1
00:00:00,000 --> 00:00:03,840
 Welcome to Office Hours, it's your weekly half hour segment where I answer questions

2
00:00:03,840 --> 00:00:04,840
 about Neos.

3
00:00:04,840 --> 00:00:09,760
 We go ahead and get questions answered based on the questions and the order that they appear

4
00:00:09,760 --> 00:00:13,620
 inside the thread, so get those questions entered into that thread, and I'll get them

5
00:00:13,620 --> 00:00:15,120
 not answered in the order that they appear in.

6
00:00:15,120 --> 00:00:18,920
 If you're not sure what's going on, check the Office Hours check chat, the chat is above

7
00:00:18,920 --> 00:00:20,120
 the one you're currently in.

8
00:00:20,120 --> 00:00:23,800
 Looking at the list we have a few questions lined up, so go ahead and get those answered

9
00:00:23,800 --> 00:00:27,040
 whilst I try and find my English language inside my brain.

10
00:00:27,040 --> 00:00:32,240
 Spex asks, "Now that MMC is done this year, what's on your plate for Neos things?

11
00:00:32,240 --> 00:00:34,680
 Got any projects you're waiting to get into?"

12
00:00:34,680 --> 00:00:38,280
 Not really, just the same projects that existed that were kind of just delayed by MMC.

13
00:00:38,280 --> 00:00:44,640
 I got to the end of last week, sort of Sunday, which is yesterday, no day before yesterday,

14
00:00:44,640 --> 00:00:47,280
 and I sort of thought like, "I haven't done anything this week!" and then I looked and

15
00:00:47,280 --> 00:00:52,320
 was like, "Um, no, you kind of, you know, ran the voting for the rest of the weekend,

16
00:00:52,320 --> 00:00:56,920
 that's last weekend, or weekend before last, sorry, time is confusing."

17
00:00:56,920 --> 00:01:03,840
 You know, ran the voting, judged the entries, rehearsed the award show, went to the award

18
00:01:03,840 --> 00:01:09,000
 show, like lots went on, it just didn't feel like it did, so I was productive last week.

19
00:01:09,000 --> 00:01:14,120
 But yeah, that's all over now, so back on other things.

20
00:01:14,120 --> 00:01:18,120
 As a reminder for those who don't know what the MMC is, or even those that do, even if

21
00:01:18,120 --> 00:01:22,520
 you won, lost, entered, didn't enter, whatever it is, just remember that the stuff that is

22
00:01:22,520 --> 00:01:27,240
 created at the MMC can be created at any point during the year, you don't need the MMC as

23
00:01:27,240 --> 00:01:31,560
 an excuse to make it, so carry on making cool stuff in EOS, I love seeing it.

24
00:01:31,560 --> 00:01:36,040
 Moving forwards to the next question there, which is from Lex, who says "When silencing

25
00:01:36,040 --> 00:01:40,120
 a user through the sessions tab, the one that mutes a user for everyone, does that user's

26
00:01:40,120 --> 00:01:44,440
 voice audio stream stop being transmitted, or does it set the volume to zero?"

27
00:01:44,440 --> 00:01:47,280
 I don't know, I'd have to open the codebase and go take a look, I'll go ahead and get

28
00:01:47,280 --> 00:01:52,480
 that open, and then I'll dig back into it when I get a chance, as we go through the

29
00:01:52,480 --> 00:01:55,160
 normal ebb and flow of questions.

30
00:01:55,160 --> 00:01:58,120
 Unnamed Cyborg Cat says "Ever heard of Copy Kitty?

31
00:01:58,120 --> 00:02:00,880
 It's an amazing game the devs worked on very hard on, super underrated."

32
00:02:00,880 --> 00:02:01,880
 I have not.

33
00:02:01,880 --> 00:02:10,400
 I have like 1,800 games on Steam, and like 2,000+ on my wishlist, so probably it's on

34
00:02:10,400 --> 00:02:12,540
 there, I don't know though.

35
00:02:12,540 --> 00:02:16,560
 And with that, all the questions are over, I'm gonna mute Tech Humongous, Tech Humongous

36
00:02:16,560 --> 00:02:18,440
 scops of water, and then I'll be back.

