uh what the function

510
00:17:33,679 --> 00:17:37,840
might look like uh get logged in users

511
00:17:35,760 --> 00:17:39,679
will return a list of users like how

512
00:17:37,840 --> 00:17:40,559
useful is this now we can just say all

513
00:17:39,679 --> 00:17:41,760
right we don't have to think about

514
00:17:40,559 --> 00:17:43,360
anything else we can just go through and

515
00:17:41,760 --> 00:17:44,799
design all the different things that

516
00:17:43,360 --> 00:17:45,840
we're going to need for our program so i

517
00:17:44,799 --> 00:17:47,520
actually went through and i designed a

518
00:17:45,840 --> 00:17:49,440
whole bunch of functions and a whole

519
00:17:47,520 --> 00:17:50,400
bunch of types that i might need later

520
00:17:49,440 --> 00:17:53,039
on

521
00:17:50,400 --> 00:17:54,799
and it was super super useful um and

522
00:17:53,039 --> 00:17:57,440
easier to sort of break down my program

523
00:17:54,799 --> 00:17:59,840
into multiple parts so types allow us to

524
00:17:57,440 --> 00:18:01,280
not only have guaranteed safety but they

525
00:17:59,840 --> 00:18:03,600
allow us to design our program a lot

526
00:18:01,280 --> 00:18:04,960
better and this is why i think typins

527
00:18:03,600 --> 00:18:07,520
and the future of python typing is gonna

528
00:18:04,960 --> 00:18:09,679
be super awesome for teaching we can now

529
00:18:07,520 --> 00:18:11,679
design and teach about python in terms

530
00:18:09,679 --> 00:18:12,960
of design first no longer do we have to

531
00:18:11,679 --> 00:18:14,559
say all right just run your program test

532
00:18:12,960 --> 00:18:16,720
for this test with that while that is a

533
00:18:14,559 --> 00:18:19,679
useful strategy sometimes we can now

534
00:18:16,720 --> 00:18:21,919
also say if we want to oh design out

535
00:18:19,679 --> 00:18:23,520
your code first and then go back and

536
00:18:21,919 --> 00:18:24,559
code it um

537
00:18:23,520 --> 00:18:26,320
so

538
00:18:24,559 --> 00:18:29,039
another awesome part of type hinting is

539
00:18:26,320 --> 00:18:31,200
documentation um for example back on our

540
00:18:29,039 --> 00:18:32,559
email function we can now see if i'll

541
00:18:31,200 --> 00:18:34,320
send this screenshot to someone and say

542
00:18:32,559 --> 00:18:35,679
hey do you know why this isn't working

543
00:18:34,320 --> 00:18:37,200
they can now see all right well message

544
00:18:35,679 --> 00:18:39,280
is supposed to be this and it's supposed

545
00:18:37,200 --> 00:18:41,280
to return this if they wanted to look at

546
00:18:39,280 --> 00:18:42,320
uh all right what is is multipart

547
00:18:41,280 --> 00:18:43,280
they'll just need to google the

548
00:18:42,320 --> 00:18:45,360
documentation for an

549
00:18:43,280 --> 00:18:47,039
email.message.emailmessage

550
00:18:45,360 --> 00:18:47,840
um or they could find it elsewhere in my

551
00:18:47,039 --> 00:18:49,200
code

552
00:18:47,840 --> 00:18:50,559
um and

553
00:18:49,200 --> 00:18:52,320
i've been sent screenshots of code

554
00:18:50,559 --> 00:18:54,400
before where you know students like help

555
00:18:52,320 --> 00:18:56,480
this is broken and i've sent screenshots

556
00:18:54,400 --> 00:18:57,679
of code before saying help this is

557
00:18:56,480 --> 00:18:59,440
broken

558
00:18:57,679 --> 00:19:01,760
and having these type hints allow us to

559
00:18:59,440 --> 00:19:03,679
uh communicate better with our tutors

560
00:19:01,760 --> 00:19:05,679
with our students about what the code is

561
00:19:03,679 --> 00:19:07,200
doing what is it expecting what is it

562
00:19:05,679 --> 00:19:08,640
going to return to us how is it going to

563
00:19:07,200 --> 00:19:10,080
operate

564
00:19:08,640 --> 00:19:11,840
and it also allows some sort of static

565
00:19:10,080 --> 00:19:14,640
guarantees so when your student runs a

566
00:19:11,840 --> 00:19:15,840
