This is a transcript of The Resonance from 2025 July 6.
00:00: Start recording.
00:02: Hello. Hello. Okay, we should be live.
00:06: Uh, let me post the announcements.
00:11: Posting in Discord.
00:14: There's actually two Discord ones
00:15: because this is office hours and live
00:16: streams.
00:19: Yes. There we go. There we go. And I'm
00:23: going to post the social media post
00:27: and post.
00:30: There we go.
00:32: Hello. Do we have people in our stream?
00:36: Hello. Hello is on Twitch. Hello PD_.
00:40: Hello.
00:42: IPC Schnopet. Oh, I do. Oh my god. We
00:45: were literally talking about it like
00:46: right before the stream stream started.
00:48: It was like one issue. But we're going
00:50: to we're going to get into it.
00:52: Um, yes.
00:55: But first, like, you know, the the
00:56: initial preamble. I'm getting kind
00:58: getting ahead of myself. Um, but anyway,
01:02: hello. Can you hear us? Fine. Is the
01:03: audio okay?
01:06: Pop Nemo. Hello. Hey, check the fox.
01:09: Thank you.
01:11: Check the fox out. Also made like a cool
01:13: new version of the Twitch chat. I can
01:15: show you like time stops now. He was
01:19: literally working on this like in a
01:20: world like yesterday.
01:22: And he was like, I need to I need to
01:24: make it work before the the end so it
01:26: doesn't explode. So, thank you, Jack.
01:30: I appreciate it. It looks
01:32: It looks fancier with the time stamps.
01:36: Oh, no. Right. The um the music the
01:40: audio is from the Twitch head and like
01:42: that's right behind the camera. So, you
01:43: probably heard it really loud. I'm
01:45: sorry. Um I'm going to actually move it
01:47: like over here. There we go.
01:50: Oh, we could probably turn the popping
01:51: noise off for the uh Twitch chat.
01:53: Yeah, let's let's turn the popping noise
01:55: off. Uh should time the message sound.
01:58: There we go.
01:59: Yeah, okay. Should be good.
02:02: Okay, there we go.
02:04: There we go. Okay. Hello everyone. Uh
02:07: can you hear us fine other than like
02:10: know the suddenly loud music?
02:13: Let's adjust the camera angle a bit.
02:19: So, hello everyone and welcome to
02:21: another episode of the resonance. Uh,
02:23: I'm FS and also have S with me. Um, you
02:26: made it this time.
02:28: Uh, and we'll be answering your
02:30: questions about Resonate. Whatever you
02:32: want to ask, uh, feel free. Uh, make
02:35: sure to put a question mark at the end
02:37: of your message. That way, like, you
02:39: know, pops on our thing. Uh, pretty much
02:41: like, you know, like so.
02:43: Um you can ask pretty much anything you
02:45: want about resite whe it's technical
02:47: philosophical it can even be like you
02:48: know personal stuff like that's kind of
02:50: nice like you know it's some lighter
02:52: stuff mixed in. Uh it can be anything
02:54: about the team, the future, the past
02:56: with the platform, whatever you want to
02:57: know. Feel free to ask some of the
02:59: questions. We might end up um hunting to
03:02: some of the other office hours. Like for
03:04: example, right before this one, that was
03:06: the moderation office hours. There's
03:07: also probably primes on Tuesday. Our
03:09: team has like ones on Wednesday. Uh J4
03:12: actually is planning one. Um I don't
03:14: remember where that one is,
03:16: unfortunately. Uh check the discord. Um
03:19: there's the events calendar and there's
03:21: also like information in de log. So you
03:23: can find you know more information about
03:25: those. Uh but we should be pretty much
03:28: ready to get started. We also have a few
03:30: questions from Discord. Uh but if you're
03:32: like you know watching this like ask ask
03:34: your questions on Twitch uh because
03:37: we're not going to be monitoring the
03:39: Discord anymore. Uh but we're going to
03:41: go through the Discord ones first
03:42: because it's like you know fixed amount
03:43: and we actually don't have too many this
03:45: uh this time around. Um, and then we're
03:48: going to get into the Twitch runs. So,
03:50: let's get started. Am I foretting
03:52: something
03:55: for anything?
03:57: I don't think so. I think we're good to
03:59: go.
04:01: Oh, okay. Uh,
04:04: let me actually also make I'm probably
04:06: going to make it. There we go. So, it
04:07: doesn't keep floating around. Um, so
04:11: let's check the first question from Oh,
04:14: it's actually one from ages. Um,
04:18: so Aegis,
04:20: I don't know if this is going to be
04:21: visible. Uh, Aegis is asking, "Been
04:25: working on a new hit scan based weapon
04:26: system that fakes projectiles as simply
04:29: as loop over uh, raycast chain following
04:31: the path of where the projectile is
04:34: going and using particle for projectile
04:36: visual. With that as context, two
04:39: questions I have. One, how is projectile
04:42: speed calculated? Is it meters/s or so?