37
00:02:18,440 --> 00:02:21,340
 And I'm back, there are no additional questions, if you have additional questions drop them

38
00:02:21,340 --> 00:02:25,360
 in the office hours texture, I'll get to them in the order that they appear in, whilst there

39
00:02:25,360 --> 00:02:27,760
 are no questions I'm just gonna sit here and think about cheese.

40
00:02:27,760 --> 00:02:31,160
 You'll be pleased to know my Amazon order has shipped, I don't know what it is, I don't

41
00:02:31,160 --> 00:02:33,440
 remember what it is.

42
00:02:33,440 --> 00:02:34,440
 What shipped?

43
00:02:34,440 --> 00:02:36,040
 I don't know, I'll figure that out later.

44
00:02:36,040 --> 00:02:39,400
 I was just reading the code on silence, looks like it will actually prevent the audio from

45
00:02:39,400 --> 00:02:46,200
 being transmitted, don't quote me, do your own testing, but based on my limited understanding

46
00:02:46,200 --> 00:02:48,640
 of how that works, it won't write to the value stream.

47
00:02:48,640 --> 00:02:52,640
 We have a question here from UnnamedCyborgCat, who says "Does Nio support any proprietary

48
00:02:52,640 --> 00:02:56,360
 mobile formats in the Source Engine?"

49
00:02:56,360 --> 00:03:04,440
 So the Nio's wiki has a list of stuff which people have looked at, and sort of work-ish.

50
00:03:04,440 --> 00:03:09,800
 Some of the problem here is that we, it's like, what we support isn't entirely up to

51
00:03:09,800 --> 00:03:17,280
 us, it's more up to Asimp, which we use to import assets.

52
00:03:17,280 --> 00:03:21,360
 If it's supported by those, then it will sometimes work.

53
00:03:21,360 --> 00:03:24,840
 Sometimes there's extra work we have to do once Asimp has handed it to us, but you can

54
00:03:24,840 --> 00:03:25,840
 see a list there.

55
00:03:25,840 --> 00:03:26,840
 Is that an alphabetical?

56
00:03:26,840 --> 00:03:27,840
 Yeah, that's an alphabetical.

57
00:03:27,840 --> 00:03:28,840
 Alphabetical.

58
00:03:28,840 --> 00:03:38,120
 It's difficult because things like FBX are technically proprietary, so yeah, BS supports

59
00:03:38,120 --> 00:03:43,560
 a proprietary file format, FBX is technically proprietary.

60
00:03:43,560 --> 00:03:44,560
 Don't use FBX, by the way.

61
00:03:44,560 --> 00:03:48,080
 I read this question like three times before unmuting it, so I'll just go ahead and read

62
00:03:48,080 --> 00:03:50,480
 it again and then think about it some more.

63
00:03:50,480 --> 00:03:54,800
 Lex asks "Why does setting the parent of a slot using the set parent node of an invalid

64
00:03:54,800 --> 00:04:02,280
 or null to slot parents the from slot to root?"

65
00:04:02,280 --> 00:04:05,720
 Set parent with an invalid to parents the from slot to root.

66
00:04:05,720 --> 00:04:08,760
 I mean, it should just cancel the operation.

67
00:04:08,760 --> 00:04:12,680
 I also wish more nodes had failure states.

68
00:04:12,680 --> 00:04:16,280
 Let me take a look here.

69
00:04:16,280 --> 00:04:21,960
 What I want is more nodes to have on failure, so if it pulsed with onFail coming out of

70
00:04:21,960 --> 00:04:32,440
 set parent, then it would be good to do that.

71
00:04:32,440 --> 00:04:42,080
 The new parent will default to the root slot if you pass in null or an invalid value.

72
00:04:42,080 --> 00:04:46,420
 I think a lot of this is just like various sort of hacky patches that we've been doing

73
00:04:46,420 --> 00:04:50,520
 over the years to try and prevent null parents from happening.

74
00:04:50,520 --> 00:04:55,920
 Anywhere that touches the slot hierarchy model, we try and basically interject and say "hey,

75
00:04:55,920 --> 00:04:59,160
 you're about to null parent this, don't do that!"

76
00:04:59,160 --> 00:05:01,360
 No, don't do that.

77
00:05:01,360 --> 00:05:07,240
 The behavior change that you're requesting where it doesn't do that is something that

78
00:05:07,240 --> 00:05:08,240
 would be cool.