code um and say hey my code is broken

567
00:19:14,640 --> 00:19:17,760
you can just say hey have you tried

568
00:19:15,840 --> 00:19:20,320
running your uh type checker yet have

569
00:19:17,760 --> 00:19:22,840
you done something obviously wrong um

570
00:19:20,320 --> 00:19:25,360
which happens all the time uh

571
00:19:22,840 --> 00:19:27,280
so i guess what i want to look up next

572
00:19:25,360 --> 00:19:29,600
is what's what's next we've talked about

573
00:19:27,280 --> 00:19:30,640
where type hinting is up to today um

574
00:19:29,600 --> 00:19:32,799
we've talked about some of the stuff

575
00:19:30,640 --> 00:19:35,280
that's available in 310 with the bar

576
00:19:32,799 --> 00:19:37,280
union operator but what's coming up next

577
00:19:35,280 --> 00:19:39,200
um what's coming up next is a lot of

578
00:19:37,280 --> 00:19:40,799
cool features uh but i realized my

579
00:19:39,200 --> 00:19:42,320
slides are in order and i'm so sorry for

580
00:19:40,799 --> 00:19:44,559
that uh

581
00:19:42,320 --> 00:19:46,320
there are a bunch more python type

582
00:19:44,559 --> 00:19:47,760
hinting uh

583
00:19:46,320 --> 00:19:49,120
shenanigans that i didn't really mention

584
00:19:47,760 --> 00:19:51,840
in today's talk so we are time

585
00:19:49,120 --> 00:19:53,440
constrained um for example generic types

586
00:19:51,840 --> 00:19:56,320
if you want to say this function accepts

587
00:19:53,440 --> 00:19:58,080
a type t and returns a list of type t

588
00:19:56,320 --> 00:20:00,400
you can do something like that you can

589
00:19:58,080 --> 00:20:03,200
create your own types uh you can

590
00:20:00,400 --> 00:20:05,200
pass in functions as types so a callable

591
00:20:03,200 --> 00:20:06,720
type you can say all right this this

592
00:20:05,200 --> 00:20:09,039
function should return anything that's

593
00:20:06,720 --> 00:20:12,159
iterable anything that implements the

594
00:20:09,039 --> 00:20:14,080
iterable uh sort of subclass

595
00:20:12,159 --> 00:20:15,840
and so you can do a whole bunch of

596
00:20:14,080 --> 00:20:17,200
complex things with type bits typings

597
00:20:15,840 --> 00:20:18,960
aren't just the basic stuff that i've

598
00:20:17,200 --> 00:20:21,360
shown today they are very very very

599
00:20:18,960 --> 00:20:23,120
advanced they have a lot of room to grow

600
00:20:21,360 --> 00:20:26,240
because python was not made for this in

601
00:20:23,120 --> 00:20:28,400
mind they were added on in 2015

602
00:20:26,240 --> 00:20:29,679
but they're they're very very advanced

603
00:20:28,400 --> 00:20:31,039
if you want to get into that level so

604
00:20:29,679 --> 00:20:33,039
don't think that you know your hsc

605
00:20:31,039 --> 00:20:34,320
students or the wiz kids can't use this

606
00:20:33,039 --> 00:20:35,360
this is really really good and i think

607
00:20:34,320 --> 00:20:36,640
it's actually encouraged that they

608
00:20:35,360 --> 00:20:38,640
should use this because you get better

609
00:20:36,640 --> 00:20:40,159
documentation and awareness of how the

610
00:20:38,640 --> 00:20:42,080
code runs and how they can design their

611
00:20:40,159 --> 00:20:44,480
code um

612
00:20:42,080 --> 00:20:46,400
here's the what's next slide so

613
00:20:44,480 --> 00:20:48,559
what's next so 310 hasn't been released

614
00:20:46,400 --> 00:20:50,400
yet and it has some really cool typing

615
00:20:48,559 --> 00:20:52,159
features so for example type union we

616
00:20:50,400 --> 00:20:55,280
talk about the type union

617
00:20:52,159 --> 00:20:56,559
there are two also extra things which uh

618
00:20:55,280 --> 00:20:58,640
i'm not really going to explain too much

619
00:20:56,559 --> 00:21:01,280
but essentially recursive types so for

620
00:20:58,640 --> 00:21:03,200
example if you had a type a tree and

621
00:21:01,280 --> 00:21:05,520
that tree has a field in it which is a

