0
00:01:56,300 --> 00:01:58,199
Hello. Hey folks.

1
00:01:58,370 --> 00:02:01,849
Happy Thanksgiving to the Americans.

2
00:02:02,319 --> 00:02:05,709
I mean, everybody can have a Happy Thanksgiving. They probably just don't care.

3
00:02:09,298 --> 00:02:10,169
So,

4
00:02:12,050 --> 00:02:17,309
so it's not gonna be a giant long nine hour opus or anything because we are having,

5
00:02:18,960 --> 00:02:23,320
oh my God, we're having eight people for dinner. So, including us. That's 10.

6
00:02:23,330 --> 00:02:25,539
So there is still plenty of prep to do,

7
00:02:25,820 --> 00:02:28,639
but I wanted to keep up the stream schedule of.

8
00:02:29,750 --> 00:02:32,089
So, speaking of the stream,

9
00:02:32,289 --> 00:02:34,889
this is Lobster's office hours

10
00:02:37,809 --> 00:02:40,070
and let's grab the browser up.

11
00:02:42,270 --> 00:02:43,130
So the first thing,

12
00:02:43,710 --> 00:02:45,830
so Lobsters is a

13
00:02:47,770 --> 00:02:50,869
social media website where people can talk to each other

14
00:02:51,830 --> 00:02:52,600
and,

15
00:02:54,949 --> 00:02:55,380
hey,

16
00:02:55,850 --> 00:02:56,559
Gaua, good morning.

17
00:03:01,419 --> 00:03:02,660
And what a funny morning.

18
00:03:02,809 --> 00:03:03,360
Um,

19
00:03:04,050 --> 00:03:04,820
so, uh

20
00:03:07,619 --> 00:03:10,679
I saw on the news, where would I just see that?

21
00:03:10,990 --> 00:03:11,839
Was it? Um

22
00:03:14,699 --> 00:03:16,039
Let's pull it off of here.

23
00:03:16,589 --> 00:03:19,440
Oh, no, it wasn't here. It was.

24
00:03:20,270 --> 00:03:22,460
Yeah, I don't have one in candy. Do I

25
00:03:23,910 --> 00:03:26,520
really prepared? Totally professional.

26
00:03:34,899 --> 00:03:36,360
So I was gonna grab the

27
00:03:38,080 --> 00:03:39,639
wild bit of news.

28
00:03:40,139 --> 00:03:41,279
This was the, uh,

29
00:03:44,210 --> 00:03:45,429
I saw a couple of links

30
00:03:46,429 --> 00:03:48,779
and uh pop up aside. This was the

31
00:03:49,490 --> 00:03:50,990
most interesting version.

32
00:03:51,649 --> 00:03:53,190
Yeah, I really like that. Um

33
00:03:54,080 --> 00:03:55,320
If you haven't seen

34
00:03:56,139 --> 00:03:58,080
HCR News. Um,

35
00:03:58,240 --> 00:03:59,970
I don't know how I found it

36
00:04:01,839 --> 00:04:03,259
but it's basically

37
00:04:05,839 --> 00:04:08,039
a filtered version of

38
00:04:08,179 --> 00:04:09,940
what we would call slash newest.

39
00:04:12,580 --> 00:04:13,279
And

40
00:04:17,209 --> 00:04:18,309
I don't know, it's, um,

41
00:04:19,558 --> 00:04:24,859
I really like that. It has the horizontal line for, here's where you last read.

42
00:04:26,459 --> 00:04:31,459
Although it is possible for things to pop in under it, it's uncommon

43
00:04:31,690 --> 00:04:33,630
and they pop in under it because,

44
00:04:33,859 --> 00:04:37,480
you know, you view at 10 a.m. and then at 10:05 a.m.

45
00:04:38,000 --> 00:04:43,040
they make it to the front page and then you come back at 11. And so the line will be

46
00:04:43,709 --> 00:04:44,359
or wait,

47
00:04:46,290 --> 00:04:47,829
maybe I'm saying this backwards.

48
00:04:51,700 --> 00:04:53,160
So after your view,

49
00:04:55,850 --> 00:04:59,040
uh anyways, the the order is based on when they're posted,

50
00:04:59,049 --> 00:05:00,890
not when they made it to the front page.

51
00:05:00,899 --> 00:05:01,630
And so

52
00:05:02,609 --> 00:05:05,989
it is occasionally possible for things to be under the line

53
00:05:06,160 --> 00:05:08,019
and new, but it's pretty rare

54
00:05:09,140 --> 00:05:09,799
and

55
00:05:12,010 --> 00:05:16,320
the uh I go on about that horizontal line because here on stream,

56
00:05:16,329 --> 00:05:18,579
a couple of months ago, I added the

57
00:05:19,010 --> 00:05:21,260
horizontal last red line to

58
00:05:21,390 --> 00:05:22,779
slash newest

59
00:05:23,019 --> 00:05:25,059
and that was pretty straight based on

60
00:05:26,000 --> 00:05:29,970
this front page for hacker news. I like it a bunch. It um

61
00:05:33,579 --> 00:05:34,920
It also just

62
00:05:35,690 --> 00:05:39,019
keeps me from refreshing the front page of hacker news.

63
00:05:39,029 --> 00:05:41,829
I'm aware I nominally run a competitor,

64
00:05:42,109 --> 00:05:44,320
maybe it's a nominal competitor.

65
00:05:44,589 --> 00:05:46,760
There's not a lot of actual competition in social

66
00:05:46,769 --> 00:05:49,559
media because everybody could be active on multiple sites.

67
00:05:49,570 --> 00:05:49,970
But

68
00:05:50,119 --> 00:05:51,709
people kind of assume that there is

69
00:05:53,429 --> 00:05:54,040
uh

70
00:05:57,380 --> 00:06:02,540
also it was a, an actual competition. Lobsters would be losing very, very badly

71
00:06:02,850 --> 00:06:04,059
or something like

72
00:06:04,200 --> 00:06:07,040
1% or less on, on any metric.

73
00:06:07,049 --> 00:06:10,600
You care to count like traffic or number of stories or anything.

74
00:06:11,869 --> 00:06:12,450
Um,

75
00:06:13,510 --> 00:06:14,489
but I enjoy reading

76
00:06:14,959 --> 00:06:16,959
hacker news and occasionally post there

77
00:06:17,720 --> 00:06:18,660
and, uh,

78
00:06:20,109 --> 00:06:21,359
I like the uh

79
00:06:22,170 --> 00:06:26,700
the slash newest kind of approach because it keeps me from just rereading the

80
00:06:26,709 --> 00:06:30,269
homepage or digging back into stories to see more comments and more comments.

81
00:06:32,500 --> 00:06:34,239
It's also why I mostly read Lobsters by

82
00:06:34,459 --> 00:06:37,350
Newest and Active and the mod dashboard.

83
00:06:38,589 --> 00:06:39,559
So yeah, g so

84
00:06:39,700 --> 00:06:42,450
this other, this other link, um

85
00:06:44,429 --> 00:06:48,329
I pulled up hacker news at the start there because that was where I first saw it.

86
00:06:48,609 --> 00:06:50,670
And then I found this link that was a

87
00:06:50,679 --> 00:06:53,959
little bit better and the most valuable thing that

88
00:06:54,940 --> 00:06:56,670
man, I am not the,

89
00:06:56,799 --> 00:06:58,279
the king of journalism,

90
00:06:58,649 --> 00:07:02,079
I mean, not that I ever was, but I have actually worked in journalism.

91
00:07:02,679 --> 00:07:06,440
And so my pet peeve is if you are going to talk about a law

92
00:07:07,070 --> 00:07:08,790
or talk about a court case,

93
00:07:09,670 --> 00:07:12,420
you have to link to it for the love of Pete.

94
00:07:14,489 --> 00:07:15,170
So

95
00:07:15,559 --> 00:07:16,290
this one,

96
00:07:18,190 --> 00:07:19,010
the uh

97
00:07:19,579 --> 00:07:20,420
uh where did it go?

98
00:07:21,880 --> 00:07:25,679
That was kind of bookkeeping stuff. So I'm real,

99
00:07:25,940 --> 00:07:29,910
I'm real cautious of playing lawyer, especially cross country

100
00:07:30,700 --> 00:07:31,760
because

101
00:07:32,160 --> 00:07:34,720
so much stuff that just looks like regular words

102
00:07:34,730 --> 00:07:38,839
turns out to be legal jargon and has special meaning

103
00:07:39,910 --> 00:07:40,690
and

104
00:07:44,559 --> 00:07:46,209
so many laws

105
00:07:46,880 --> 00:07:50,660
and contracts assume some basic knowledge of

106
00:07:52,380 --> 00:07:54,220
some legal precept

107
00:07:54,510 --> 00:07:56,160
that I'm oblivious to.

108
00:07:56,459 --> 00:08:00,559
And so it's really easy to be like, well, why don't you ask or to miss

109
00:08:01,380 --> 00:08:02,709
that? Something is implied.

110
00:08:02,720 --> 00:08:04,399
I don't have a great example off the top of my

111
00:08:04,410 --> 00:08:07,070
head because I'm talking about things I don't know about.

112
00:08:07,079 --> 00:08:07,459
But

113
00:08:07,730 --> 00:08:09,369
I've seen it enough times that,

114
00:08:10,230 --> 00:08:12,559
you know, I've asked a lawyer, can we include this clause?

115
00:08:12,570 --> 00:08:15,130
And they're like, no, that's illegal under the so,

116
00:08:15,329 --> 00:08:16,890
so active

117
00:08:17,250 --> 00:08:19,079
whoever and I'm like, well,

118
00:08:19,369 --> 00:08:20,570
really and they're like, yes.

119
00:08:20,579 --> 00:08:23,730
And then also it was litigated in Smith versus Ferguson. And

120
00:08:24,880 --> 00:08:26,489
you know, you just don't know this stuff.

121
00:08:27,429 --> 00:08:28,170
But the,

122
00:08:30,250 --> 00:08:33,140
the gist of this one is pretty short

123
00:08:35,669 --> 00:08:36,467
and

124
00:08:37,590 --> 00:08:38,250
all

125
00:08:38,380 --> 00:08:39,630
commentators,

126
00:08:40,080 --> 00:08:47,450
whether they are legal or journalist or otherwise seem to all be saying wtf at this.

127
00:08:49,979 --> 00:08:53,609
So I'm not super freaked out because it, there's a like

128
00:08:54,039 --> 00:08:55,520
we are ordering our ourselves.

129
00:08:55,530 --> 00:08:58,989
What I believe is ourselves the government to come up with guidelines

130
00:08:59,090 --> 00:09:01,590
for reasonable steps to prevent

131
00:09:02,729 --> 00:09:07,070
when it says age restricted, it means 16 and under or I'm sorry, it means under 16

132
00:09:07,719 --> 00:09:11,289
from having accounts with age restricted social media platforms.

133
00:09:11,299 --> 00:09:14,989
That is a, a new term this bill makes up and says

134
00:09:16,830 --> 00:09:19,380
it means anything that this bill applies to

135
00:09:20,609 --> 00:09:23,109
and that's the, the meat of it

136
00:09:23,630 --> 00:09:27,109
is right here and this part is pretty small.

137
00:09:29,140 --> 00:09:29,859
So

138
00:09:30,380 --> 00:09:34,030
the term it invents it age restricted social media platform.

139
00:09:35,309 --> 00:09:38,080
And if the sole purpose or a significant purpose of the

140
00:09:38,090 --> 00:09:42,580
service is to enable social interaction between two or more users.

141
00:09:42,590 --> 00:09:44,599
That that's certainly true of lobsters.

142
00:09:45,250 --> 00:09:47,799
The service allows end users, which

143
00:09:51,260 --> 00:09:54,159
why are they called end users instead of just users?

144
00:09:54,890 --> 00:09:55,469
I don't know,

145
00:09:57,020 --> 00:10:01,460
allows users to link to or interact with some of the other end users.

146
00:10:02,989 --> 00:10:04,969
I mean, that's a like yes,

147
00:10:05,210 --> 00:10:07,049
social interaction

148
00:10:07,530 --> 00:10:10,729
is interacting. So this one's kind of redundant but yes,

149
00:10:11,710 --> 00:10:15,080
the service allows users to post material on the service.

150
00:10:15,619 --> 00:10:18,809
Well, I mean, yes, you can leave stories and comments and

151
00:10:19,559 --> 00:10:21,859
even direct messages on lobster. So yes,

152
00:10:22,770 --> 00:10:26,119
such conditions if any are set out in the legislative rules,

153
00:10:26,830 --> 00:10:31,169
I I don't know what this means. Maybe this is a, we're gonna punt it to a uh

154
00:10:34,090 --> 00:10:34,950
what's it called?

155
00:10:35,140 --> 00:10:35,840
Um

156
00:10:36,299 --> 00:10:38,349
the government office for

157
00:10:38,849 --> 00:10:41,159
making every forum into a nightclub,

158
00:10:41,929 --> 00:10:46,030
whatever that bit about, you know, set up guidelines was earlier. So I don't know,

159
00:10:46,830 --> 00:10:48,409
I I am admitting right up front.

160
00:10:48,419 --> 00:10:51,679
I don't know what part of this means and this could be a pretty big escape hatch.

161
00:10:52,840 --> 00:10:56,320
But this is so weird of we're just gonna assume that

162
00:10:56,619 --> 00:10:59,140
social media is bad for people under 16.

163
00:11:02,979 --> 00:11:03,909
And uh

164
00:11:05,299 --> 00:11:06,119
where is it?

165
00:11:06,559 --> 00:11:07,869
Oh, yeah, there's a bunch of,

166
00:11:14,330 --> 00:11:16,909
so it's in legal jargon here

167
00:11:17,280 --> 00:11:19,390
of 500 penalty units,

168
00:11:20,960 --> 00:11:23,260
which sounds very, very video gamey.

169
00:11:23,489 --> 00:11:23,969
But

170
00:11:25,500 --> 00:11:26,979
the important bit is

171
00:11:27,919 --> 00:11:34,320
that translates to a Australian $50 million fine, which is $32 million.

172
00:11:34,330 --> 00:11:35,179
United States

173
00:11:36,099 --> 00:11:40,919
if they don't take reasonable steps to stop people under 16 from having accounts.