79
00:05:08,240 --> 00:05:14,440
 Unfortunately it would also be a somewhat breaking change though, because it would increase

80
00:05:14,440 --> 00:05:19,320
 the amount of times where the set parent node wouldn't do something.

81
00:05:19,320 --> 00:05:27,640
 Like ran out, it would basically not do something if the thing you're trying to parent is null,

82
00:05:27,640 --> 00:05:34,600
 or if the new parent is a child of the target object, so you can't put something inside

83
00:05:34,600 --> 00:05:35,760
 itself.

84
00:05:35,760 --> 00:05:40,000
 In those cases it won't output failure, it will just literally do nothing, there will

85
00:05:40,000 --> 00:05:41,000
 be no output.

86
00:05:41,000 --> 00:05:46,640
 I hate nodes that don't have an output, because it's just very difficult to operate with.

87
00:05:46,640 --> 00:05:48,960
 It just defaults to world root if there's a problem.

88
00:05:48,960 --> 00:05:52,560
 Rampa says "nulling parent should just make nears destroy stuff, it has no parent so it

89
00:05:52,560 --> 00:05:53,560
 can't exist."

90
00:05:53,560 --> 00:05:58,880
 No not really, so while you, and I say you with a bunch of people who write mods in here

91
00:05:58,880 --> 00:06:05,480
 that know I'm wrong, but generally speaking, you are only able to access one hierarchy,

92
00:06:05,480 --> 00:06:08,120
 which is the world root hierarchy.

93
00:06:08,120 --> 00:06:12,680
 There are other hierarchies of slots and other types of components, and other lists and other

94
00:06:12,680 --> 00:06:19,920
 tree-based systems that exist outside of that system that some of this stuff operates in,

95
00:06:19,920 --> 00:06:25,800
 and so it's incorrect to say that if it's outside of the world root, it should be deleted.

96
00:06:25,800 --> 00:06:29,040
 It's kind of correct to say that it should be deleted if it doesn't have a parent, but

97
00:06:29,040 --> 00:06:33,200
 then again we should try and correct that rather than deleting it.

98
00:06:33,200 --> 00:06:37,920
 Your MMC project accidentally parents itself to null.

99
00:06:37,920 --> 00:06:41,320
 Yeah we delete it instead of trying to save your work for you and send it back to the

100
00:06:41,320 --> 00:06:42,320
 world root.

101
00:06:42,320 --> 00:06:48,160
 It's one of those things where it's like, this is gonna, yeah, um, I, Lex says "people

102
00:06:48,160 --> 00:06:51,280
 who are intentionally using that to parent things to root, or people who have different

103
00:06:51,280 --> 00:06:54,000
 logics to catch an unintentional root parent shouldn't be doing that."

104
00:06:54,000 --> 00:06:55,680
 Yeah, they shouldn't be doing that.

105
00:06:55,680 --> 00:07:00,840
 In fact if you're using root slot and you're not the owner of the world, I question your

106
00:07:00,840 --> 00:07:01,840
 motives.

107
00:07:01,840 --> 00:07:09,440
 Always use local user space, it is for the most part the world root, but in those one

108
00:07:09,440 --> 00:07:14,560
 few cases where you have world management systems using local user space, will keep

109
00:07:14,560 --> 00:07:18,400
 those world management systems more performant and easier to use.

110
00:07:18,400 --> 00:07:20,400
 So local user space.

111
00:07:20,400 --> 00:07:25,880
 Local user space returns the parent of the user who is running the set parent operation.

112
00:07:25,880 --> 00:07:29,800
 There is usually no downside of using local user space, I can think of a couple of cases

113
00:07:29,800 --> 00:07:33,780
 but only in world management systems where that might be problematic, so just use local

114
00:07:33,780 --> 00:07:34,780
 user space.

115
00:07:34,780 --> 00:07:43,680
 The amount of guns, RPGs, rocket launchers, grenades, bows and arrows, rifles, crossbows,

116
00:07:43,680 --> 00:07:49,620
 drink systems I see that are just like "cool, because I'm inconsiderate, I'll parent a bunch

117
00:07:49,620 --> 00:07:51,680
 of stuff to root when people use me."

118
00:07:51,680 --> 00:07:54,520
 It's like "cool, how about no."

119
00:07:54,520 --> 00:07:56,320
 So yeah, do think about that.