04:45: Two, could I potentially use particle
04:47: attractors to make particle follow an
04:49: object or player? Uh, so one, it kind of
04:53: depends how you Thank you.
04:56: What was that? Uh, thank you Nikki for
04:58: the raid.
05:01: Uh so the first one I mean it kind of
05:03: depends how you calculate things like
05:05: like usually the math you use you know
05:07: that's your own but the units on the
05:09: resonite uh the distance uh that's
05:13: pretty much you can treat it as meters
05:16: uh especially in VR like you know if you
05:17: got like one unit that's going to like
05:20: it's going to look like one meter for
05:21: user at one scale. So if you're
05:24: calculating how much something how how
05:27: fast is something moving that would
05:28: probably be like you know meters/s which
05:30: are like sort of the standard units but
05:33: you can convert to units or whatever
05:35: internal system you're using it you
05:37: might be using different units and then
05:38: you need to kind of convert it but by
05:40: default yes it's going to be m/s.
05:43: Uh the second question, yes. Uh there's
05:45: the particle force. Uh so you could
05:48: like, you know, you could set up a
05:50: module that has the particle radial
05:51: force for example and you can position
05:53: that or drive, you know, the center of
05:56: that uh at an object or a player and
05:59: that will make the particles, you know,
06:01: follow the target. So you should be able
06:03: to do that. The only limitation is like
06:05: you know you need to actually add the
06:06: module to the particle system itself to
06:10: the style uh and make sure like you know
06:12: it's driven you know correctly like the
06:15: positions driven for like where the
06:16: player is and it's all in the correct
06:19: you know coordinate spaces but uh yep
06:22: pretty much like that should be
06:25: I don't think there's much more to add
06:26: to that.
06:29: Uh the next questions from Josh or Yosh,
06:33: sorry. Um
06:40: you're sure Josh knows. Um Yosh is
06:43: asking uh will the switch to system
06:45: numeric involve kicking the bucket on
06:47: most if not all of the very low play
06:50: code class? I don't know if I would use
06:52: that adjective on it like that thing has
06:54: been turning.
06:58: Oh, your microphone died.
07:06: Droo is getting his voice back, guys.
07:08: One moment.
07:28: I believe in you.
07:31: I I'll do a little jig while he uh fixes
07:33: his mic.
07:37: Oh, this one too. Okay, I think it's
07:39: back.
07:40: Ah, there you are.
07:42: Hello. Sorry, my microphone decided to
07:44: die and then not want to turn back on.
07:46: Um, but yeah, like every time I have to
07:49: touch the code class, I'm like, please
07:51: no more. We want to switch to net like09
07:54: and switch to nicer way. But um reading
07:57: the rest of the question, system memor
07:59: provides budent interfaces for basically
08:01: every mod operation that coder t
08:03: wrangles and it generic templating
08:05: madness uh which pro simplifies code a
08:08: bunch also lead a pretty neat speed
08:09: boost. A friend of mine benchmark uh 1
08:11: million additions using coder t versus
08:14: system numeric interfaces and float and
08:16: was a difference of 2.5 milliseconds
08:19: versus 0.2 20 milliseconds for them
08:21: after JIT caught up with them both. To
08:23: it seems system numeric vector types
08:25: don't implement any of these interfaces.
08:27: There's some interesting discussion as
08:29: to what's up with that. So uh it's not
08:31: going to be the switch to system numeric
08:34: itself. That's uh but we do want to kind
08:36: of get rid of the coder sheet um or at
08:39: least like get rid of like the large
08:42: chunk of it. Um but it's a separate
08:45: separate kind of task from the switch to
08:47: system numeric. um pretty much like one
08:50: of the really nice things with net 9 is
08:53: and actually I don't remember which
08:54: version they introduced in is but a lot
08:56: of the base data types they implement in
08:58: interfaces for stuff like addition so
09:01: you can have like you know two types
09:02: that can be added together and then you
09:04: can use it as a generic constraint and
09:06: it's something that we were not able to
09:08: use with mono because it just straight
09:10: up doesn't exist and that's kind of like
09:12: why code three was created and it just
09:14: kind of bloated over time as we needed
09:16: more and more functionality Um but with
09:19: net 9 once the splittering is like you
09:22: know the main part is done we definitely
09:25: do want to get rid of it. We want to
09:26: like rework a lot of the parts and use
09:29: the interfaces and the generic
09:31: constraints because that's going to make
09:32: the code so much nicer and also like
09:34: more performant. Generally I don't think
09:36: it's going to like um affect like that
09:40: much in terms of like practical
09:42: performance because usually we try to
09:44: avoid using the coder for like a lot of
09:46: heavy math stuff. there's like few
09:48: places where it's used a bit but usually
09:51: we kind of avoid it like you know for
09:52: the reason that there's a fair bit of
09:54: overhead with it. Um so it's only used
09:56: like in cases where it's like very
09:58: unavoidable
10:00: and it shouldn't be like you know where
10:01: most of the code is like being where
10:03: most of the time is being spent. So I
10:06: don't know how much of a difference it's
10:07: going to make for you know overall
10:08: performance but um every little bit
10:11: helps and it's going to make the code
10:13: much nicer to work with anyway. So we're
10:16: going to do it just not as part of the
10:18: switch to system numeric. It's it's a
10:20: separate task.