174
00:11:42,969 --> 00:11:43,390
That's

175
00:11:44,729 --> 00:11:46,510
so the penalty unit thing,

176
00:11:48,159 --> 00:11:50,809
my guess would be like, it sounds a little silly,

177
00:11:50,820 --> 00:11:53,750
but my guess is it's a way of standardizing

178
00:11:55,039 --> 00:11:59,710
penalties across legislation. Like this is just an abstraction. This is

179
00:11:59,950 --> 00:11:59,960
a,

180
00:12:00,530 --> 00:12:02,429
the legal equivalent of a variable

181
00:12:03,020 --> 00:12:04,750
I've thought for a long time that

182
00:12:05,479 --> 00:12:07,530
the law looks a lot like code

183
00:12:08,299 --> 00:12:09,880
if code had,

184
00:12:11,979 --> 00:12:15,159
if you could argue with the compiler because it was just a guy

185
00:12:16,570 --> 00:12:21,390
and if you had basically no tooling and 400 years of legacy code to deal with.

186
00:12:25,349 --> 00:12:26,080
So

187
00:12:28,260 --> 00:12:29,229
this is like

188
00:12:30,530 --> 00:12:32,789
from civil penalties import

189
00:12:34,030 --> 00:12:34,750
unit.

190
00:12:39,190 --> 00:12:39,380
Yes,

191
00:12:46,719 --> 00:12:48,210
I suppose that'll um

192
00:12:48,500 --> 00:12:49,179
I don't know.

193
00:12:49,549 --> 00:12:50,020
I'm

194
00:12:51,109 --> 00:12:52,280
maybe there's a uh

195
00:12:53,799 --> 00:12:55,429
maybe they shouldn't have written the whole penalty

196
00:12:55,440 --> 00:12:57,039
unit thing and there's a Yagi thing,

197
00:12:57,049 --> 00:12:57,340
but

198
00:12:57,520 --> 00:13:00,539
maybe it'll be really, really helpful if they ever have to uh

199
00:13:00,830 --> 00:13:02,919
replace their currency entirely.

200
00:13:04,280 --> 00:13:07,440
They're gonna drop Australian dollars and switch to the yen or something,

201
00:13:09,039 --> 00:13:12,210
then all they have to do is update that one define

202
00:13:12,219 --> 00:13:15,219
and I'm sure none of their legacy code will break.

203
00:13:19,070 --> 00:13:20,289
So this one's not

204
00:13:20,659 --> 00:13:22,090
super urgent.

205
00:13:23,020 --> 00:13:24,020
There's um

206
00:13:24,890 --> 00:13:28,780
like a year before it takes effect and there's this stuff about

207
00:13:30,580 --> 00:13:31,440
where the

208
00:13:31,679 --> 00:13:34,000
link to the PDF go and put that in the notes.

209
00:13:37,340 --> 00:13:40,320
There's that stuff about we're gonna pass it off to the other

210
00:13:41,609 --> 00:13:46,090
government department for research and implementation and guidelines and stuff.

211
00:13:46,099 --> 00:13:47,549
So, we'll see what this looks like.

212
00:13:48,270 --> 00:13:52,169
But it's just so weird. I had to mention it.

213
00:13:53,270 --> 00:13:55,530
I'm aware that this is politics and I try and stay

214
00:13:55,539 --> 00:13:59,729
out of politics but this one very directly affects the site.

215
00:14:03,250 --> 00:14:04,510
The whole thing where, uh,

216
00:14:08,320 --> 00:14:09,619
some random government

217
00:14:10,780 --> 00:14:11,799
decides that

218
00:14:12,400 --> 00:14:17,239
it's going to make a law that applies to every website in every country is,

219
00:14:19,940 --> 00:14:21,559
it's not scaling well,

220
00:14:25,090 --> 00:14:28,179
and it's not that Australia. I mean, Australia is the size of

221
00:14:28,580 --> 00:14:30,700
population wise, is the size of

222
00:14:30,950 --> 00:14:32,460
a medium US state.

223
00:14:33,830 --> 00:14:35,210
So this isn't, you know,

224
00:14:37,049 --> 00:14:39,280
a small Pacific Atoll or something.

225
00:14:44,520 --> 00:14:46,340
Ah, the whole thing is very strange.

226
00:14:47,440 --> 00:14:50,969
I would like to not ban Australia. I would miss Australians.

227
00:14:55,289 --> 00:14:56,549
So, all that aside,

228
00:14:58,989 --> 00:14:59,010
I

229
00:14:59,140 --> 00:15:01,969
wanted to do, uh, some pull request review.

230
00:15:02,849 --> 00:15:05,150
Oh, you know, let's swap that around,

231
00:15:05,460 --> 00:15:06,830
do a little out of order.

232
00:15:07,419 --> 00:15:10,750
I can do the other, uh, small bit, which is, this is Thanksgiving

233
00:15:12,690 --> 00:15:14,849
the, uh US holiday with,

234
00:15:15,650 --> 00:15:19,030
I mean, every US holiday everywhere has a complicated history.

235
00:15:19,950 --> 00:15:22,489
So we're gonna skip the meaning of the holiday

236
00:15:22,880 --> 00:15:25,750
and talk about the practical meaning of the holiday which is,

237
00:15:25,909 --> 00:15:27,390
have nice food with folks.

238
00:15:27,669 --> 00:15:30,489
Try and remember to, uh, enjoy things in your life.

239
00:15:31,760 --> 00:15:32,580
Don't ask

240
00:15:33,820 --> 00:15:37,080
lots of historical questions because you probably won't like the answers.

241
00:15:38,859 --> 00:15:39,239
Hm.

242
00:15:42,239 --> 00:15:45,260
So hopping back here to code,

243
00:15:47,539 --> 00:15:51,479
there was no movement on this one since the last stream. That's OK.

244
00:15:52,919 --> 00:15:53,809
And then

245
00:15:53,960 --> 00:15:56,669
action mailbox, you might remember

246
00:15:57,099 --> 00:15:58,289
a week ago.

247
00:15:58,500 --> 00:16:02,739
I left a review for this one. So I wanted to check in on this because

248
00:16:03,900 --> 00:16:06,390
GM had a minute to

249
00:16:07,169 --> 00:16:08,830
look at my couple of comments.

250
00:16:09,530 --> 00:16:13,309
I have glanced, but I didn't actually read

251
00:16:13,969 --> 00:16:15,349
his replies.

252
00:16:15,919 --> 00:16:17,270
So there was the one,

253
00:16:20,109 --> 00:16:23,409
ok. It says outdated. So presumably he incorporated that change

254
00:16:26,340 --> 00:16:29,469
and then I would suppose he rebased because

255
00:16:30,429 --> 00:16:32,289
there is a bunch of stuff here.

256
00:16:32,630 --> 00:16:34,530
Yeah, force pushed. Sure.

257
00:16:36,010 --> 00:16:37,309
So let's give it a quick scan.

258
00:16:37,320 --> 00:16:40,669
The other thing, I don't see the comment but it must have been rebased out

259
00:16:41,080 --> 00:16:45,070
was there was a to do comment left in the middle of the mailbox code.

260
00:16:45,640 --> 00:16:47,789
And if that is gone or explained,

261
00:16:48,260 --> 00:16:49,510
I will go ahead and

262
00:16:54,250 --> 00:16:58,950
yeah. So the to do was right here and I guess there wasn't anything extra.

263
00:17:00,809 --> 00:17:04,479
So we figure out if it's a comment, we figure out if the parent is a comment,

264
00:17:07,708 --> 00:17:09,769
if stuff is broken,

265
00:17:10,939 --> 00:17:13,800
we'll just kinda drop it in the bounced queue.

266
00:17:16,338 --> 00:17:17,739
We have to look at the

267
00:17:18,368 --> 00:17:19,118
oh man,

268
00:17:20,578 --> 00:17:22,198
years ago I

269
00:17:24,698 --> 00:17:26,279
made a

270
00:17:27,650 --> 00:17:30,800
ruby project to process mailing list archives

271
00:17:30,810 --> 00:17:32,510
because occasionally I like to read them.

272
00:17:32,949 --> 00:17:34,770
And so it would take a directory,

273
00:17:35,109 --> 00:17:36,349
well, take

274
00:17:36,699 --> 00:17:38,689
email in a bunch of forms and then try

275
00:17:38,699 --> 00:17:41,469
and figure out the threading which was fairly complicated.

276
00:17:41,800 --> 00:17:44,109
Yes, I've seen JW Z's algorithm.

277
00:17:44,670 --> 00:17:47,069
It is imperfect but a great start.

278
00:17:48,189 --> 00:17:49,010
And then

279
00:17:51,150 --> 00:17:52,930
every email client in the world.

280
00:17:53,130 --> 00:17:54,829
It's funny. Email is one of those

281
00:17:56,300 --> 00:17:57,579
long live standards.

282
00:17:57,589 --> 00:18:00,140
Where as soon as you open up the hood on something where you're like,

283
00:18:00,150 --> 00:18:01,469
this should be reasonable

284
00:18:02,310 --> 00:18:05,959
if you click reply in an email, how do I know which email you're replying to?

285
00:18:05,969 --> 00:18:08,390
And then you just immediately fall into

286
00:18:09,290 --> 00:18:13,260
40 years of legacy compatibility. Actually, it's longer with email.

287
00:18:13,270 --> 00:18:14,459
It's gotta be about 50 now.

288
00:18:18,819 --> 00:18:20,319
Asynchronous slack.

289
00:18:21,550 --> 00:18:22,329
Oh man,

290
00:18:25,609 --> 00:18:28,229
Slack is a Slack is one of those apps.

291
00:18:28,239 --> 00:18:31,469
I find a different bug in every single time I use it.

292
00:18:32,650 --> 00:18:34,000
They dragged me up the wall.

293
00:18:34,520 --> 00:18:35,109
Um

294
00:18:37,520 --> 00:18:38,020
It uh

295
00:18:39,319 --> 00:18:42,859
they keep putting all of their energy into weird collaboration

296
00:18:42,869 --> 00:18:46,349
features and then the actual but it's a chat room.

297
00:18:46,359 --> 00:18:48,010
I should be able to chat with people,

298
00:18:48,930 --> 00:18:51,390
the chatting features just don't work

299
00:18:52,349 --> 00:18:53,439
and her buggy

300
00:18:55,010 --> 00:18:55,560
and like

301
00:18:56,689 --> 00:18:58,479
I can just watch the UI

302
00:18:58,819 --> 00:19:00,420
slide around the app

303
00:19:01,260 --> 00:19:04,199
taking 200 milliseconds at a time to

304
00:19:04,560 --> 00:19:08,560
like update three or four times as I change from one channel to the other.

305
00:19:09,420 --> 00:19:10,079
It's just,

306
00:19:11,339 --> 00:19:15,199
it is the weirdest case of management prioritization where

307
00:19:16,619 --> 00:19:21,589
I I expect that they just keep wanting headline features to close deals instead of

308
00:19:22,800 --> 00:19:25,089
making the core thing work

309
00:19:30,489 --> 00:19:32,849
provided you use it in the browser. No, I can,

310
00:19:34,060 --> 00:19:35,849
I've seen so many browser bugs.

311
00:19:39,180 --> 00:19:40,079
Their app is

312
00:19:42,180 --> 00:19:42,469
Janer.

313
00:19:43,510 --> 00:19:44,810
I agree the app is Janer

314
00:19:44,949 --> 00:19:46,410
than the browser version but

315
00:19:48,189 --> 00:19:48,560
uh

316
00:20:00,510 --> 00:20:02,380
that's a totally reasonable couple of

317
00:20:03,079 --> 00:20:04,369
regular expressions

318
00:20:07,689 --> 00:20:11,099
looking at these. I just reread that classic. Um

319
00:20:14,489 --> 00:20:17,390
Actually, that would be a good one. If you have the day off,

320
00:20:22,010 --> 00:20:23,510
here's a classic,

321
00:20:23,819 --> 00:20:28,280
classic five year old, but still just such an excellent blog post

322
00:20:28,449 --> 00:20:29,260
about

323
00:20:30,260 --> 00:20:31,530
structure and code.

324
00:20:32,030 --> 00:20:34,109
It's probably a little more complicated than

325
00:20:35,229 --> 00:20:37,569
you could read while listening to me burble along.

326
00:20:37,579 --> 00:20:39,910
So just mute me if you wanna read it right now.

327
00:20:40,250 --> 00:20:42,000
But otherwise save it for the,

328
00:20:42,750 --> 00:20:44,170
it's a great holiday read.

329
00:20:46,410 --> 00:20:47,420
I put it on the,

330
00:20:48,180 --> 00:20:49,719
let's bring the show notes back up.

331
00:20:58,300 --> 00:20:59,670
She just um

332
00:21:00,119 --> 00:21:04,119
really nails with some clear examples. Um

333
00:21:05,439 --> 00:21:06,369
a great

334
00:21:06,979 --> 00:21:10,500
idea for how to think about code. And it's one of those where

335
00:21:12,479 --> 00:21:14,869
as I've interacted with junior developers a

336
00:21:15,219 --> 00:21:16,430
bunch of times

337
00:21:16,770 --> 00:21:20,239
they say, oh well, you know, I'm writing X and in my head,

338
00:21:20,790 --> 00:21:23,680
I can see like eight ways to write it and then I can't

339
00:21:23,689 --> 00:21:27,339
explain why I like one over the other and working with junior developers,

340
00:21:27,349 --> 00:21:31,310
which I started doing more of maybe 15 years ago, forced me to

341
00:21:33,099 --> 00:21:34,640
introspect a lot more

342
00:21:35,089 --> 00:21:38,890
and think about what it is that I like about a design versus another design.

343
00:21:38,900 --> 00:21:40,359
And how do I explain that?

344
00:21:40,819 --> 00:21:41,430
And

345
00:21:42,869 --> 00:21:45,849
explaining that forced me to get more rigorous about it and

346
00:21:46,050 --> 00:21:48,430
clear about the values even to myself.

347
00:21:48,439 --> 00:21:48,630
So,

348
00:21:48,640 --> 00:21:52,329
working with junior developers really made me a better programmer and this led to

349
00:21:53,430 --> 00:21:56,260
mostly that led to the Rails con talk I gave in.