120
00:07:56,320 --> 00:08:01,920
 If you have a gun that produces, I should say entities, in a more traditional gaming

121
00:08:01,920 --> 00:08:07,880
 sense, like for those who are unaware, most guns in games are non-entity producing, so

122
00:08:07,880 --> 00:08:16,280
 they do bullets, and yeah sure bullets have animations, muzzle flash, smoke, explosions,

123
00:08:16,280 --> 00:08:19,120
 that sometimes traces the line that they take etc.

124
00:08:19,120 --> 00:08:23,120
 Those don't produce additional entities in the world, there's basically sort of a hitscan

125
00:08:23,120 --> 00:08:26,320
 with various complexities depending on what type of game it is.

126
00:08:26,320 --> 00:08:30,040
 Guns that produce entities are things like grenades, rocket launchers, the Needler from

127
00:08:30,040 --> 00:08:32,400
 Halo would be another good example.

128
00:08:32,400 --> 00:08:35,280
 They add additional stuff to the world when they are fired.

129
00:08:35,280 --> 00:08:38,960
 Crossbow and bow and arrow being a good sense, it sticks an entity to you.

130
00:08:38,960 --> 00:08:44,400
 Like a big rifle doesn't stick a world entity to the person, a bow and arrow does.

131
00:08:44,400 --> 00:08:51,440
 Which is kind of strange, because all weapons leave residue, guns leave impacted bullets

132
00:08:51,440 --> 00:08:53,240
 and casings and stuff like that.

133
00:08:53,240 --> 00:08:57,280
 I've never really seen a game make use of casings, sometimes they'll fall on the floor

134
00:08:57,280 --> 00:09:01,280
 but usually they're particle systems so they'll clear up after a bit, but I dunno, maybe it'll

135
00:09:01,280 --> 00:09:07,240
 be funny in a random shooting game, you start tripping over yourself if you stand still

136
00:09:07,240 --> 00:09:11,680
 for too long and start firing, you just can't keep your feet still when there's like casings

137
00:09:11,680 --> 00:09:12,680
 rolling under your feet.

138
00:09:12,680 --> 00:09:18,400
 Casings are usually quite easy to squish though, like dropping a bag of marbles on the floor

139
00:09:18,400 --> 00:09:22,240
 but just because you've been firing in the same place for too long.

140
00:09:22,240 --> 00:09:27,760
 Specs, actually let's go through a couple of comments here, so, Rampus says "the only

141
00:09:27,760 --> 00:09:29,840
 thing I can imagine not using user space is death effect".

142
00:09:29,840 --> 00:09:36,200
 No, user space should still work, because it'll parent above the user, not in the user.

143
00:09:36,200 --> 00:09:40,080
 If your world management system deletes the slot pertaining to the user when they leave,

144
00:09:40,080 --> 00:09:43,800
 then yes though, depends what's going on.

145
00:09:43,800 --> 00:09:46,600
 And test those, like extremely test them, it's difficult to test something where it's

146
00:09:46,600 --> 00:09:50,560
 like "when I leave the world, do something I can't see", because you're gone, but you

147
00:09:50,560 --> 00:09:52,640
 know what I mean, be careful.

148
00:09:52,640 --> 00:09:56,080
 Lex says "it's always funny when I see people using custom parenting systems to attach to

149
00:09:56,080 --> 00:09:59,640
 someone, then de-parent themselves to what a parent or parent thought they were in, and

150
00:09:59,640 --> 00:10:03,000
 get deleted when the user they previously parented to leaves".

151
00:10:03,000 --> 00:10:05,800
 Yeah, I've seen that happen a lot.

152
00:10:05,800 --> 00:10:10,560
 Virtual parent might be a solution there, if you know that it might be an unstable user

153
00:10:10,560 --> 00:10:12,640
 connection who might leave.

154
00:10:12,640 --> 00:10:15,480
 Specs says "there's a lot of cases where I've seen folks rely on uninitialized behaviour

155
00:10:15,480 --> 00:10:18,720
 and logics like that, and I have to admit it makes me wince.

156
00:10:18,720 --> 00:10:23,640
 No should users specifically initialize logics nodes with values instead of relying on default

157
00:10:23,640 --> 00:10:27,680
 behaviour that may change at any moment, unless a node is like a specifically duplicated design

158
00:10:27,680 --> 00:10:32,040
 to have an initial value that has some promise not to change without notice".

159
00:10:32,040 --> 00:10:35,440
 Yeah, nulls are terrible.