622
00:21:03,200 --> 00:21:07,440
tree aren't very well done in python you

623
00:21:05,520 --> 00:21:09,679
have to do a quote around the second

624
00:21:07,440 --> 00:21:11,200
type it's a bit weird if you have a look

625
00:21:09,679 --> 00:21:14,480
at it later you'll probably run into

626
00:21:11,200 --> 00:21:16,960
this problem um but they are adding a

627
00:21:14,480 --> 00:21:17,840
feature to help that be a lot better um

628
00:21:16,960 --> 00:21:19,679
so

629
00:21:17,840 --> 00:21:20,640
a quick disclaimer before we wrap it up

630
00:21:19,679 --> 00:21:22,640
um

631
00:21:20,640 --> 00:21:25,280
my pi and a lot of python

632
00:21:22,640 --> 00:21:26,400
type checkers are still in beta and

633
00:21:25,280 --> 00:21:28,400
they've been in beta for the last couple

634
00:21:26,400 --> 00:21:30,720
of years for last five years because the

635
00:21:28,400 --> 00:21:32,240
language is constantly evolving even as

636
00:21:30,720 --> 00:21:35,120
i was running this talk i was trying to

637
00:21:32,240 --> 00:21:36,720
show off the the lowercase l list syntax

638
00:21:35,120 --> 00:21:38,720
uh when i was running my error my

639
00:21:36,720 --> 00:21:40,400
programs um and the latest version of my

640
00:21:38,720 --> 00:21:42,640
pi actually didn't let me do that i had

641
00:21:40,400 --> 00:21:45,280
to jump onto the uh the nightly branch

642
00:21:42,640 --> 00:21:46,799
to have a look at using that so just be

643
00:21:45,280 --> 00:21:48,320
in mind there there are some things that

644
00:21:46,799 --> 00:21:49,360
are impossible to do with python type

645
00:21:48,320 --> 00:21:51,360
ins

646
00:21:49,360 --> 00:21:53,440
for example a very common restriction is

647
00:21:51,360 --> 00:21:55,280
higher kind of type so if you want a a

648
00:21:53,440 --> 00:21:57,679
type that has a type

649
00:21:55,280 --> 00:21:59,280
and which you probably won't need but if

650
00:21:57,679 --> 00:22:01,840
you do want to do that that is something

651
00:21:59,280 --> 00:22:04,320
you can't do um recursive types can also

652
00:22:01,840 --> 00:22:06,320
be a pain sometimes uh and a bunch of

653
00:22:04,320 --> 00:22:08,000
super complex things for the most part

654
00:22:06,320 --> 00:22:09,039
though python types are really really

655
00:22:08,000 --> 00:22:10,320
really cool

656
00:22:09,039 --> 00:22:11,600
and i'm very excited for the future i'm

657
00:22:10,320 --> 00:22:13,600
very excited for how they can

658
00:22:11,600 --> 00:22:15,039
supercharge our education how we teach

659
00:22:13,600 --> 00:22:16,799
how we talk about design first how we

660
00:22:15,039 --> 00:22:19,520
can improve documentation how we can

661
00:22:16,799 --> 00:22:20,880
improve clarity of our code um and

662
00:22:19,520 --> 00:22:21,919
overall i think they're a really great

663
00:22:20,880 --> 00:22:24,240
thing for the language and i'm very

664
00:22:21,919 --> 00:22:25,360
excited to see where the future goes um

665
00:22:24,240 --> 00:22:27,039
but yeah thank you so much for coming

666
00:22:25,360 --> 00:22:29,360
along everyone i hope you've learned

667
00:22:27,039 --> 00:22:31,280
something today um and i'll be taking

668
00:22:29,360 --> 00:22:32,559
questions for the next few minutes if

669
00:22:31,280 --> 00:22:34,960
that's okay

670
00:22:32,559 --> 00:22:34,960
nikki

671
00:22:36,960 --> 00:22:40,480
hello

672
00:22:38,640 --> 00:22:43,440
i cannot see you thank you so much thank

673
00:22:40,480 --> 00:22:46,400
you sorry the the duck maybe isn't

674
00:22:43,440 --> 00:22:47,120
adding to the uh mystique here um

675
00:22:46,400 --> 00:22:49,200
uh

676
00:22:47,120 --> 00:22:51,360
thank you so much uh we do have a few

677
00:22:49,200 --> 00:22:54,159
questions and there is time

678
00:22:51,360 --> 00:22:57,120
uh for a few more questions to be added

679
00:22:54,159 --> 00:22:59,760