350
00:21:56,390 --> 00:21:57,640
What was that? 2015?

351
00:21:58,959 --> 00:21:59,619
Um

352
00:22:00,520 --> 00:22:01,650
And this uh

353
00:22:02,099 --> 00:22:03,979
Lexi's Post here is

354
00:22:06,319 --> 00:22:08,339
exactly in that same vein. Of

355
00:22:08,810 --> 00:22:12,000
let's think through explicitly

356
00:22:12,199 --> 00:22:14,939
what's going on with this intuitive design decision.

357
00:22:15,640 --> 00:22:18,260
So I, I just love that post. It's a favorite.

358
00:22:21,270 --> 00:22:22,329
All right. So this is,

359
00:22:22,770 --> 00:22:26,189
this is the wild regular expression feature that I suggested. So

360
00:22:27,349 --> 00:22:28,130
that's in,

361
00:22:29,420 --> 00:22:30,349
that's fine.

362
00:22:33,599 --> 00:22:36,699
I occasionally get tempted to uncommon action text. And

363
00:22:38,130 --> 00:22:39,420
if we weren't a uh

364
00:22:40,719 --> 00:22:43,310
website that is aimed at programmers who are,

365
00:22:43,540 --> 00:22:44,780
I feel like mark down is

366
00:22:45,609 --> 00:22:48,859
ubiquitous enough that pretty much every programmer is comfortable with it.

367
00:22:50,729 --> 00:22:52,380
But action text is a little tempting,

368
00:22:54,790 --> 00:22:55,540
right? So

369
00:22:56,109 --> 00:22:56,939
there's that

370
00:22:57,739 --> 00:22:58,709
there's this

371
00:23:05,270 --> 00:23:06,810
active storage attachments.

372
00:23:13,160 --> 00:23:13,849
Oh,

373
00:23:14,119 --> 00:23:16,199
there's a topic um

374
00:23:18,660 --> 00:23:20,060
didn't put it on the um

375
00:23:20,239 --> 00:23:21,839
the stream topic list.

376
00:23:22,530 --> 00:23:26,189
So I'll think about this. I'll put it down for after

377
00:23:26,660 --> 00:23:27,339
um

378
00:23:29,989 --> 00:23:31,910
might be worth talking about Rail's credentials.

379
00:23:33,250 --> 00:23:36,959
So I see a few viewers have picked up real quiet here on a holiday. But

380
00:23:37,219 --> 00:23:41,420
since this is lobster's office hours, if you have any questions about the

381
00:23:42,020 --> 00:23:46,829
site or the community or the code base, feel free to

382
00:23:47,439 --> 00:23:48,790
ask any kind of questions.

383
00:23:50,079 --> 00:23:51,579
And even if you wanna,

384
00:23:53,270 --> 00:23:55,849
I should probably put it in the channel description, right? Like

385
00:23:56,209 --> 00:23:57,369
the queries thing of

386
00:23:58,050 --> 00:24:00,469
if you want to run queries against a production database,

387
00:24:01,959 --> 00:24:02,900
we can even do that.

388
00:24:05,199 --> 00:24:08,189
All right, then these tests are all

389
00:24:08,680 --> 00:24:11,349
the same. This is great. This is ready to go.

390
00:24:11,579 --> 00:24:12,709
There's that get help thing.

391
00:24:46,050 --> 00:24:47,150
This is actually

392
00:24:50,839 --> 00:24:53,939
thinking about this. This is safe to merge down, right?

393
00:24:54,359 --> 00:24:56,380
Because until

394
00:24:57,119 --> 00:24:58,060
I

395
00:24:59,459 --> 00:25:02,910
go in the answerable repo and change over the

396
00:25:06,849 --> 00:25:08,469
G sorry, any chance you're here

397
00:25:08,890 --> 00:25:09,709
or I'm sorry.

398
00:25:09,849 --> 00:25:10,780
Uh GM

399
00:25:11,459 --> 00:25:12,959
so many G's, I actually

400
00:25:13,510 --> 00:25:15,930
have a hard time remembering the difference between G

401
00:25:16,050 --> 00:25:16,119
so

402
00:25:16,290 --> 00:25:18,109
GM and G Wozniak.

403
00:25:22,030 --> 00:25:22,569
All right.

404
00:25:23,079 --> 00:25:26,890
Yeah, I can just merge this down because until I wire up

405
00:25:27,619 --> 00:25:30,319
post fix to deliver emails into it,

406
00:25:30,640 --> 00:25:32,270
it's just gonna sit here and hang out

407
00:25:33,209 --> 00:25:34,030
and then

408
00:25:34,420 --> 00:25:35,979
also

409
00:25:36,300 --> 00:25:39,390
I can test it from the command line myself.

410
00:25:40,189 --> 00:25:41,050
Bye.

411
00:25:42,839 --> 00:25:44,709
Putting an email in the post fix.

412
00:25:45,430 --> 00:25:45,439
Uh

413
00:25:46,400 --> 00:25:48,130
No, go a step higher.

414
00:25:49,439 --> 00:25:51,579
Yeah. So I contested in Prague before

415
00:25:53,069 --> 00:25:55,349
and I don't think there's anything wrong with this poor quest.

416
00:25:55,359 --> 00:25:57,930
It's just before I change over, I would really like to

417
00:25:58,239 --> 00:26:02,369
see it work because there's always, there's always those prod plumbing issues

418
00:26:03,329 --> 00:26:04,270
that was the uh

419
00:26:05,060 --> 00:26:08,920
the work from a previous stream I deployed that I just remembered a minute ago was

420
00:26:09,930 --> 00:26:12,469
I occasionally joke about how some of these raise the

421
00:26:12,479 --> 00:26:14,520
hair on the back of my neck or feel like they're

422
00:26:14,630 --> 00:26:16,469
more likely to break in prod

423
00:26:17,229 --> 00:26:17,949
and

424
00:26:18,189 --> 00:26:19,739
had a funny experience where

425
00:26:20,050 --> 00:26:22,930
the one I thought was gonna break prod was totally fine.

426
00:26:23,219 --> 00:26:24,849
And then the other one did

427
00:26:25,050 --> 00:26:26,119
break prod a bit.

428
00:26:27,449 --> 00:26:28,140
That's ok.

429
00:26:51,699 --> 00:26:52,959
Um Not very

430
00:26:53,270 --> 00:26:54,119
good but

431
00:27:02,140 --> 00:27:02,920
great.

432
00:27:04,680 --> 00:27:05,760
This is so nice

433
00:27:06,209 --> 00:27:07,339
to have this um

434
00:27:09,939 --> 00:27:11,150
rails plumbing done.

435
00:27:12,079 --> 00:27:14,609
It's funny, a big theme of

436
00:27:20,040 --> 00:27:24,609
a big theme of these recent

437
00:27:29,119 --> 00:27:31,119
pull request has been

438
00:27:31,989 --> 00:27:34,579
leaning more and more into rails features

439
00:27:34,890 --> 00:27:37,800
because Rails has picked up features for

440
00:27:38,930 --> 00:27:41,270
things that we had to do custom

441
00:27:43,979 --> 00:27:44,680
and

442
00:27:49,979 --> 00:27:52,040
as much as I've been pushing hard for it,

443
00:27:52,050 --> 00:27:58,119
the one place I really am increasingly tempted to lean away from it is Flex.

444
00:28:02,170 --> 00:28:03,709
Flex is a

445
00:28:03,939 --> 00:28:09,229
and it's a beta. The 2.0 release candidate just came out a couple of days ago.

446
00:28:09,589 --> 00:28:11,550
I know I've mentioned this on stream before.

447
00:28:14,800 --> 00:28:15,380
It's,

448
00:28:18,930 --> 00:28:19,790
let's see.

449
00:28:20,760 --> 00:28:22,589
How did I put it in that other?

450
00:28:25,300 --> 00:28:27,780
So it is a, a templating library

451
00:28:28,500 --> 00:28:29,319
or

452
00:28:33,550 --> 00:28:36,560
Ruby that allows you to

453
00:28:39,599 --> 00:28:40,459
just sort of

454
00:28:41,500 --> 00:28:43,599
uh it would replace er B

455
00:28:43,819 --> 00:28:47,859
and er B is a big pile of strings and not

456
00:28:48,180 --> 00:28:49,119
in a great way.

457
00:28:50,109 --> 00:28:52,640
So I said it over on, on blue sky that

458
00:28:53,969 --> 00:28:56,510
there was this thread I really liked. Let's

459
00:28:56,670 --> 00:28:59,660
just, I'll grab this, I'll share this so folks can click around.

460
00:29:00,150 --> 00:29:04,430
Uh Joel Draper here is the primary author of Flax

461
00:29:05,209 --> 00:29:05,869
and

462
00:29:08,099 --> 00:29:11,400
I said it's just Ruby all the right ways where

463
00:29:11,699 --> 00:29:12,520
it has

464
00:29:13,119 --> 00:29:16,280
less mental overhead because there's less stuff happening.

465
00:29:16,290 --> 00:29:20,180
You're not like context switching between strings in multiple languages.

466
00:29:20,699 --> 00:29:24,599
It's composable because it's Ruby, it's reusable because it's Ruby

467
00:29:25,030 --> 00:29:27,719
the tooling. Like instead of

468
00:29:29,099 --> 00:29:31,130
we get no linting on our

469
00:29:31,459 --> 00:29:35,439
b, I've been hesitant to add it because it's really complicated and

470
00:29:36,599 --> 00:29:38,489
the value is maybe not there.

471
00:29:39,310 --> 00:29:40,160
And then

472
00:29:40,689 --> 00:29:45,010
uh Joel also had a talk about safety where because

473
00:29:46,660 --> 00:29:53,160
Flex really does not work by slapping strings together in the way er B does.

474
00:29:57,010 --> 00:29:59,920
Flex allows a lot

475
00:30:00,239 --> 00:30:02,869
better safety around

476
00:30:03,119 --> 00:30:04,310
cross site scripting.

477
00:30:06,949 --> 00:30:07,619
So

478
00:30:08,189 --> 00:30:09,930
let me jump in. Um

479
00:30:11,900 --> 00:30:12,479
come here,

480
00:30:13,760 --> 00:30:14,849
where is my

481
00:30:16,890 --> 00:30:18,319
always hate this floating

482
00:30:20,060 --> 00:30:22,010
headers and footers that cover the page.

483
00:30:22,959 --> 00:30:23,560
So

484
00:30:23,930 --> 00:30:27,500
Flex has no dependencies. It's 2100 lines of standard ruby

485
00:30:28,709 --> 00:30:31,150
EB is not 2150 lines

486
00:30:31,530 --> 00:30:32,449
and it is not

487
00:30:34,199 --> 00:30:36,010
small and straightforward.

488
00:30:36,020 --> 00:30:39,849
I've gone into the air B code base a couple of times and it's, it's not bad.

489
00:30:39,859 --> 00:30:41,079
I'm not trying to insult it

490
00:30:41,699 --> 00:30:43,689
but it is complicated. It is

491
00:30:44,520 --> 00:30:45,969
a little bit of a climb

492
00:30:46,150 --> 00:30:49,030
and I've read a little bit of the flex code and it's just,

493
00:30:50,510 --> 00:30:52,410
there's very, very little magic.

494
00:30:53,550 --> 00:30:54,300
And

495
00:30:54,569 --> 00:30:56,300
so this person

496
00:30:56,459 --> 00:30:58,959
came along and had a couple of questions around

497
00:31:02,989 --> 00:31:07,880
what things look like. So they wrote themselves a helper called Simple Table

498
00:31:09,209 --> 00:31:12,680
or maybe this is from some gem I've never heard of called Simple Table.

499
00:31:12,689 --> 00:31:13,550
It doesn't matter

500
00:31:14,329 --> 00:31:17,030
where they passed in some records and then

501
00:31:17,900 --> 00:31:19,359
I guess they have columns.

502
00:31:19,540 --> 00:31:23,359
This is the list of columns and then it's gonna iterate over the records and print,

503
00:31:24,020 --> 00:31:28,890
you know, for each record call, record dot employee or record dot unit price

504
00:31:29,229 --> 00:31:30,719
and then build up a table

505
00:31:31,150 --> 00:31:32,030
which,

506
00:31:32,500 --> 00:31:33,560
ok, like

507
00:31:34,119 --> 00:31:35,670
this is fine,

508
00:31:37,689 --> 00:31:38,619
but it's sort of

509
00:31:39,369 --> 00:31:44,290
like, is this really er B at this point? No, this is pretty much just straight ruby.

510
00:31:44,300 --> 00:31:47,579
The only er B feature that's getting used here is

511
00:31:49,849 --> 00:31:52,239
we're gonna call these functions and slap them together.

512
00:31:52,250 --> 00:31:54,979
You will be almost might not as well be used here,

513
00:31:54,989 --> 00:31:58,150
but you have to have this special syntax before and after.

514
00:32:03,609 --> 00:32:04,859
Uh yeah, it's

515
00:32:05,069 --> 00:32:05,079
a

516
00:32:06,890 --> 00:32:07,670
partial.

517
00:32:08,260 --> 00:32:08,949
So this is

518
00:32:09,709 --> 00:32:10,359
he says you

519
00:32:10,500 --> 00:32:12,250
be, but he means

520
00:32:12,969 --> 00:32:15,780
action view here because this whole thing with

521
00:32:16,459 --> 00:32:19,109
how render builds on this stuff is

522
00:32:23,229 --> 00:32:24,760
happening in action view, not

523
00:32:25,000 --> 00:32:25,410
B

524
00:32:26,280 --> 00:32:26,939
and so

525
00:32:27,739 --> 00:32:30,420
Joel kind of ported his thing to,

526
00:32:30,949 --> 00:32:31,280
all right.

527
00:32:31,290 --> 00:32:34,569
Well, like if you just made a simple table component in flex,

528
00:32:34,579 --> 00:32:35,949
this is what it would look like.

529
00:32:37,500 --> 00:32:39,280
And it's basically the same code

530
00:32:42,640 --> 00:32:45,670
where it's like, OK, well, we have all these records, we're gonna loop them,

531
00:32:45,680 --> 00:32:47,099
we're gonna call the block for each.