160
00:10:35,440 --> 00:10:43,280
 I am quite liking nullables in some C# projects I'm in right now, but it's just really annoying

161
00:10:43,280 --> 00:10:45,200
 like how much it comes up.

162
00:10:45,200 --> 00:10:49,320
 If you look it up, there's various articles, various quotes, various versions of it.

163
00:10:49,320 --> 00:10:52,480
 Someone in the computing world invented null, and they regret it.

164
00:10:52,480 --> 00:10:55,840
 I don't remember exactly who it was, it was either, I don't remember the name, it was

165
00:10:55,840 --> 00:11:00,080
 one of the big computer people from early on in computing.

166
00:11:00,080 --> 00:11:04,760
 It's the same thing with, the easiest way to identify null is if you look at a boolean.

167
00:11:04,760 --> 00:11:08,840
 So a boolean has true or false, on or off, yes or no, whatever you want to say.

168
00:11:08,840 --> 00:11:13,280
 But then with null, you introduce a third state of boolean, which is "I don't have a

169
00:11:13,280 --> 00:11:16,900
 value", and that's what null means in that case, "I don't have a value".

170
00:11:16,900 --> 00:11:20,980
 And that's perfectly fine to be used as "I don't know" or "I don't have a value because

171
00:11:20,980 --> 00:11:25,540
 I'm not initialised yet", but make that a value in your program.

172
00:11:25,540 --> 00:11:28,960
 Make the value be "I am initialised", not "I am null".

173
00:11:28,960 --> 00:11:35,600
 A good example of that comes up is in the sort of task-based system that C# has.

174
00:11:35,600 --> 00:11:39,000
 You have to wait for a task to complete, that's great.

175
00:11:39,000 --> 00:11:42,640
 Instead you could have written that and C# developers would have had hate about it, where

176
00:11:42,640 --> 00:11:47,680
 whilst you're waiting for a web request to happen, you have to while loop over the result

177
00:11:47,680 --> 00:11:51,720
 of that web request and check if it's not null, and you'd be like "Ah, it's happened!"

178
00:11:51,720 --> 00:11:55,220
 Instead they made this whole task system where it's like "Has the task completed?

179
00:11:55,220 --> 00:11:58,540
 The task has completed, here is the task's result".

180
00:11:58,540 --> 00:12:02,820
 And so it's like sort of implicitly built into the API that null doesn't occur, even

181
00:12:02,820 --> 00:12:07,740
 though it still does when people do tasks that return null for some reason.

182
00:12:07,740 --> 00:12:13,040
 When you're programming you should always be wondering "What do I do when X Y Z happens?

183
00:12:13,040 --> 00:12:14,980
 What do I do when X Y Z happens?"

184
00:12:14,980 --> 00:12:19,740
 If the code is telling you there's a possibility that null could exist here, you need to handle

185
00:12:19,740 --> 00:12:20,740
 it.

186
00:12:20,740 --> 00:12:26,020
 Like "Oh, the result of this method operation could be null, what are we going to do about

187
00:12:26,020 --> 00:12:27,020
 that?"

188
00:12:27,020 --> 00:12:29,140
 Same thing with objects though, we need more onFails.

189
00:12:29,140 --> 00:12:31,660
 What if the parenting fails, I need to know about it.

190
00:12:31,660 --> 00:12:33,480
 Yeah I don't like people who do that.

191
00:12:33,480 --> 00:12:40,080
 I always add the default of 0 anyway and people question why and I say "Well, it's there,

192
00:12:40,080 --> 00:12:44,760
 I can change it if I need to, I don't usually need to, but I also find it a little bit easier

193
00:12:44,760 --> 00:12:50,580
 to sort of compute in my head when I can see the default value, otherwise I have to remember

194
00:12:50,580 --> 00:12:55,440
 extra information, not only what the default value is, but is the default value acceptable

195
00:12:55,440 --> 00:12:56,700
 for that occurrence?"

196
00:12:56,700 --> 00:12:59,420
 For those who are like "Well I don't like adding the default value because that's an

197
00:12:59,420 --> 00:13:02,340
 extra slot", just go away.

198
00:13:02,340 --> 00:13:04,380
 Like it's one extra slot, come on.

199
00:13:04,380 --> 00:13:08,220
 I mean in most cases it will default to 0.