as well um now the first question is

680
00:22:57,120 --> 00:23:02,320
simply a question of a duck

681
00:22:59,760 --> 00:23:03,919
uh with a question mark um how would you

682
00:23:02,320 --> 00:23:05,280
like to respond

683
00:23:03,919 --> 00:23:11,600
uh duck

684
00:23:05,280 --> 00:23:14,159
quack excellent fantastic um great um uh

685
00:23:11,600 --> 00:23:17,360
can you make my pie figure out what

686
00:23:14,159 --> 00:23:19,679
types are instead of finding them

687
00:23:17,360 --> 00:23:21,120
yourself is our next question yes you

688
00:23:19,679 --> 00:23:22,559
definitely can so my pi has some

689
00:23:21,120 --> 00:23:24,640
functions there's a particular function

690
00:23:22,559 --> 00:23:26,480
called reveal type so if you if you

691
00:23:24,640 --> 00:23:28,080
print reveal type of a variable it'll

692
00:23:26,480 --> 00:23:29,840
print out this type this variables of

693
00:23:28,080 --> 00:23:32,320
blah blah blah which is really useful

694
00:23:29,840 --> 00:23:34,720
for big complex code bases of you know a

695
00:23:32,320 --> 00:23:36,240
couple files or even even larger

696
00:23:34,720 --> 00:23:37,679
and you can just get a variable and you

697
00:23:36,240 --> 00:23:40,400
can say all right revealed what this

698
00:23:37,679 --> 00:23:41,760
type is um so yes you definitely can and

699
00:23:40,400 --> 00:23:42,960
a lot of them are built in mind to do

700
00:23:41,760 --> 00:23:44,880
that

701
00:23:42,960 --> 00:23:48,080
fantastic excellent

702
00:23:44,880 --> 00:23:50,480
um uh another question here which

703
00:23:48,080 --> 00:23:54,720
version of python will this work for you

704
00:23:50,480 --> 00:23:57,520
took about 3.9 onwards what about 3.6

705
00:23:54,720 --> 00:23:59,919
so uh lovely chart type painting was

706
00:23:57,520 --> 00:24:01,919
first implemented in three five

707
00:23:59,919 --> 00:24:03,679
three six onwards has support for type

708
00:24:01,919 --> 00:24:05,279
hinting there are some features that

709
00:24:03,679 --> 00:24:07,360
aren't there like for example if you

710
00:24:05,279 --> 00:24:10,240
really want generic types or super

711
00:24:07,360 --> 00:24:12,480
complex stuff 373839 are your are your

712
00:24:10,240 --> 00:24:14,320
best but yes it does exist in three six

713
00:24:12,480 --> 00:24:16,799
and you can do the basic stuff you may

714
00:24:14,320 --> 00:24:18,559
have to do typing dot uppercase d dict

715
00:24:16,799 --> 00:24:20,400
instead of a lowercase d dick but that's

716
00:24:18,559 --> 00:24:22,000
just semantics everything else will work

717
00:24:20,400 --> 00:24:24,000
fine

718
00:24:22,000 --> 00:24:24,880
excellent fantastic

719
00:24:24,000 --> 00:24:26,720
um

720
00:24:24,880 --> 00:24:29,200
here's a question on

721
00:24:26,720 --> 00:24:30,240
uh why would you pipe int

722
00:24:29,200 --> 00:24:33,200
to none

723
00:24:30,240 --> 00:24:34,720
um can none be implied if there's no int

724
00:24:33,200 --> 00:24:35,520
available

725
00:24:34,720 --> 00:24:37,679
uh

726
00:24:35,520 --> 00:24:39,760
so this is a really good question

727
00:24:37,679 --> 00:24:42,559
i propose you a counter question imagine

728
00:24:39,760 --> 00:24:43,840
i pass in a variable to a function um

729
00:24:42,559 --> 00:24:46,720
and for some reason that variable's

730
00:24:43,840 --> 00:24:48,960
turned to none now if i am implying now

731
00:24:46,720 --> 00:24:51,200
that my function of you know string or

732
00:24:48,960 --> 00:24:52,400
int should also take in none the type

733
00:24:51,200 --> 00:24:54,480
checker won't error it will say all

734
00:24:52,400 --> 00:24:56,080
right none is expected but often times

735
00:24:54,480 --> 00:24:58,000
you want it to error you want to say hey

736
00:24:56,080 --> 00:25:00,159
for some reason this is actually turned

737
00:24:58,000 --> 00:25:01,919