532
00:32:48,160 --> 00:32:50,459
You can imagine what this is gonna look like. He doesn't

533
00:32:50,900 --> 00:32:51,739
give the

534
00:32:52,589 --> 00:32:53,729
components. But

535
00:32:55,050 --> 00:32:59,750
once you've read the 101 for Flex, you can kind of write them in your head

536
00:33:02,670 --> 00:33:03,609
and he gave

537
00:33:05,010 --> 00:33:06,170
maybe a slightly

538
00:33:06,750 --> 00:33:08,390
better example where

539
00:33:11,650 --> 00:33:14,420
this is what the Flex would look like and this is what the

540
00:33:14,530 --> 00:33:15,560
B would look like.

541
00:33:15,910 --> 00:33:18,449
And I think that he likes it in the

542
00:33:18,709 --> 00:33:19,349
B,

543
00:33:20,359 --> 00:33:24,540
you're just directly writing the html. It's kind of nice to be direct.

544
00:33:25,030 --> 00:33:26,900
But then again, you have to know

545
00:33:27,150 --> 00:33:29,939
this syntax and you have to know this syntax

546
00:33:30,439 --> 00:33:31,810
and all of this is string

547
00:33:31,949 --> 00:33:34,880
munging. And so like right here, either

548
00:33:35,180 --> 00:33:36,369
action view

549
00:33:37,430 --> 00:33:38,000
and

550
00:33:40,369 --> 00:33:41,550
Brakeman

551
00:33:41,760 --> 00:33:43,369
protect you from

552
00:33:45,219 --> 00:33:46,560
CS S injection

553
00:33:48,229 --> 00:33:49,300
or they don't

554
00:33:50,079 --> 00:33:52,550
and it's not at all visible and they have

555
00:33:52,560 --> 00:33:55,140
to know the context and you're kind of guessing that

556
00:33:56,010 --> 00:33:58,989
cause this percent equals, you could use it here

557
00:33:59,849 --> 00:34:00,300
in

558
00:34:00,540 --> 00:34:03,810
html text or you could use it here in like, you know,

559
00:34:03,819 --> 00:34:06,280
tr class equals column dot whatever.

560
00:34:09,040 --> 00:34:09,580
And so

561
00:34:10,629 --> 00:34:12,668
Joel did this really clever

562
00:34:13,688 --> 00:34:17,309
comparison where on the left here is the flex version of,

563
00:34:17,688 --> 00:34:19,618
well, if you kinda uncross your eyes,

564
00:34:19,628 --> 00:34:23,018
what actual feature is happening on these lines and you can see that

565
00:34:23,830 --> 00:34:25,570
these method calls

566
00:34:26,139 --> 00:34:27,770
and the IVAR look up

567
00:34:28,399 --> 00:34:29,280
and the block,

568
00:34:29,989 --> 00:34:31,050
these match

569
00:34:31,610 --> 00:34:32,610
that list of

570
00:34:33,270 --> 00:34:35,580
T head T RT D at

571
00:34:36,469 --> 00:34:37,250
column

572
00:34:38,458 --> 00:34:39,989
dot each with a block,

573
00:34:40,850 --> 00:34:41,350
right?

574
00:34:41,478 --> 00:34:43,750
You can see the, the shape of it matches

575
00:34:45,418 --> 00:34:48,197
and he says, well, there's one language, it's Ruby,

576
00:34:48,358 --> 00:34:50,158
you don't change to another language.

577
00:34:50,768 --> 00:34:51,829
It's a little bit of

578
00:34:51,958 --> 00:34:56,079
fib, you gotta know that you're output html, but that's, that's OK. That's fine.

579
00:34:57,419 --> 00:35:00,600
I have our look up, you do one of those, you have some method calls

580
00:35:01,300 --> 00:35:06,159
so you're not using a lot of language features. This is a smaller vocabulary.

581
00:35:07,010 --> 00:35:11,790
And then on the right here's the er B version that Gregory gave where

582
00:35:12,310 --> 00:35:14,290
there's the raw string, the raw string.

583
00:35:14,919 --> 00:35:15,550
Then

584
00:35:16,360 --> 00:35:19,280
after this syntax, you're changing into

585
00:35:19,810 --> 00:35:24,739
this er B context and you have to know that er B can handle

586
00:35:27,000 --> 00:35:28,580
what's effectively

587
00:35:29,800 --> 00:35:32,419
a string of Ruby that's gonna get

588
00:35:32,780 --> 00:35:33,100
veiled.

589
00:35:33,870 --> 00:35:34,570
And then

590
00:35:34,850 --> 00:35:36,370
this end down here,

591
00:35:38,530 --> 00:35:41,189
you know, if you've never looked at the er B source,

592
00:35:41,199 --> 00:35:45,070
it's worth looking at because something very clever is happening

593
00:35:45,620 --> 00:35:47,139
to make it possible

594
00:35:47,320 --> 00:35:47,850
to

595
00:35:48,659 --> 00:35:49,070
like

596
00:35:49,290 --> 00:35:50,159
email code

597
00:35:50,879 --> 00:35:53,969
with a raw string in the middle and run this all together.

598
00:35:55,899 --> 00:35:56,719
If you haven't,

599
00:35:57,379 --> 00:35:59,439
I it's, it's really worth a read.

600
00:35:59,449 --> 00:36:01,149
I don't wanna deep dive it now cause I'm

601
00:36:01,159 --> 00:36:03,199
trying to keep this stream short cause I've got

602
00:36:03,399 --> 00:36:03,840
like

603
00:36:04,850 --> 00:36:07,629
put some buns in the oven and that kind of stuff, but

604
00:36:10,370 --> 00:36:13,989
just an enormous amount goes on under the hood to make that work.

605
00:36:14,000 --> 00:36:18,340
And it's weird to look at where you're switching languages

606
00:36:19,100 --> 00:36:21,120
and I think Joel correctly

607
00:36:21,659 --> 00:36:23,000
highlights it here.

608
00:36:24,620 --> 00:36:28,459
Uh Well, you're working in two languages and I think you were before but

609
00:36:28,659 --> 00:36:30,639
you change in and out of the languages a bunch

610
00:36:30,649 --> 00:36:33,179
of times you use a whole bunch of different elements,

611
00:36:33,189 --> 00:36:35,179
you got different syntax, you've got

612
00:36:35,649 --> 00:36:39,879
the same amount of eye bars and lookups, but you've got all this extra stuff on top.

613
00:36:42,129 --> 00:36:43,429
So as much as I'm

614
00:36:45,850 --> 00:36:46,209
Yas

615
00:36:46,399 --> 00:36:47,739
finding the uh

616
00:36:50,840 --> 00:36:54,320
app more and more to be closer and closer to just

617
00:36:54,739 --> 00:36:56,060
stock rails,

618
00:36:56,810 --> 00:36:58,780
I keep getting tempted by flex.

619
00:37:08,750 --> 00:37:11,979
One of the things that's really impressive about flex is

620
00:37:12,739 --> 00:37:14,350
and I've, I've played with this a bit.

621
00:37:14,780 --> 00:37:16,270
It's really easy

622
00:37:16,850 --> 00:37:17,320
to,

623
00:37:17,679 --> 00:37:19,800
you don't have to do a big bang migration.

624
00:37:20,100 --> 00:37:24,870
You can do a partial migration like one component at a time.

625
00:37:26,899 --> 00:37:30,399
So that is the, the Siren song that keeps singing to me

626
00:37:31,199 --> 00:37:32,820
and pulling me away from

627
00:37:33,729 --> 00:37:38,909
rails aside from the general annoyance with rails. And that one guy who

628
00:37:41,639 --> 00:37:42,870
keeps wanting attention,

629
00:37:44,739 --> 00:37:46,399
not gonna rant about him on stream.

630
00:37:48,959 --> 00:37:49,770
So,

631
00:37:51,459 --> 00:37:55,320
got two pull requests from a first time contributor,

632
00:37:56,229 --> 00:37:57,479
Ashwin 47.

633
00:38:00,330 --> 00:38:01,679
Oh, what a super for.

634
00:38:02,850 --> 00:38:04,989
Turn your Discord into an open form.

635
00:38:05,570 --> 00:38:06,389
Oh, interesting.

636
00:38:11,669 --> 00:38:13,280
And I guess it's something that

637
00:38:14,000 --> 00:38:15,800
he is actively working on.

638
00:38:18,030 --> 00:38:18,899
That's pretty clever.

639
00:38:21,419 --> 00:38:24,360
I suppose he likes forums if he's over here contributing.

640
00:38:26,689 --> 00:38:28,469
So I picked today as pull requests

641
00:38:30,570 --> 00:38:31,260
and

642
00:38:34,820 --> 00:38:37,159
this first one is,

643
00:38:39,800 --> 00:38:40,610
it's searching

644
00:38:42,229 --> 00:38:43,110
where

645
00:38:44,419 --> 00:38:46,169
it's kind of a parsing hassle.

646
00:38:47,300 --> 00:38:48,379
But I've been trying to,

647
00:38:48,389 --> 00:38:53,639
to throw in different shapes of feature requests and bug fixes to the

648
00:38:53,949 --> 00:38:55,080
issue tracker

649
00:38:58,209 --> 00:38:59,310
cause a lot of stuff,

650
00:39:01,310 --> 00:39:05,840
a lot of stuff just hangs out here and people don't pick it up. But the things that are

651
00:39:06,780 --> 00:39:10,030
kind of plumbing shaped or have interesting

652
00:39:10,219 --> 00:39:13,320
CS puzzles seem people seem to pick up and like them.

653
00:39:13,840 --> 00:39:17,459
So I'm trying to post more of those. So this one was,

654
00:39:17,739 --> 00:39:19,989
it would be neat if when you search for stuff,

655
00:39:20,219 --> 00:39:22,520
it had mark tags around it to highlight it.

656
00:39:23,729 --> 00:39:24,310
But

657
00:39:26,060 --> 00:39:27,949
there is one big hassle

658
00:39:28,199 --> 00:39:29,909
and I wrote about it here. A corner case

659
00:39:31,110 --> 00:39:32,760
if you search for

660
00:39:33,639 --> 00:39:35,979
not a single term but a quoted term

661
00:39:38,600 --> 00:39:41,620
that ignores punctuation that are in the original markdown,

662
00:39:41,750 --> 00:39:42,580
let alone

663
00:39:43,689 --> 00:39:45,189
output in the html.

664
00:39:48,449 --> 00:39:51,010
So if you search for a phrase that

665
00:39:51,449 --> 00:39:53,149
is split across

666
00:39:55,100 --> 00:39:56,040
elements,

667
00:39:56,610 --> 00:39:59,080
it would be really nice to still highlight that.

668
00:39:59,540 --> 00:40:01,340
And so the search engine can't

669
00:40:01,479 --> 00:40:02,580
just be

670
00:40:04,629 --> 00:40:06,530
find the text and the page and

671
00:40:06,830 --> 00:40:07,570
highlight it.

672
00:40:14,989 --> 00:40:15,899
However,

673
00:40:16,760 --> 00:40:18,250
Ashwin left a comment that

674
00:40:19,020 --> 00:40:20,600
they missed that corner case.

675
00:40:21,129 --> 00:40:21,860
And so they,

676
00:40:22,439 --> 00:40:24,320
he did do the

677
00:40:24,770 --> 00:40:26,320
kind of naive version of it.

678
00:40:34,080 --> 00:40:36,159
So, you know, painting the bike shed

679
00:40:38,090 --> 00:40:41,909
a less nuclear color would be nice. That's fine. That's not a big deal.

680
00:40:43,580 --> 00:40:45,969
I like to be soft in the corners on it though.

681
00:40:50,760 --> 00:40:54,000
So basically he takes the search terms

682
00:40:54,639 --> 00:40:56,870
and he makes a regular expression

683
00:40:57,219 --> 00:40:58,120
out of them

684
00:41:08,439 --> 00:41:09,199
and then

685
00:41:10,050 --> 00:41:12,550
join stuff together

686
00:41:12,979 --> 00:41:15,070
so that he gets one regular expression,

687
00:41:15,959 --> 00:41:20,689
which all right, this is, this is pretty reasonable. I'm

688
00:41:22,699 --> 00:41:22,709
a

689
00:41:22,860 --> 00:41:27,189
little suspicious of this and would wanna see a test or two that

690
00:41:27,939 --> 00:41:29,600
punctuation would work.

691
00:41:32,129 --> 00:41:34,179
It really encodes a

692
00:41:35,330 --> 00:41:36,919
assumption about terms,

693
00:41:38,989 --> 00:41:43,199
I suppose, Redex dot escape handles it. But then there was something else that

694
00:41:43,939 --> 00:41:44,590
didn't.

695
00:41:45,969 --> 00:41:49,419
So we added this highlight search terms helper.

696
00:41:55,530 --> 00:41:55,939
Mm

697
00:41:58,879 --> 00:42:01,459
And then there's a bunch of stuff to thread it through

698
00:42:02,790 --> 00:42:03,800
the uh

699
00:42:04,209 --> 00:42:05,820
the various functions

700
00:42:06,459 --> 00:42:08,010
or the various templates.

701
00:42:13,169 --> 00:42:16,820
And it really, it caught my eye that he calls strip operators. Here.

702
00:42:17,719 --> 00:42:17,739
I

703
00:42:17,840 --> 00:42:19,610
gotta go look at the search model.

704
00:42:21,310 --> 00:42:21,929
So

705
00:42:25,979 --> 00:42:26,000
I

706
00:42:26,110 --> 00:42:27,229
started the darn

707
00:42:28,500 --> 00:42:30,959
them from the wrong buffer. Come here.

708
00:42:32,020 --> 00:42:35,610
I'm not gonna chase that around all stream. That's my uh my blog directory

709
00:42:35,770 --> 00:42:37,560
mean to start from the

710
00:42:39,600 --> 00:42:40,810
lobster source code.

711
00:42:50,590 --> 00:42:52,290
So we have a

712
00:42:52,770 --> 00:42:57,689
search parser that kind of parses an A ST out of the search to find operators like

713
00:42:58,050 --> 00:42:59,129
searching in the title

714
00:43:02,979 --> 00:43:06,439
and it splits things out into like terms and quoted

715
00:43:06,659 --> 00:43:09,479
which are the the parts of the search for, how to search them.