200
00:13:08,220 --> 00:13:12,240
 In cases where we have changed the default value of something, we are able to upgrade

201
00:13:12,240 --> 00:13:15,940
 logics to inject the old default value.

202
00:13:15,940 --> 00:13:19,700
 We did that once with findChildByName or something like that.

203
00:13:19,700 --> 00:13:24,780
 Say for example, I don't know, the inputs to + default to 0, let's say they change to

204
00:13:24,780 --> 00:13:29,120
 1, for some ungodly reason, the world would be quite screwed up for that to happen.

205
00:13:29,120 --> 00:13:33,580
 But we would basically, when we load your logics back in, we would detect "Hey you haven't

206
00:13:33,580 --> 00:13:40,660
 specified a default, so we will default this to 0".

207
00:13:40,660 --> 00:13:51,860
 And preserve the old behavior by automatically injecting a 0 input node for you, and that

208
00:13:51,860 --> 00:13:52,860
 way you'll be fine.

209
00:13:52,860 --> 00:13:59,460
 It might be the divide node, let me see, div float, nah, I forgot about that, yeah.

210
00:13:59,460 --> 00:14:05,220
 The divide nodes will allow divide by 0 errors, I forgot about that.

211
00:14:05,220 --> 00:14:08,680
 That's why we have the filter invalid node, to get rid of those cases where you have an

212
00:14:08,680 --> 00:14:09,680
 invalid value.

213
00:14:09,680 --> 00:14:16,460
 I always put, or try to, I often forget, I try to put filter invalid in places where

214
00:14:16,460 --> 00:14:19,140
 a 0 might occur with user input.

215
00:14:19,140 --> 00:14:25,260
 For example if there's a form field where they're entering a number, if they need to

216
00:14:25,260 --> 00:14:29,500
 clear out all of the number to enter their new number, then it might, for an update or

217
00:14:29,500 --> 00:14:31,260
 two, it might have a 0 in it.

218
00:14:31,260 --> 00:14:34,580
 And a couple of times that stung me and broken something, because I don't expect there to

219
00:14:34,580 --> 00:14:37,580
 be a 0 there, so filter invalid helps with that.

220
00:14:37,580 --> 00:14:42,440
 For example the value graph recorder, or it might be the graphing mesh, can't remember

221
00:14:42,440 --> 00:14:48,280
 which, it doesn't like if the number of values that it can store is less than 2.

222
00:14:48,280 --> 00:14:52,720
 It has to be 2 or greater, or it breaks and disables itself.

223
00:14:52,720 --> 00:14:54,240
 >> Are there any additional questions?

224
00:14:54,240 --> 00:14:58,040
 We've got about 5 minutes left in this office hours segment, we could talk about much more

225
00:14:58,040 --> 00:15:02,640
 programming stuff for a long time, or we could, any questions that happen, go for them.

226
00:15:02,640 --> 00:15:04,440
 Whatever you feel like.

227
00:15:04,440 --> 00:15:07,080
 >> Whilst I wait for questions.

228
00:15:07,080 --> 00:15:13,360
 There's a TV show on Netflix called Shadow and Bone, I don't know what it's called.

229
00:15:13,360 --> 00:15:15,240
 Shadow and Bone, yeah there we go.

230
00:15:15,240 --> 00:15:16,240
 Shadow and Bone.

231
00:15:16,240 --> 00:15:20,280
 I was quite enjoying it until they literally resolved the entire storyline in the finale.

232
00:15:20,280 --> 00:15:24,000
 Like there were a couple of episodes in the middle of the first season where nothing was

233
00:15:24,000 --> 00:15:28,400
 happening, and in the last episode an entire season worth of content happens and I'm just

234
00:15:28,400 --> 00:15:31,240
 like, "Guys, this isn't how you end a season."

235
00:15:31,240 --> 00:15:37,040
 You know, it's just like, "Oh yeah, it's mid-season episode, the characters have a cup of tea."

236
00:15:37,040 --> 00:15:38,880
 End of the season.

237
00:15:38,880 --> 00:15:44,200
 The characters travel halfway across the world, defeat the big baddie, throw the ring into

238
00:15:44,200 --> 00:15:46,240
 Mordor, and go home.

239
00:15:46,240 --> 00:15:47,240
 Very annoying.