to none whereas i'm expecting a string

738
00:25:00,159 --> 00:25:04,080
or an int so you should explicitly

739
00:25:01,919 --> 00:25:05,520
specify this should also accept none if

740
00:25:04,080 --> 00:25:06,960
that's a valid use case because

741
00:25:05,520 --> 00:25:08,720
sometimes things will turn to none and

742
00:25:06,960 --> 00:25:10,159
that's not what's supposed to happen and

743
00:25:08,720 --> 00:25:12,640
you want to yell you an error and say

744
00:25:10,159 --> 00:25:14,400
hey something's gone wrong so

745
00:25:12,640 --> 00:25:15,760
that's a good question um but yes you

746
00:25:14,400 --> 00:25:17,440
should definitely be very explicit about

747
00:25:15,760 --> 00:25:19,760
those things

748
00:25:17,440 --> 00:25:20,799
that is a great answer excellent

749
00:25:19,760 --> 00:25:23,520
um

750
00:25:20,799 --> 00:25:24,720
two more questions

751
00:25:23,520 --> 00:25:26,960
um

752
00:25:24,720 --> 00:25:30,640
uh one of my favorite things about

753
00:25:26,960 --> 00:25:32,559
typing as a dev is it enables ide magic

754
00:25:30,640 --> 00:25:33,919
for navigating around your code

755
00:25:32,559 --> 00:25:35,440
there's more there's more this question

756
00:25:33,919 --> 00:25:37,919
continues it's like an essay it's

757
00:25:35,440 --> 00:25:40,000
fantastic our ide is the editor tooling

758
00:25:37,919 --> 00:25:42,000
types sorry our ids and the editor

759
00:25:40,000 --> 00:25:44,000
tooling types uh

760
00:25:42,000 --> 00:25:46,720
are they able to enable useful

761
00:25:44,000 --> 00:25:48,640
things for teaching or is the magic a

762
00:25:46,720 --> 00:25:51,679
hindrance in understanding how the

763
00:25:48,640 --> 00:25:53,679
system works what are your thoughts so a

764
00:25:51,679 --> 00:25:56,000
lot of editors will

765
00:25:53,679 --> 00:25:57,840
analyze your types as you write the code

766
00:25:56,000 --> 00:25:59,039
and so when you call the function so

767
00:25:57,840 --> 00:26:00,640
oftentimes you write the function and

768
00:25:59,039 --> 00:26:02,480
use it later right so you write the

769
00:26:00,640 --> 00:26:04,559
function you go to call the function and

770
00:26:02,480 --> 00:26:06,000
it will actually hover in your preview

771
00:26:04,559 --> 00:26:07,360
saying this function expects two

772
00:26:06,000 --> 00:26:08,720
arguments the first argument between

773
00:26:07,360 --> 00:26:10,880
integer blur and the second argument

774
00:26:08,720 --> 00:26:13,039
should be of string blah

775
00:26:10,880 --> 00:26:14,640
and so oftentimes it won't do the magic

776
00:26:13,039 --> 00:26:16,640
of checking if you pass in a variable

777
00:26:14,640 --> 00:26:19,919
blah blah so it won't actually hinder

778
00:26:16,640 --> 00:26:21,600
you that sense um i think that editors

779
00:26:19,919 --> 00:26:23,919
can be super super useful when it comes

780
00:26:21,600 --> 00:26:26,320
to hinting because they will enable that

781
00:26:23,919 --> 00:26:28,400
magic feature and so

782
00:26:26,320 --> 00:26:29,360
as to whether it's a it's too magic

783
00:26:28,400 --> 00:26:31,600
that's a question that gets asked

784
00:26:29,360 --> 00:26:33,440
sometimes like i don't want my students

785
00:26:31,600 --> 00:26:35,279
to have this magical pop-over of this is

786
00:26:33,440 --> 00:26:37,440
what it should expect i want to fail and

787
00:26:35,279 --> 00:26:40,240
learn the hard way um which is which is

788
00:26:37,440 --> 00:26:41,919
a valid strategy that's totally okay um

789
00:26:40,240 --> 00:26:42,799
but i think you can disable those if

790
00:26:41,919 --> 00:26:45,039
that's something you're interested in

791
00:26:42,799 --> 00:26:46,320
doing um so i'm not entirely sure if

792
00:26:45,039 --> 00:26:48,000
that answers your question feel free to

793
00:26:46,320 --> 00:26:50,640
ask me more on the hallway channel

794