716
00:43:10,389 --> 00:43:10,929
And then

717
00:43:11,080 --> 00:43:14,969
strip operators is my least favorite part of the code.

718
00:43:14,979 --> 00:43:17,050
So I was really surprised to see it mentioned

719
00:43:17,830 --> 00:43:18,810
because

720
00:43:25,610 --> 00:43:26,709
this is

721
00:43:28,239 --> 00:43:30,889
not the same abstraction

722
00:43:31,139 --> 00:43:31,750
as

723
00:43:34,360 --> 00:43:35,449
I mean, this is the

724
00:43:35,770 --> 00:43:36,659
Oh Hey Hegy.

725
00:43:37,510 --> 00:43:42,689
So I was talking about one of the things I liked about flex is it can do context aware

726
00:43:43,469 --> 00:43:44,260
escaping.

727
00:43:45,310 --> 00:43:48,169
And this is sort of the opposite of that where

728
00:43:49,370 --> 00:43:49,860
it's

729
00:43:50,649 --> 00:43:53,070
just kind of slapping strings around.

730
00:43:54,290 --> 00:43:55,219
And so

731
00:43:56,419 --> 00:44:00,979
I'm just trying to match the databases ideas of what characters have to

732
00:44:00,989 --> 00:44:04,429
be escaped when you are searching in a string in boolean mode.

733
00:44:04,800 --> 00:44:07,110
And then it's nested inside of quotes.

734
00:44:08,879 --> 00:44:10,899
It is so hard to trust this code.

735
00:44:11,409 --> 00:44:12,250
I've put it off

736
00:44:13,080 --> 00:44:16,419
tests around it that I'm comfortable with it in production,

737
00:44:17,979 --> 00:44:21,379
but that it's actually an unusually large amount of tests

738
00:44:21,709 --> 00:44:23,419
for our code base.

739
00:44:23,709 --> 00:44:24,320
And

740
00:44:27,419 --> 00:44:28,030
I just,

741
00:44:30,070 --> 00:44:33,850
I hate that sequel is a string instead of a structured thing.

742
00:44:37,399 --> 00:44:40,060
If somebody would like to port link

743
00:44:40,370 --> 00:44:42,260
to Ruby, that would be very nice.

744
00:44:42,649 --> 00:44:45,459
That's almost Ruby object mapper.

745
00:44:46,070 --> 00:44:46,459
But

746
00:44:47,709 --> 00:44:48,090
yeah,

747
00:44:52,679 --> 00:44:56,860
so I was really surprised that he calls that strip operators.

748
00:44:57,709 --> 00:44:59,729
I think he's trying to get

749
00:45:00,909 --> 00:45:01,459
to,

750
00:45:02,000 --> 00:45:05,520
to switch between. What is the user searching

751
00:45:06,419 --> 00:45:06,959
input.

752
00:45:07,939 --> 00:45:08,479
Yeah,

753
00:45:08,969 --> 00:45:09,290
I mean,

754
00:45:09,300 --> 00:45:13,689
my favorite part of C# is how they keep stealing functional programming ideas from

755
00:45:16,010 --> 00:45:16,840
F sharp.

756
00:45:18,820 --> 00:45:20,709
All right. Not stealing, obviously

757
00:45:20,919 --> 00:45:22,899
boarding, migrating, translating

758
00:45:24,669 --> 00:45:27,370
and link is one of those. And it is such a good idea.

759
00:45:36,649 --> 00:45:39,550
I suppose this is the one place in the code.

760
00:45:44,090 --> 00:45:45,830
So it tracks

761
00:45:45,939 --> 00:45:49,360
the input that's cue the query variable,

762
00:45:50,909 --> 00:45:54,929
but it doesn't really have an internal abstraction layer of

763
00:45:57,669 --> 00:46:02,189
how that translates into a query. This class kind of is that abstraction layer.

764
00:46:03,419 --> 00:46:06,560
So I guess that makes sense that he called strip operators

765
00:46:07,929 --> 00:46:11,219
to know what's the thing that's gonna actually end up in the query?

766
00:46:17,840 --> 00:46:18,260
Just

767
00:46:19,560 --> 00:46:22,449
there's something about it that rubs me the wrong way seeing.

768
00:46:31,050 --> 00:46:32,649
So let's think about what it's gonna do.

769
00:46:33,169 --> 00:46:35,540
So if somebody searches for

770
00:46:49,250 --> 00:46:50,659
it translates

771
00:46:51,520 --> 00:46:53,340
or I guess it escapes,

772
00:46:53,520 --> 00:46:54,340
what do we call it

773
00:46:55,469 --> 00:46:56,409
formats?

774
00:46:58,669 --> 00:47:00,590
FDF

775
00:47:01,320 --> 00:47:02,860
kind of build up a table here.

776
00:47:03,989 --> 00:47:07,370
Like if you have an input with nothing interesting, it's just translated.

777
00:47:07,699 --> 00:47:08,850
You have, can't

778
00:47:09,530 --> 00:47:11,709
let me get rid of these quotes or this is gonna become

779
00:47:13,139 --> 00:47:13,840
confusing

780
00:47:17,139 --> 00:47:18,629
that still becomes can't.

781
00:47:19,850 --> 00:47:21,590
But if you search for

782
00:47:23,969 --> 00:47:25,439
like food

783
00:47:26,280 --> 00:47:27,159
bar

784
00:47:27,770 --> 00:47:28,750
that turns into

785
00:47:29,000 --> 00:47:29,189
a boo

786
00:47:29,439 --> 00:47:30,219
bar,

787
00:47:33,639 --> 00:47:36,179
we do not pass that punctuation through.

788
00:47:36,370 --> 00:47:37,659
And then obviously the

789
00:47:38,199 --> 00:47:41,939
the scary one is something more like this where

790
00:47:42,709 --> 00:47:47,489
if someone passed a single quote or a double quote through, we really,

791
00:47:48,159 --> 00:47:52,139
we don't even wanna try matching the perfect level of escaping,

792
00:47:52,300 --> 00:47:55,060
especially because it's just gonna get thrown away.

793
00:48:02,679 --> 00:48:05,139
And so when we're choosing what to highlight,

794
00:48:14,360 --> 00:48:16,969
when we're choosing what to highlight on the page.

795
00:48:19,120 --> 00:48:19,739
If

796
00:48:21,139 --> 00:48:21,709
come here,

797
00:48:23,620 --> 00:48:25,010
let's go look at the uh

798
00:48:27,820 --> 00:48:30,479
text here again. Now, I would like to see my example.

799
00:48:39,870 --> 00:48:43,290
So someone searched for types, comma whatever

800
00:48:49,429 --> 00:48:51,620
or they just searched for types comma

801
00:48:53,110 --> 00:48:56,689
the actual search executed would only look for the word types.

802
00:48:59,540 --> 00:49:02,699
And I think it's better to highlight

803
00:49:02,850 --> 00:49:04,479
and not include the comma

804
00:49:07,610 --> 00:49:10,070
cause we're not actually using it anyways.

805
00:49:12,530 --> 00:49:14,479
So here in my demo

806
00:49:15,399 --> 00:49:20,050
and I generated this demo by just editing the html in the inspector.

807
00:49:22,810 --> 00:49:24,939
You know, there's another possibility there

808
00:49:25,120 --> 00:49:25,830
of

809
00:49:28,689 --> 00:49:32,780
what is making some of these recent issues more attractive than others.

810
00:49:34,209 --> 00:49:38,550
And I guess one of the issues or one of the differences here is not just

811
00:49:39,080 --> 00:49:41,689
you're playing with the parser, it's kind of a CS problem,

812
00:49:42,310 --> 00:49:43,229
but also

813
00:49:44,129 --> 00:49:48,300
there's mock ups. It is very clear what the output is going to be.

814
00:49:50,719 --> 00:49:55,780
So I made this by hand and I highlighted the period just because it was there.

815
00:49:55,790 --> 00:49:57,060
And I was thinking of,

816
00:49:57,510 --> 00:49:57,719
well,

817
00:49:57,729 --> 00:49:59,979
if you found the start of the search term and then you found the

818
00:49:59,989 --> 00:50:02,939
end of the search term and you highlighted everything between you'd be fine.

819
00:50:03,679 --> 00:50:07,530
But if they had highlighted just correctly and sometimes

820
00:50:09,500 --> 00:50:12,780
it's a little funny that that period would be hanging out here.

821
00:50:12,790 --> 00:50:14,719
And if we imagine that um

822
00:50:15,560 --> 00:50:19,750
they were quoting a fellow Gen Xer who ended with a ellipse,

823
00:50:19,870 --> 00:50:21,580
you know, so if this ended dot dot dot

824
00:50:22,389 --> 00:50:25,090
It would be a little more obvious that the dot is

825
00:50:26,459 --> 00:50:28,439
not getting highlighted but

826
00:50:28,600 --> 00:50:30,679
that would be fine.

827
00:50:33,209 --> 00:50:35,840
It doesn't feel like it's incorrect or misleading.

828
00:50:35,850 --> 00:50:38,840
We're, we're kind of indicating what actual search happened.

829
00:50:40,189 --> 00:50:44,320
I think I'm talking myself into the use of strip operators being ok.

830
00:50:44,989 --> 00:50:46,209
And the result of

831
00:50:47,060 --> 00:50:49,080
finding these terms without.

832
00:50:51,219 --> 00:50:51,590
Yeah.

833
00:50:58,310 --> 00:50:58,989
Yeah. OK.

834
00:51:03,250 --> 00:51:04,050
All right. So if

835
00:51:06,620 --> 00:51:09,370
strip operators is fine,

836
00:51:10,629 --> 00:51:12,600
a little weird but fine,

837
00:51:24,600 --> 00:51:25,419
then I see.

838
00:51:27,570 --> 00:51:28,620
Why would the

839
00:51:30,010 --> 00:51:31,330
rent the seas get

840
00:51:33,129 --> 00:51:34,030
highlighted

841
00:51:38,919 --> 00:51:42,260
cause they get escaped but these get thrown away.

842
00:51:44,770 --> 00:51:46,379
Yeah, I don't understand that test

843
00:51:47,479 --> 00:51:48,719
and I have to come back to that.

844
00:51:56,149 --> 00:51:57,850
Yeah, I think my only gripe is

845
00:52:03,280 --> 00:52:05,360
guest strip operators is public.

846
00:52:08,219 --> 00:52:09,719
It feels weird.

847
00:52:10,590 --> 00:52:10,969
Yeah.

848
00:52:14,879 --> 00:52:15,739
All right. So

849
00:52:18,370 --> 00:52:19,550
I'll take the

850
00:52:20,320 --> 00:52:21,649
search model

851
00:52:21,889 --> 00:52:22,530
said it

852
00:52:23,280 --> 00:52:24,379
exposes,

853
00:52:32,530 --> 00:52:35,010
I guess the thing that confuses me here is

854
00:52:36,139 --> 00:52:36,959
it's using

855
00:52:37,469 --> 00:52:40,360
poems very directly instead of

856
00:52:41,629 --> 00:52:42,770
instead of using

857
00:52:45,370 --> 00:52:46,159
the

858
00:52:52,300 --> 00:52:53,570
search dot QE

859
00:52:54,320 --> 00:52:56,600
and it is readable. So yeah,

860
00:52:57,649 --> 00:52:59,399
let's just go ahead and make that a suggestion.

861
00:53:00,780 --> 00:53:02,169
I guess the part that really

862
00:53:02,629 --> 00:53:05,080
there's like so many layers of abstraction happening

863
00:53:14,070 --> 00:53:15,659
and then does it make sense

864
00:53:16,199 --> 00:53:16,929
or?

865
00:53:25,340 --> 00:53:27,409
So strip operators works on a single

866
00:53:29,870 --> 00:53:33,800
works on a single string. You can't actually pass in an array like

867
00:53:34,080 --> 00:53:34,889
he does.

868
00:53:44,689 --> 00:53:45,110
Yeah.

869
00:54:08,270 --> 00:54:09,489
Yeah. I guess the thing that

870
00:54:10,000 --> 00:54:13,889
most bugs me is he's kind of recreating some of the search work.

871
00:54:16,479 --> 00:54:18,939
So here we build up this terms array

872
00:54:19,419 --> 00:54:21,189
when we're doing a comment search

873
00:54:21,610 --> 00:54:23,689
because there are many kinds of operator.

874
00:54:24,820 --> 00:54:27,360
And I think what we actually want is to say,

875
00:54:32,419 --> 00:54:35,050
let's loop the parse tree

876
00:54:37,649 --> 00:54:38,419
and

877
00:54:40,540 --> 00:54:42,760
for the term or catchalls

878
00:54:43,330 --> 00:54:44,800
we'll call strip on it.

879
00:54:46,270 --> 00:54:47,139
And then, yeah.

880
00:54:48,959 --> 00:54:49,560
So

881
00:55:03,739 --> 00:55:06,320
what was the name of that function? I can see it off screen.

882
00:55:41,320 --> 00:55:41,500
Hm.

883
00:55:47,020 --> 00:55:47,050
Ok.

884
00:56:03,389 --> 00:56:03,459
I

885
00:56:04,199 --> 00:56:04,489
hope

886
00:56:06,719 --> 00:56:08,139
actually this should just be on the,

887
00:56:08,959 --> 00:56:09,379
yeah,

888
00:57:34,469 --> 00:57:35,580
operators.

889
00:57:36,510 --> 00:57:37,020
All right.

890
00:58:25,139 --> 00:58:25,919
Maybe.

891
00:58:36,699 --> 00:58:37,439
Yeah,

892
00:58:37,590 --> 00:58:39,229
I'll let him come up with his own name.

893
00:58:54,790 --> 00:58:55,459
Here we are.

894
00:58:56,290 --> 00:58:58,729
Hopefully that kind of nudges him towards a uh

895
00:58:59,070 --> 00:59:00,389
functional style.

896
00:59:02,750 --> 00:59:06,850
I don't think I need to be explicit because I can see a few ways to do this

897
00:59:07,399 --> 00:59:10,290
and if it's got a good obstruction on it, it is. OK.