240
00:15:47,240 --> 00:15:50,560
 >> So Lex asks, "Would you possibly know when closing a session, sometimes it doesn't happen,

241
00:15:50,560 --> 00:15:53,320
 immediately it takes a few seconds to kick everyone and then close."

242
00:15:53,320 --> 00:15:55,560
 I think it's resolving stuff.

243
00:15:55,560 --> 00:15:56,560
 There's stuff it has to resolve.

244
00:15:56,560 --> 00:15:59,360
 Like for example if you're saving the world, that's going to add extra time to it closing

245
00:15:59,360 --> 00:16:03,320
 of course, but if you're not saving the world, there's got to be something that's resolving.

246
00:16:03,320 --> 00:16:04,960
 I'd have to look into it.

247
00:16:04,960 --> 00:16:08,300
 There's got to be something that's preventing it from closing, some sort of process that

248
00:16:08,300 --> 00:16:09,800
 needs to finish or something.

249
00:16:09,800 --> 00:16:13,520
 >> Nationwide asks, "Why do sessions that have been open for about three days start

250
00:16:13,520 --> 00:16:14,520
 to get all funky?

251
00:16:14,520 --> 00:16:19,280
 Do my headless pack all logics into the world into a single random thought?"

252
00:16:19,280 --> 00:16:23,040
 It did that by itself, or did it do that when you next tried to pack logics?

253
00:16:23,040 --> 00:16:24,520
 That sounds like it shouldn't...

254
00:16:24,520 --> 00:16:27,680
 I can understand if it's funky how you packed it.

255
00:16:27,680 --> 00:16:30,480
 Yeah, so after about three days, numbers start getting weird.

256
00:16:30,480 --> 00:16:33,640
 Like I know tea has issues, but I haven't really worried about that.

257
00:16:33,640 --> 00:16:38,300
 Like if you see tea getting issues, it's like you're doing something wrong probably.

258
00:16:38,300 --> 00:16:41,280
 So input tea will get funky after a little while.

259
00:16:41,280 --> 00:16:45,640
 I don't remember exactly how long, but it's basically when floating point imprecision

260
00:16:45,640 --> 00:16:51,320
 takes over and the value of tea starts getting funky if you use tea for everything.

261
00:16:51,320 --> 00:16:53,440
 Then maybe it ran out of like...

262
00:16:53,440 --> 00:16:56,000
 I don't know if it ran out of references or something?

263
00:16:56,000 --> 00:16:59,080
 I don't know why that would happen.

264
00:16:59,080 --> 00:17:00,480
 Dynamic impulses should be fine as well.

265
00:17:00,480 --> 00:17:05,320
 I'd have to look into the numbers involved with packing.

266
00:17:05,320 --> 00:17:07,400
 It's by sure a number problem.

267
00:17:07,400 --> 00:17:09,060
 It just doesn't look like it.

268
00:17:09,060 --> 00:17:10,060
 It's something...

269
00:17:10,060 --> 00:17:12,200
 We have one more question.

270
00:17:12,200 --> 00:17:16,920
 So I was distracted by reading through the logics to see what was going on there.

271
00:17:16,920 --> 00:17:21,460
 Ishimwade asks, "one more very quick, does the headless client have a memory leak?"

272
00:17:21,460 --> 00:17:25,740
 It's possible that the entirety of Neos has a memory leak.

273
00:17:25,740 --> 00:17:29,560
 Sometimes I have to remind people, and I get all the circumstances and everything, I'm

274
00:17:29,560 --> 00:17:32,480
 not here to talk about them, but I'm here to state facts.

275
00:17:32,480 --> 00:17:39,000
 When it comes to the headless client, you are playing with an alpha pre-release of a

276
00:17:39,000 --> 00:17:44,480
 feature for an early access beta video game.

277
00:17:44,480 --> 00:17:47,360
 So yeah, it's possible it has a memory leak.

278
00:17:47,360 --> 00:17:50,760
 All right, that is half past and I don't see any more questions coming.

279
00:17:50,760 --> 00:17:52,240
 So I'm going to go ahead and leave things there.

280
00:17:52,240 --> 00:17:55,160
 As a reminder, if you have questions, please feel free to direct message them, drop them

281
00:17:55,160 --> 00:17:57,680
 in the questions and help channel, or send a carrier pigeon my way.

282
00:17:57,680 --> 00:18:01,120
 I'll get to them as soon as I can and I will speak to you guys next week