00:26:48,000 --> 00:26:52,400
afterwards um but yes they're they're

795
00:26:50,640 --> 00:26:54,960
super helpful five years

796
00:26:52,400 --> 00:26:57,360
i i think i have to agree that those

797
00:26:54,960 --> 00:26:58,720
that special magic can be ridiculously

798
00:26:57,360 --> 00:27:00,799
useful when

799
00:26:58,720 --> 00:27:02,640
um helping students

800
00:27:00,799 --> 00:27:04,159
gain that sense of confidence like oh

801
00:27:02,640 --> 00:27:06,159
wow i can actually do this this isn't

802
00:27:04,159 --> 00:27:08,559
just in fact exactly almost ironically

803
00:27:06,159 --> 00:27:11,360
this isn't just magic i can i can i can

804
00:27:08,559 --> 00:27:13,840
use the magic um but then it also

805
00:27:11,360 --> 00:27:14,880
um is handy to to occasionally

806
00:27:13,840 --> 00:27:18,080
um

807
00:27:14,880 --> 00:27:19,120
to feel like uh that

808
00:27:18,080 --> 00:27:20,080
that they understand they really

809
00:27:19,120 --> 00:27:21,039
understand what's going on under the

810
00:27:20,080 --> 00:27:22,000
hood yeah

811
00:27:21,039 --> 00:27:24,880
yeah so

812
00:27:22,000 --> 00:27:28,480
everything's optional which is great

813
00:27:24,880 --> 00:27:31,279
um one last question um unless there's

814
00:27:28,480 --> 00:27:37,039
another one that comes in now uh could

815
00:27:31,279 --> 00:27:39,120
we run my pie as part of the gitlab ci

816
00:27:37,039 --> 00:27:41,039
i am almost certain that is a

817
00:27:39,120 --> 00:27:43,120
possibility i would be very surprised if

818
00:27:41,039 --> 00:27:45,679
gitlab has no way to

819
00:27:43,120 --> 00:27:47,200
enable some static type checking so i'm

820
00:27:45,679 --> 00:27:48,720
going to say 95

821
00:27:47,200 --> 00:27:50,320
yes

822
00:27:48,720 --> 00:27:52,799
and leave that last five percent as an

823
00:27:50,320 --> 00:27:54,399
exercise to the homework for the viewer

824
00:27:52,799 --> 00:27:56,159
google it i'm guaranteed there's

825
00:27:54,399 --> 00:27:57,200
something there yeah

826
00:27:56,159 --> 00:28:00,960
excellent

827
00:27:57,200 --> 00:28:03,520
well those were the questions um

828
00:28:00,960 --> 00:28:04,320
that our fantastic audience have come up

829
00:28:03,520 --> 00:28:06,880
with

830
00:28:04,320 --> 00:28:09,279
um so far and we are

831
00:28:06,880 --> 00:28:12,320
pretty much at time

832
00:28:09,279 --> 00:28:16,720
so i am going to invite you

833
00:28:12,320 --> 00:28:18,080
uh dear viewer and you sray um to join

834
00:28:16,720 --> 00:28:19,440
us in the

835
00:28:18,080 --> 00:28:22,240
hallway

836
00:28:19,440 --> 00:28:24,640
in the education hallway track um now

837
00:28:22,240 --> 00:28:28,480
byo beveridges please

838
00:28:24,640 --> 00:28:30,000
um and we'll have a good old chat about

839
00:28:28,480 --> 00:28:31,039
ducks and type checking and all sorts of

840
00:28:30,000 --> 00:28:34,320
things

841
00:28:31,039 --> 00:28:35,760
before you all rejoin us here for our

842
00:28:34,320 --> 00:28:38,559
next talk

843
00:28:35,760 --> 00:28:41,440
uh by andre roberge uh entitled it

844
00:28:38,559 --> 00:28:41,440
doesn't work

845
00:28:42,320 --> 00:28:45,600
i just love it i'm not sure what it is

846
00:28:44,159 --> 00:28:48,000
but it doesn't work it definitely

847
00:28:45,600 --> 00:28:52,000
doesn't work and uh you'll have to come

848
00:28:48,000 --> 00:28:53,760
back uh in 15 minutes to find out more

849
00:28:52,000 --> 00:28:56,000
but trey thank you so very very very

850
00:28:53,760 --> 00:28:57,840
much um that talk really resonated with

851
00:28:56,000 --> 00:29:00,399
that audience and resonated with me it

852
00:28:57,840 --> 00:29:01,360
resonated with my hundreds of ducks

853
00:29:00,399 --> 00:29:02,559
and

854
00:29:01,360 --> 00:29:06,200
thank you

855
00:29:02,559 --> 00:29:06,200
thank you for having me

856
00:29:13,600 --> 00:29:15,679
you