898
01:00:08,969 --> 01:00:09,629
Here we go.

899
01:00:11,090 --> 01:00:12,139
I'm trying to,

900
01:00:12,699 --> 01:00:18,110
I am hedging a little to avoid saying you have to deal with quoted for this pr versus

901
01:00:18,820 --> 01:00:22,760
I will take it without, I'm not totally ready to commit because I can't

902
01:00:23,399 --> 01:00:26,439
imagine yet what the whole thing will look like.

903
01:00:31,600 --> 01:00:34,419
This is. This is a pretty decent start.

904
01:01:15,389 --> 01:01:15,879
I know

905
01:01:17,189 --> 01:01:18,360
strip note that

906
01:01:19,110 --> 01:01:20,090
punctuation.

907
01:01:29,219 --> 01:01:29,590
So

908
01:03:03,919 --> 01:03:05,229
yeah, I guess part of this is

909
01:03:05,879 --> 01:03:06,959
let me say that

910
01:03:36,320 --> 01:03:38,340
the understanding

911
01:03:40,550 --> 01:03:42,830
uh search syntax

912
01:03:43,060 --> 01:03:47,060
and how to format term stays within the search.

913
01:03:51,280 --> 01:03:54,379
I can hear my spouse starting to knock around in the kitchen. So we're

914
01:03:55,229 --> 01:03:59,149
immediately starting to wind up towards the end of this stream.

915
01:03:59,159 --> 01:04:03,510
Even if it's only been an hour, I'm not gonna hang up in the next five minutes,

916
01:04:03,520 --> 01:04:05,649
but in the next 30.

917
01:04:06,699 --> 01:04:07,719
So let me.

918
01:04:09,699 --> 01:04:10,199
Yeah.

919
01:04:12,479 --> 01:04:13,120
OK.

920
01:04:19,290 --> 01:04:20,100
I think

921
01:04:20,800 --> 01:04:24,550
if I look at the spec for the search model,

922
01:04:32,719 --> 01:04:34,500
I think the other thing would be,

923
01:04:36,500 --> 01:04:36,989
yeah,

924
01:04:38,590 --> 01:04:41,129
this uh doesn't permit sequel injection.

925
01:04:47,379 --> 01:04:51,040
This was such an unfun test to write because

926
01:04:51,590 --> 01:04:58,459
I am searching for inserting quotes into a string that is quoted multiple times and

927
01:04:58,469 --> 01:05:03,239
then in the spec I have to write quotes around my search and so it's just

928
01:05:04,129 --> 01:05:06,270
how many layers of escaping are there?

929
01:05:08,969 --> 01:05:08,989
Yeah.

930
01:05:16,649 --> 01:05:17,310
This one

931
01:05:18,129 --> 01:05:19,090
as a fun

932
01:05:19,439 --> 01:05:20,040
uh

933
01:05:20,310 --> 01:05:25,149
Actually all these 30 I even left a comment. These are real things.

934
01:05:25,159 --> 01:05:28,770
People run vulnerability scanners against lobsters all the time.

935
01:05:29,620 --> 01:05:30,840
And the way the

936
01:05:30,989 --> 01:05:34,520
vulnerability scanner works is it sees a box with

937
01:05:34,530 --> 01:05:37,229
the word search nearby and it starts stuffing random

938
01:05:38,399 --> 01:05:39,100
uh

939
01:05:39,889 --> 01:05:41,110
strings into it

940
01:05:42,850 --> 01:05:44,270
and it includes

941
01:05:44,709 --> 01:05:46,909
so it's trying to do sequel injection here.

942
01:05:48,209 --> 01:05:48,860
And these,

943
01:05:48,879 --> 01:05:51,689
the thing that's happening with these random

944
01:05:51,699 --> 01:05:54,669
strings like this or these random values

945
01:05:55,540 --> 01:05:56,229
R

946
01:05:57,949 --> 01:06:00,969
the bot is checking to see if

947
01:06:02,780 --> 01:06:06,760
on the page that comes back from the submitting the search form.

948
01:06:07,090 --> 01:06:08,540
Do any of these appear?

949
01:06:10,149 --> 01:06:12,290
Because if they appear, it's

950
01:06:12,820 --> 01:06:15,409
it kind of implies or that

951
01:06:17,209 --> 01:06:18,969
you have mis parched the query.

952
01:06:25,260 --> 01:06:27,189
Yeah. Yeah. Here's the other one. Pu

953
01:06:28,000 --> 01:06:28,580
So

954
01:06:30,080 --> 01:06:31,739
on the other hand, it also makes your

955
01:06:31,939 --> 01:06:34,600
queries really, really stand out

956
01:06:35,020 --> 01:06:35,739
in the

957
01:06:35,949 --> 01:06:38,530
weblog when I see you searching for gibberish,

958
01:06:42,709 --> 01:06:46,209
I understand some of this is just the background radiation of the internet.

959
01:06:46,219 --> 01:06:47,489
But also this is

960
01:06:49,070 --> 01:06:52,010
we're at nerd oriented site. And so nerds are like,

961
01:06:52,320 --> 01:06:55,090
I need an example site. So we get a lot of like

962
01:06:56,129 --> 01:06:58,449
the first web crawler someone has ever written

963
01:06:58,689 --> 01:07:00,500
with no rate limiting, no

964
01:07:00,899 --> 01:07:02,979
robots. TXT, no anything.

965
01:07:03,879 --> 01:07:03,949
Uh,

966
01:07:04,969 --> 01:07:07,840
and we get people who are like, what's a vulnerability scanner?

967
01:07:07,850 --> 01:07:10,270
Oh, I just installed one. How do I use it?

968
01:07:10,370 --> 01:07:10,560
Oh,

969
01:07:10,570 --> 01:07:13,570
let me run it against this website that I was just looking at

970
01:07:13,580 --> 01:07:15,879
in any other tab and then they run it against our prod,

971
01:07:16,620 --> 01:07:17,229
which is,

972
01:07:18,290 --> 01:07:19,419
it's real rude.

973
01:07:19,800 --> 01:07:21,489
That's also a good way to get your IP band.

974
01:07:27,959 --> 01:07:29,699
And I'm real aggressive about

975
01:07:31,510 --> 01:07:33,969
if production throws an exception, I'm gonna fix it.

976
01:07:34,590 --> 01:07:37,689
There isn't because our site traffic is low enough that, you know,

977
01:07:37,699 --> 01:07:39,669
I don't really worry about the,

978
01:07:39,830 --> 01:07:41,330
is it a one in a million

979
01:07:42,090 --> 01:07:45,439
ram got hit by a cosmic ray kind of error?

980
01:07:45,590 --> 01:07:47,070
That thing happens at scale

981
01:07:47,969 --> 01:07:49,639
doesn't happen at lobster scale.

982
01:07:50,860 --> 01:07:52,830
Although I have seen one or two things that were,

983
01:07:54,780 --> 01:07:57,979
it's strongly indicative of hardware failure on the client.

984
01:08:04,169 --> 01:08:06,739
So at this point, vulnerability scanners don't

985
01:08:07,399 --> 01:08:08,689
throw five hundreds.

986
01:08:10,580 --> 01:08:12,350
They do make me annoyed if I see them though.

987
01:08:30,660 --> 01:08:32,080
Yeah, I don't want this here.

988
01:08:33,509 --> 01:08:34,060
All right.

989
01:08:42,799 --> 01:08:42,850
OK.

990
01:09:00,269 --> 01:09:00,809
Well,

991
01:09:01,358 --> 01:09:03,328
does he need any for sequel injection

992
01:09:04,368 --> 01:09:05,417
for highlighting?

993
01:09:06,759 --> 01:09:08,890
No, as long as strip operators happens internally,

994
01:09:08,899 --> 01:09:11,339
I think that's fine because this is just,

995
01:09:11,709 --> 01:09:12,799
this is just highlighting.

996
01:09:13,709 --> 01:09:14,160
Yeah.

997
01:09:17,839 --> 01:09:18,529
Hey, chill.

998
01:09:18,810 --> 01:09:21,990
Uh so I am reviewing this pull request

999
01:09:22,220 --> 01:09:23,109
right now,

1000
01:09:24,099 --> 01:09:25,249
which is

1001
01:09:25,389 --> 01:09:28,758
my own feature request for highlighting search terms.

1002
01:09:28,769 --> 01:09:30,398
Is there supposed to be music?

1003
01:09:30,809 --> 01:09:32,398
No. Why do you hear some?

1004
01:09:38,339 --> 01:09:39,350
Oh I think

1005
01:09:43,620 --> 01:09:48,310
Chill. You're like a popular streamer, aren't you? As opposed to you know me a yacht?

1006
01:09:49,099 --> 01:09:51,979
I think I started following you on Twitch recently

1007
01:09:51,988 --> 01:09:54,559
cause I heard a podcast interview with you.

1008
01:09:58,169 --> 01:10:00,839
Was that you or was that somebody else? Were you on the

1009
01:10:01,240 --> 01:10:04,390
Dead Code podcast? No, it's just mentioned in the rules.

1010
01:10:04,770 --> 01:10:05,240
OK.

1011
01:10:06,040 --> 01:10:07,649
But you do stream code, right?

1012
01:10:19,620 --> 01:10:20,240
OK.

1013
01:10:20,899 --> 01:10:22,080
Yeah, I've clicked over.

1014
01:10:22,290 --> 01:10:24,729
I have an offscreen browser. I've clicked over your,

1015
01:10:25,200 --> 01:10:26,399
I see you do um

1016
01:10:29,240 --> 01:10:32,979
a bunch of stuff. You're on maternity leave.

1017
01:10:34,959 --> 01:10:38,029
You aren't the person I just heard on the now I have to find it.

1018
01:10:42,660 --> 01:10:47,620
So I ran into Jared Norman earlier this year at Madison Ruby.

1019
01:10:52,609 --> 01:10:53,560
No. OK.

1020
01:10:54,379 --> 01:10:55,439
Yeah, this is you.

1021
01:10:55,810 --> 01:10:56,680
I did just hear

1022
01:10:56,819 --> 01:10:59,439
a, an interview with you. This is why I was following you.

1023
01:10:59,689 --> 01:11:00,290
Uh

1024
01:11:00,660 --> 01:11:04,600
is you were just on this podcast and talking about how

1025
01:11:05,779 --> 01:11:08,970
streaming has meant a bunch of opportunities in your career.

1026
01:11:09,859 --> 01:11:14,240
This was a really nice, let's, let's share this in the chat and I'll stick it in the um

1027
01:11:14,979 --> 01:11:15,839
show notes.

1028
01:11:16,660 --> 01:11:18,040
Anyways, thanks for dropping in.

1029
01:11:18,879 --> 01:11:19,439
Uh

1030
01:11:25,040 --> 01:11:29,399
you know, it's not just a very black metal piece of podcast art.

1031
01:11:29,410 --> 01:11:35,359
He actually commissioned a musician to make a little like metal bumper.

1032
01:11:36,069 --> 01:11:38,390
And so it has someone doing the screamy metal

1033
01:11:38,399 --> 01:11:41,169
voice saying like Dead Code and then like playing

1034
01:11:41,439 --> 01:11:44,209
really heavy distorted guitar.

1035
01:11:45,000 --> 01:11:46,140
So uh

1036
01:11:49,729 --> 01:11:55,600
the first couple of episodes, he didn't know what a compressor was yet. I think.

1037
01:11:56,109 --> 01:12:00,669
So. The first couple, like the first three or four episodes,

1038
01:12:01,479 --> 01:12:03,740
his metal intro is a little bit loud.

1039
01:12:04,439 --> 01:12:08,339
Uh I mentioned if you're sensitive to that and then he started working with

1040
01:12:09,750 --> 01:12:12,950
um Mandy Moore who is a

1041
01:12:14,120 --> 01:12:15,399
sort of famous

1042
01:12:15,879 --> 01:12:16,680
tech, I mean,

1043
01:12:16,790 --> 01:12:17,620
to me,

1044
01:12:18,029 --> 01:12:21,799
a tech podcast producer and as soon as she came on board,

1045
01:12:21,810 --> 01:12:23,959
his audio took a step up in quality.

1046
01:12:24,620 --> 01:12:26,359
Yeah. Yeah, it does slap. Oh

1047
01:12:27,060 --> 01:12:27,069
I

1048
01:12:27,180 --> 01:12:28,120
actually have,

1049
01:12:29,689 --> 01:12:32,370
I think I have desktop audio wired up

1050
01:12:32,870 --> 01:12:35,500
and I could probably play it. But I, I will just

1051
01:12:36,000 --> 01:12:37,120
let's see if it works.

1052
01:12:37,540 --> 01:12:39,700
So if I unmute that

1053
01:12:40,290 --> 01:12:41,140
and that,

1054
01:12:45,459 --> 01:12:46,020
I think it'll

1055
01:12:46,680 --> 01:12:50,049
actually notice that you, you choose the advent of code.

1056
01:12:50,060 --> 01:12:51,759
If you ever want help with your,

1057
01:12:52,069 --> 01:12:55,709
if you like the box in the corner, you can put like a little excerpt.

1058
01:12:56,930 --> 01:12:59,120
It was real nice if you could just sweep it

1059
01:12:59,279 --> 01:13:00,919
and you could like the chat in there.

1060
01:13:01,580 --> 01:13:02,140
Maybe

1061
01:13:02,669 --> 01:13:03,189
he's pretty over

1062
01:13:03,290 --> 01:13:03,540
late.

1063
01:13:04,819 --> 01:13:06,080
Good idea. Good idea.

1064
01:13:07,330 --> 01:13:08,620
So it's the um

1065
01:13:09,149 --> 01:13:14,589
it is definitely not wired up correctly because you are hearing,

1066
01:13:15,279 --> 01:13:18,049
especially if that was faint. You were hearing the

1067
01:13:22,330 --> 01:13:25,779
my monitor is playing the audio and then my microphone

1068
01:13:25,790 --> 01:13:28,279
is picking it up as it reflects off of me.

1069
01:13:28,970 --> 01:13:33,850
So I'm not gonna debug audio on stream. That's a little too painful.

1070
01:13:34,770 --> 01:13:36,910
Also, it's gotta be a trip for

1071
01:13:37,779 --> 01:13:39,089
uh jail codes to

1072
01:13:39,830 --> 01:13:40,549
hear me

1073
01:13:40,899 --> 01:13:42,930
playing an interview with her. So

1074
01:13:43,959 --> 01:13:45,169
in any case, uh

1075
01:13:45,750 --> 01:13:51,140
where I was going with that trail is uh if you wanna start streaming after me,

1076
01:13:51,149 --> 01:13:55,540
let me know and cause I'm getting towards wrapping up in like 1520 minutes

1077
01:13:56,439 --> 01:13:59,339
and I will throw over to you or raid.

1078
01:14:00,129 --> 01:14:02,439
Yeah, when I had to start,

1079
01:14:04,200 --> 01:14:08,819
when I uh, like, started giving conference talks and I was like, ok, well,

1080
01:14:08,830 --> 01:14:11,930
now you have to watch your own conference talks so that you can improve.

1081
01:14:12,350 --> 01:14:12,370
Ah,

1082
01:14:12,490 --> 01:14:13,609
that's such a trip.

1083
01:14:14,330 --> 01:14:14,990
And

1084
01:14:19,870 --> 01:14:21,990
doing it with streams is even worse.

1085
01:14:22,000 --> 01:14:25,620
Cuz it's just me talking to an empty room for hours.

1086
01:14:25,629 --> 01:14:27,419
I've watched one or two of those streams.

1087
01:14:29,560 --> 01:14:31,740
Oh, that's ok. Your baby can code on stream.

1088
01:14:33,799 --> 01:14:36,299
Congrats. I hope you're having a good holiday. Actually,

1089
01:14:38,209 --> 01:14:38,990
I am.

1090
01:14:39,000 --> 01:14:41,160
So I'm gonna roll out in a bit because we are

1091
01:14:41,169 --> 01:14:44,229
also hosting a Thanksgiving and there's plenty of prep to do,

1092
01:14:44,450 --> 01:14:48,439
although we did most of it over the last couple of days. I feel so adult.

1093
01:14:48,549 --> 01:14:50,270
Like, we weren't up all night

1094
01:14:50,490 --> 01:14:52,310
the night before a big dinner party.

1095
01:14:54,669 --> 01:14:55,129
All right.

1096
01:14:59,700 --> 01:15:04,160
Ah, no, I have to watch my talk so that I can get better at them and I don't watch

1097
01:15:05,250 --> 01:15:08,629
all of my stream playbacks because it's like, all right, if I'm doing, like,

1098
01:15:08,640 --> 01:15:09,850
six hours of stream a week,

1099
01:15:09,859 --> 01:15:13,669
that's already kind of a full work day of streaming every week.

1100
01:15:14,180 --> 01:15:15,470
I don't need to double that,

1101
01:15:16,439 --> 01:15:16,979
but

1102
01:15:17,160 --> 01:15:17,919
I do

1103
01:15:18,580 --> 01:15:22,299
watch bits of it to make sure that stuff came across. Or

1104
01:15:22,750 --> 01:15:26,899
if I remember that, I sounded particularly funny. I wanna hear what that sounds like

1105
01:15:28,660 --> 01:15:33,439
or I'm just debugging my stream archives and then I have to like click play on things.

1106
01:15:35,910 --> 01:15:36,350
Yeah.

1107
01:15:38,720 --> 01:15:39,979
All right. So this is,

1108
01:15:40,910 --> 01:15:41,970
this is really nice.

1109
01:15:41,979 --> 01:15:44,509
I should make more screenshots in my future requests

1110
01:15:44,520 --> 01:15:46,370
cause maybe that's the thing that people are like.

1111
01:15:46,379 --> 01:15:46,770
Yeah,

1112
01:15:47,009 --> 01:15:48,720
I understand what this person wants.

1113
01:15:55,140 --> 01:15:59,290
I almost wonder if they should pass in the search object instead of

1114
01:16:00,240 --> 01:16:02,379
extracting stuff and passing it in.

1115
01:16:03,419 --> 01:16:04,439
I'll leave it as a No.

1116
01:16:51,810 --> 01:16:52,339
Great.

1117
01:16:55,649 --> 01:16:57,029
That's what I'm gonna say.

1118
01:16:58,370 --> 01:16:59,870
Let's request some changes

1119
01:17:23,479 --> 01:17:23,549
which

1120
01:17:24,580 --> 01:17:24,600
um

1121
01:17:33,459 --> 01:17:33,470
I,

1122
01:18:11,680 --> 01:18:14,100
I don't, I guess they are term terms,

1123
01:18:14,490 --> 01:18:16,459
not yet quoted term. Yeah. All right.

1124
01:18:36,729 --> 01:18:37,350
Oh Chill.

1125
01:18:37,359 --> 01:18:40,669
You know, if this is your first time tuning in, maybe I should have answered though.

1126
01:18:40,680 --> 01:18:42,790
What are you doing a little more broadly of

1127
01:18:43,819 --> 01:18:48,450
I run Lobsters, which is a big web form for discussing programming. So we are

1128
01:18:50,250 --> 01:18:52,629
reviewing code to search

1129
01:18:53,069 --> 01:18:54,169
a site about code

1130
01:18:54,830 --> 01:18:55,379
very meta

1131
01:18:59,009 --> 01:19:04,060
and on these streams also on off, they are kind of lobsters office hours.

1132
01:19:04,069 --> 01:19:07,750
There is a link under the description to

1133
01:19:08,220 --> 01:19:12,029
my web page where I kind of introduce things and have the archive for the stream

1134
01:19:12,970 --> 01:19:13,790
and

1135
01:19:15,069 --> 01:19:15,540
good.

1136
01:19:15,779 --> 01:19:16,479
That's so great.

1137
01:19:20,970 --> 01:19:21,950
All right. So

1138
01:19:22,279 --> 01:19:26,390
Ashwin, Ashwin actually opened two poll requests

1139
01:19:27,060 --> 01:19:28,259
and this one,

1140
01:19:28,549 --> 01:19:29,640
where am I on time?

1141
01:19:32,640 --> 01:19:34,890
I think we can take a look at it. So

1142
01:19:35,930 --> 01:19:39,740
as I mentioned, I like to make sure the site doesn't throw exceptions in prod.

1143
01:19:41,620 --> 01:19:43,700
Oh, as long as I'm between poll requests,

1144
01:19:43,709 --> 01:19:47,100
I am saying the bumper thing the same amount of times, but you know,

1145
01:19:48,020 --> 01:19:50,669
four times in 90 minutes instead of four times in

1146
01:19:51,109 --> 01:19:52,040
three hours.

1147
01:19:52,629 --> 01:19:55,640
But if anybody has any questions about the site, the community,

1148
01:19:56,339 --> 01:19:57,399
the guidelines,

1149
01:19:58,290 --> 01:19:59,859
why things are the way they are

1150
01:20:01,040 --> 01:20:04,700
happy to talk about those happy to run queries against the Proud Database

1151
01:20:06,200 --> 01:20:12,020
or otherwise, I do code review and try and keep the code base moving and in good shape

1152
01:20:12,709 --> 01:20:13,640
or at least

1153
01:20:14,350 --> 01:20:15,850
break it less.

1154
01:20:17,250 --> 01:20:18,029
So

1155
01:20:19,379 --> 01:20:21,490
something broke the

1156
01:20:22,129 --> 01:20:24,810
stats page which has our nice graphs on it.

1157
01:20:27,750 --> 01:20:31,810
So if I start up my rail server in the background

1158
01:20:31,959 --> 01:20:33,770
and let's go to

1159
01:20:39,129 --> 01:20:41,129
this will probably throw 500

1160
01:20:43,899 --> 01:20:45,879
I mean production is. So

1161
01:20:47,509 --> 01:20:49,720
um stats also,

1162
01:20:49,859 --> 01:20:50,830
yeah, there we go

1163
01:20:50,930 --> 01:20:53,689
is cashed because these are big queries that are like,

1164
01:20:53,700 --> 01:20:56,000
let's look at every comment in the database.

1165
01:20:56,009 --> 01:20:57,970
Let's look at every story in the database.

1166
01:21:00,009 --> 01:21:01,709
And as long as we're gonna write

1167
01:21:02,879 --> 01:21:04,959
giant complicated queries,

1168
01:21:05,410 --> 01:21:06,819
let's not add any tests.

1169
01:21:09,279 --> 01:21:11,689
I'm a little embarrassed that I didn't even have a

1170
01:21:11,700 --> 01:21:13,979
smoke test to see that the stats page loads.

1171
01:21:14,649 --> 01:21:16,990
So this is probably related to

1172
01:21:18,370 --> 01:21:19,299
a recent

1173
01:21:21,270 --> 01:21:25,529
change I asked for replacing, execute with select all.

1174
01:21:25,839 --> 01:21:26,689
So, I wonder

1175
01:21:26,910 --> 01:21:29,890
if I change this back to execute?

1176
01:21:33,939 --> 01:21:35,720
Will this immediately go away?

1177
01:21:39,540 --> 01:21:40,149
Um,

1178
01:21:47,779 --> 01:21:49,100
yeah, still thinking about it.

1179
01:21:52,060 --> 01:21:55,990
If it is thinking longer it's probably working. Yeah.

1180
01:21:56,200 --> 01:21:59,080
So this is what the stats page is supposed to look like.

1181
01:22:08,439 --> 01:22:09,799
So it is.

1182
01:22:12,259 --> 01:22:14,250
Yeah, I don't think I need to add that. So then

1183
01:22:15,100 --> 01:22:16,439
the fix that

1184
01:22:17,500 --> 01:22:18,520
Ashwin maid,

1185
01:22:19,839 --> 01:22:21,649
I'm a little suspicious of it.

1186
01:22:24,209 --> 01:22:26,149
So there's a thing

1187
01:22:26,250 --> 01:22:28,910
I've talked about how sometimes I look at code and I'm like,

1188
01:22:28,919 --> 01:22:30,629
ah a junior developer wrote that

1189
01:22:30,810 --> 01:22:33,029
this one pinged that sense where

1190
01:22:35,270 --> 01:22:39,890
instead of trying to fix the query or

1191
01:22:40,890 --> 01:22:43,109
fix something downstream

1192
01:22:45,439 --> 01:22:46,979
and say, OK, well,

1193
01:22:47,270 --> 01:22:49,799
the query is gonna output a slightly different format.

1194
01:22:49,810 --> 01:22:51,970
Let's change the way it consumes

1195
01:22:52,149 --> 01:22:52,740
data.

1196
01:22:53,779 --> 01:22:54,870
It says, well,

1197
01:22:54,879 --> 01:22:57,779
let's loop over all of the data and inspect it and

1198
01:22:57,790 --> 01:23:00,319
tweak things and then put it back and pass it off

1199
01:23:02,149 --> 01:23:03,450
and it's such a

1200
01:23:03,700 --> 01:23:03,709
a,

1201
01:23:06,529 --> 01:23:10,080
it's a thing that Junior devs do often where

1202
01:23:10,209 --> 01:23:14,109
they don't try and find the root cause of something or change

1203
01:23:14,629 --> 01:23:17,919
the output one place. They just kind of look at

1204
01:23:18,060 --> 01:23:21,189
data and munch it where the exception happens

1205
01:23:21,359 --> 01:23:22,629
to it occur.

1206
01:23:35,689 --> 01:23:37,609
The exception occurs

1207
01:23:38,919 --> 01:23:41,290
line 90

1208
01:23:42,839 --> 01:23:43,649
it was 90.

1209
01:23:46,669 --> 01:23:50,649
So we grab all the data, we turn it into an array and we flatten it.

1210
01:23:51,680 --> 01:23:52,379
Is this

1211
01:23:52,859 --> 01:23:53,660
all right? Hang on.

1212
01:23:54,490 --> 01:23:58,229
So let's grab a console and let's say, what's the model here

1213
01:23:58,870 --> 01:23:59,910
user? All right,

1214
01:24:03,870 --> 01:24:08,169
cause I think this might be as simple as is stuff coming back as a hash or an array.

1215
01:24:18,209 --> 01:24:20,990
What do you mean user out from users?

1216
01:24:23,390 --> 01:24:23,930
Fucking?

1217
01:24:31,930 --> 01:24:31,939
I

1218
01:24:32,040 --> 01:24:32,250
get it. So

1219
01:24:32,720 --> 01:24:36,729
this is like a weird trilogy thing where it doesn't play well with my pager.

1220
01:24:50,100 --> 01:24:52,129
And so I get just useless.

1221
01:24:52,509 --> 01:24:53,879
Ask the art stuff.

1222
01:24:54,859 --> 01:24:56,390
Julia Evans just had a great

1223
01:24:58,870 --> 01:25:01,399
a string of posts on blue sky about

1224
01:25:02,240 --> 01:25:04,080
how it is a like

1225
01:25:04,689 --> 01:25:09,169
guideline of writing shell stuff that you're supposed to detect. If your pager is

1226
01:25:13,120 --> 01:25:16,060
not a person and then strip out your

1227
01:25:17,120 --> 01:25:19,640
export pager equals

1228
01:25:19,979 --> 01:25:21,319
see if this wants to do it.

1229
01:25:28,120 --> 01:25:29,020
All right. That's better.

1230
01:25:32,000 --> 01:25:34,140
So it comes back as a result

1231
01:25:34,899 --> 01:25:38,709
and then it has rows and this is an array of arrays.

1232
01:25:40,410 --> 01:25:41,509
If I said

1233
01:25:41,939 --> 01:25:42,910
select

1234
01:25:44,089 --> 01:25:45,810
oh part of that is

1235
01:25:46,430 --> 01:25:48,770
I should just say select one because I care about the structure.

1236
01:25:48,779 --> 01:25:50,100
I don't care how many I get.

1237
01:25:50,810 --> 01:25:52,250
So select all

1238
01:25:52,660 --> 01:25:53,709
gets me

1239
01:25:55,000 --> 01:25:56,089
a result

1240
01:25:56,299 --> 01:25:57,640
with the columns one.

1241
01:25:58,709 --> 01:26:01,089
And if I say each

1242
01:26:05,830 --> 01:26:05,850
the

1243
01:26:08,370 --> 01:26:10,069
I get hashes back,

1244
01:26:11,609 --> 01:26:12,200
OK?

1245
01:26:12,750 --> 01:26:14,259
But if I say execute,

1246
01:26:16,339 --> 01:26:17,870
I get an array back. So

1247
01:26:19,479 --> 01:26:21,379
this is not a pull request,

1248
01:26:21,390 --> 01:26:23,020
I'm gonna accept because what happened is

1249
01:26:23,029 --> 01:26:25,899
we switched from execute to select all because

1250
01:26:26,029 --> 01:26:29,259
rails, plumbing works a little different. We don't need to blow out the

1251
01:26:30,379 --> 01:26:31,049
cash.

1252
01:26:34,379 --> 01:26:36,390
But having made this change,

1253
01:26:40,770 --> 01:26:42,109
this is never gonna get

1254
01:26:44,089 --> 01:26:45,490
an array back anymore.

1255
01:26:48,109 --> 01:26:49,700
It shouldn't be possible.

1256
01:27:09,970 --> 01:27:12,120
So if it's always gonna come back as

1257
01:27:13,839 --> 01:27:14,799
an array,

1258
01:27:25,649 --> 01:27:27,520
it's always going to come back as an

1259
01:27:28,990 --> 01:27:29,299
right?

1260
01:27:30,319 --> 01:27:32,439
No, it's always gonna come back as a hash.

1261
01:27:33,229 --> 01:27:35,250
Do I wanna just call values on everything?

1262
01:27:35,569 --> 01:27:36,890
Would that fix this?

1263
01:27:40,240 --> 01:27:41,120
All right. So

1264
01:27:41,629 --> 01:27:43,330
put it back into this state

1265
01:27:44,540 --> 01:27:46,589
and then down here

1266
01:27:51,709 --> 01:27:54,180
yield should grab the results.

1267
01:28:08,240 --> 01:28:08,560
So

1268
01:28:16,750 --> 01:28:17,660
like I might need

1269
01:28:17,919 --> 01:28:21,740
an extra layer of two A dot values dot flatten, but

1270
01:28:24,660 --> 01:28:26,339
it's something more like that

1271
01:28:26,740 --> 01:28:27,700
rather than,

1272
01:28:29,259 --> 01:28:31,990
yeah, for an instance of active record result.

1273
01:28:32,859 --> 01:28:34,470
So I'm gonna tweak that to say

1274
01:28:36,040 --> 01:28:36,990
to a

1275
01:28:37,589 --> 01:28:39,990
and then that should get me an array of hashes.

1276
01:28:47,560 --> 01:28:47,609
OK.

1277
01:28:54,040 --> 01:28:57,529
As I look at this, I'm not even sure what the flat is doing there knowing

1278
01:28:59,459 --> 01:29:00,540
and breaking things,

1279
01:29:15,790 --> 01:29:17,979
a defined method for an instance of array.

1280
01:29:20,020 --> 01:29:21,660
But the problem is that it's

1281
01:29:28,689 --> 01:29:30,950
so this is gonna throw an exception. I just wanna see the

1282
01:29:32,589 --> 01:29:33,240
glasses.

1283
01:29:34,200 --> 01:29:35,939
So these are all arrays.

1284
01:30:52,609 --> 01:30:52,640
OK?

1285
01:30:53,229 --> 01:30:53,259
OK.

1286
01:30:56,330 --> 01:30:57,930
This give me multiple roads,

1287
01:30:59,779 --> 01:31:00,240
huh?

1288
01:31:05,540 --> 01:31:05,669
It will,

1289
01:31:17,660 --> 01:31:19,399
this gets me two hashes.

1290
01:31:30,430 --> 01:31:30,490
Wait a

1291
01:31:30,759 --> 01:31:30,819
moment,

1292
01:31:32,270 --> 01:31:33,930
this gets me to a race.

1293
01:31:38,870 --> 01:31:40,430
So why does this not?

1294
01:31:45,160 --> 01:31:47,830
This goes and selects this stuff out?

1295
01:32:16,359 --> 01:32:17,419
Let's go into a bugger

1296
01:32:22,020 --> 01:32:22,370
like

1297
01:32:22,569 --> 01:32:23,209
this is

1298
01:32:23,390 --> 01:32:25,169
I'm clearly wasting time for de

1299
01:32:25,270 --> 01:32:25,790
debugging

1300
01:32:27,250 --> 01:32:27,879
debug.

1301
01:32:28,080 --> 01:32:28,520
Uh

1302
01:32:32,790 --> 01:32:33,330
There we go.

1303
01:32:34,450 --> 01:32:35,939
It's data class.

1304
01:32:36,819 --> 01:32:37,790
It's a hash.

1305
01:32:52,180 --> 01:32:52,209
OK.

1306
01:32:53,640 --> 01:32:55,180
Yeah. And I'm, I'm such a

1307
01:32:55,759 --> 01:32:58,729
old school din line guy that I rarely remember

1308
01:32:58,919 --> 01:33:00,810
to just drop into a debugger.

1309
01:33:01,669 --> 01:33:03,549
There was a really nice talk at

1310
01:33:04,779 --> 01:33:06,370
Rocky Mountain ruby.

1311
01:33:06,500 --> 01:33:07,200
I think.

1312
01:33:07,879 --> 01:33:08,689
Where was that?

1313
01:33:12,520 --> 01:33:13,540
If you wanna write.

1314
01:33:14,180 --> 01:33:14,879
Um,

1315
01:33:16,069 --> 01:33:17,129
a bunch of,

1316
01:33:18,649 --> 01:33:19,379
let's see.

1317
01:33:19,770 --> 01:33:20,910
I make sure I'm muted.

1318
01:33:28,310 --> 01:33:29,209
Where was this?

1319
01:33:29,649 --> 01:33:31,759
Oh, this must be a year or something ago.

1320
01:33:33,660 --> 01:33:34,200
Come here.

1321
01:33:36,270 --> 01:33:37,720
Where is the actual

1322
01:33:41,350 --> 01:33:42,850
try until you die? Here you go.

1323
01:33:43,609 --> 01:33:45,839
This was a nice talk about

1324
01:33:46,020 --> 01:33:47,770
features in pride.

1325
01:33:49,080 --> 01:33:52,310
And I started watching it and then I learned like five things and I went,

1326
01:33:52,319 --> 01:33:53,040
my brain is full.

1327
01:33:53,049 --> 01:33:54,939
I can't actually learn five things at once.

1328
01:33:58,979 --> 01:34:00,279
Uh, So I have to

1329
01:34:01,180 --> 01:34:02,669
come back and watch that again.

1330
01:34:08,810 --> 01:34:09,870
All right. So

1331
01:34:11,770 --> 01:34:12,589
that's it. Right.

1332
01:34:13,779 --> 01:34:15,129
Could not parse time.

1333
01:34:25,509 --> 01:34:27,149
Yeah, I don't understand. I'm

1334
01:34:28,020 --> 01:34:30,600
kind of dinking around cause I'm trying to,

1335
01:34:33,270 --> 01:34:35,759
I'm just gonna get that. Why am exception? Aren't I?

1336
01:34:36,729 --> 01:34:39,080
Cause this is what the coat already is? Yeah.

1337
01:34:47,310 --> 01:34:48,850
So this is not all right.

1338
01:34:49,350 --> 01:34:52,020
That's enough to say that this pull request isn't gonna do it.

1339
01:34:52,029 --> 01:34:56,069
Let's just punt it back. I like I look at these things and I want to tweak it, but

1340
01:35:15,470 --> 01:35:19,399
I gotta get it up on screen so I can actually say it correctly.

1341
01:35:22,740 --> 01:35:22,759
So

1342
01:35:24,049 --> 01:35:25,509
the race and

1343
01:35:26,600 --> 01:35:29,580
select all returns and array

1344
01:35:30,310 --> 01:35:31,290
hatches.

1345
01:35:37,919 --> 01:35:39,009
Thanks for

1346
01:35:42,509 --> 01:35:43,459
Tinker a

1347
01:35:43,560 --> 01:35:44,330
bit

1348
01:35:45,370 --> 01:35:48,450
and learn that this is bad. This is death.

1349
01:35:54,589 --> 01:35:56,580
This isn't going to be

1350
01:35:58,549 --> 01:35:59,680
changing the

1351
01:36:23,060 --> 01:36:25,830
today, the data of

1352
01:36:27,830 --> 01:36:28,439
the

1353
01:36:29,020 --> 01:36:32,049
class of the array elements

1354
01:36:37,109 --> 01:36:38,399
should always be

1355
01:36:39,089 --> 01:36:39,790
ashes.

1356
01:36:43,049 --> 01:36:47,990
Don't like this style of inspecting data in the middle

1357
01:36:48,910 --> 01:36:49,580
and

1358
01:36:51,270 --> 01:36:52,660
we came back

1359
01:36:53,890 --> 01:36:55,109
is either

1360
01:36:59,250 --> 01:37:01,270
either update the

1361
01:37:03,560 --> 01:37:04,339
query

1362
01:37:04,979 --> 01:37:06,379
and I don't wanna tell him to

1363
01:37:24,220 --> 01:37:24,839
There we go.

1364
01:37:27,109 --> 01:37:28,850
I feel like a very,

1365
01:37:29,819 --> 01:37:32,819
I hope I am being a competent open source maintainer

1366
01:37:33,069 --> 01:37:35,359
by saying like, hey, yeah, you got this.

1367
01:37:36,919 --> 01:37:41,020
It's really tempting, especially with this kind of like, oh, I almost have it bug.

1368
01:37:41,029 --> 01:37:45,339
Like I did not mean to rabbit hole for five or 10 minutes and try and fix it myself.

1369
01:37:45,720 --> 01:37:48,729
What I should have said to encourage a contributor is just like,

1370
01:37:50,000 --> 01:37:52,770
yeah, the styles off, you're mostly there. Why don't you?

1371
01:37:52,850 --> 01:37:54,870
I could have written this comment 10 minutes ago.

1372
01:37:55,669 --> 01:37:57,870
As soon as I saw the difference between execute and

1373
01:37:57,879 --> 01:38:00,330
select all I should have just come back and said,

1374
01:38:01,100 --> 01:38:02,100
please keep going.

1375
01:38:03,279 --> 01:38:04,370
It's the

1376
01:38:04,819 --> 01:38:06,569
uh the wanting to do it myself.

1377
01:38:07,259 --> 01:38:09,419
I am trying to get away from that and

1378
01:38:09,430 --> 01:38:12,080
encourage more people to contribute to the code base.

1379
01:38:12,089 --> 01:38:14,319
It is a heck of a lot more scalable than

1380
01:38:15,600 --> 01:38:17,310
me trying to do everything myself.

1381
01:38:18,970 --> 01:38:19,549
All right.

1382
01:38:20,259 --> 01:38:21,439
So with that,

1383
01:38:22,100 --> 01:38:25,339
I am gonna roll out. This is gonna be a very short stream.

1384
01:38:25,370 --> 01:38:29,259
Um Unless anybody has any last minute questions about

1385
01:38:29,750 --> 01:38:32,000
the lobster site or code base.

1386
01:38:32,930 --> 01:38:36,419
Uh And I understand from tail codes that she's gonna

1387
01:38:36,430 --> 01:38:39,189
start writing rails and contributing to the code base.

1388
01:38:41,740 --> 01:38:44,430
Oops, why is this bug still first, still open?

1389
01:38:45,479 --> 01:38:46,149
This is done

1390
01:38:49,819 --> 01:38:50,509
and

1391
01:38:53,009 --> 01:38:59,680
given that we have 100 and 26 issues open and she has a baby in her arms.

1392
01:38:59,689 --> 01:39:02,109
Probably she's only gonna close half of them today.

1393
01:39:02,250 --> 01:39:03,149
That seems reasonable.

1394
01:39:05,950 --> 01:39:06,350
No.

1395
01:39:08,669 --> 01:39:09,649
So, anyways, uh,

1396
01:39:09,660 --> 01:39:15,890
I hope everybody has a nice holiday and if you don't celebrate Thanksgiving

1397
01:39:16,160 --> 01:39:17,779
because you're not an American,

1398
01:39:18,459 --> 01:39:19,049
uh,

1399
01:39:19,060 --> 01:39:21,279
the most important part of it is eating too

1400
01:39:21,290 --> 01:39:23,649
much food and hanging out with your family.

1401
01:39:23,919 --> 01:39:25,450
So I hope you can do that.

1402
01:39:28,910 --> 01:39:30,069
That's about it.

1403
01:39:31,229 --> 01:39:33,899
Oh, you already write rails and you're considering contributing well,

1404
01:39:33,910 --> 01:39:36,899
then I will try and not actually like, put you on the spot.

1405
01:39:36,910 --> 01:39:39,430
I hope I don't put you off. I'm just being silly

1406
01:39:39,890 --> 01:39:41,299
but um

1407
01:39:41,640 --> 01:39:42,250
I am

1408
01:39:42,979 --> 01:39:47,359
very welcoming to contributors if any of this stuff interests you.

1409
01:39:47,830 --> 01:39:50,339
I mean, I break stuff on the site all the time.

1410
01:39:50,569 --> 01:39:53,080
There is lots of stuff that's tagged good first issue.

1411
01:39:53,089 --> 01:39:55,279
So you don't have to know much about the code base

1412
01:39:56,109 --> 01:39:57,000
and

1413
01:39:59,000 --> 01:40:00,250
if you uh

1414
01:40:00,700 --> 01:40:02,799
want an invite to the site,

1415
01:40:02,939 --> 01:40:06,450
just let me know and I will email you one so that you can poke around.

1416
01:40:06,990 --> 01:40:08,649
Uh I just have to get

1417
01:40:09,399 --> 01:40:11,620
an email address for you. If you're not already on the site.

1418
01:40:14,589 --> 01:40:16,419
That is true for anybody,

1419
01:40:16,640 --> 01:40:18,819
please please help me fix my bugs.

1420
01:40:20,379 --> 01:40:21,000
Alrighty.

1421
01:40:22,220 --> 01:40:23,209
Have a good holiday.

1422
01:40:23,950 --> 01:40:24,549
Take care.