10:22: I think like more so what would help is
10:26: like switching like the vector types out
10:28: themselves for like system numeric
10:30: vectors or like basing them off of the
10:32: system numeric types cuz like
10:34: even just changing the types around like
10:37: just changing the types around uh to
10:40: provide like
10:41: that's what that's what a switch to
10:42: system numeric is. It's pretty much like
10:44: using using those for like you know the
10:46: vector operations and that like I do
10:49: expect to have like kind of substantial
10:50: boost um because that's like where a lot
10:53: of the heavy math is but it also you
10:55: know it doesn't have anything to do with
10:57: the code that's more like you know just
10:59: the core data types.
11:00: So that's like what I'm saying like
11:01: those two are separate tasks.
11:04: Um
11:06: and then we got one more question from
11:08: the discord. Uh Mintshock is asking
11:13: uh what is the most expensive part of
11:16: resinite back end cloud storage as a
11:17: distribution variation generation cloud
11:19: wars or anything else? Um so it kind of
11:23: depends. I don't have the numbers at
11:25: hand but it depends a little bit on how
11:29: you would slice it. I can see if I can
11:31: log in and thing real quick but I might
11:34: not be able to do that in VR. Um
11:38: let's see if this loads up.
11:42: other video I can get into it. Uh
11:44: actually no, never mind. This is not
11:45: this is my personal Azure. Um
11:49: yeah, I have to like pull the numbers
11:50: together.
11:52: Yeah, I can't I can't log in into like
11:54: in VR unfortunately. Um
11:57: I would say it kind of depends how you
11:58: slice it because there's like you know
12:00: for example what do you consider storage
12:02: because um the biggest part with that
12:07: cost the most like like they would
12:08: probably see is compute uh I thought
12:11: storage also takes a chunk but like it's
12:14: kind of hard to like separate some of
12:15: these costs because you know in order to
12:17: use some storage we have to use compute
12:19: and then to use compute you also need
12:21: some storage like for example you know
12:24: the compute services for API that's
12:26: going to use stuff like uh cues and it's
12:29: going to use stuff like you know uh like
12:31: a radius cache which you can consider
12:33: kind of storage. So
12:37: it depends a bit like how you slice it.
12:41: It's it's it's like um it's a little bit
12:44: hard to like you know separate them. uh
12:46: cloud it's also like some of these are
12:48: not like individual things like cloud
12:49: wars like you know that's mostly using
12:51: like uh storage like tables uh but
12:54: there's also like you know compute
12:56: involved because you need the API to
12:57: actually serve the requests and there
12:59: might be harder to kind of isolate some
13:01: of these uh some of them as well for
13:04: example the variation generation we have
13:07: like a server with a fixed cost that
13:09: like you know is multi-core server and
13:12: that has a good amount of like
13:15: processing power and that like is
13:16: currently enough. So, but that cost is
13:18: fixed and the server is also doing lots
13:20: of you know other stuff. It's like
13:21: running other services. So, it's harder
13:24: to separate those costs. Um I would say
13:27: like if you like look at the raw numbers
13:29: I would probably say compute is the
13:31: biggest one
13:33: but again like you know it kind of
13:34: depends how you break it down.
13:39: So that's all our questions from
13:42: Discord. uh which means we can get like
13:44: you know also the uh Twitch ones and the
13:47: first one like already kind of started
13:48: like answering um BD is asking IPC
13:53: schnopit so for those of you who don't
13:56: know um what what I've been like working
13:59: on like what what I've been also
14:01: specifically working on pretty heavily
14:03: has been the spliting which is
14:05: separating you know fruxyen from unity
14:08: into two separate processes and having
14:10: those two processes communicate over IPC
14:12: which is interprocess communication. Um,
14:17: and so like I think like you know we
14:19: need like a mechanism for the two
14:21: processes to communicate. So like the
14:23: main engine can send data to Unity uh to
14:26: be rendered with those two being
14:29: completely separate processes. Um and
14:32: for that like we pretty much like
14:34: majorly using shared memory uh because
14:36: it's like really really fast. And then
14:38: there's like a this library called
14:40: interprocess that actually cyro um help
14:43: kind of like prep and like test and fix
14:45: up a bunch of like issues with it uh to
14:47: make it usable especially uh even across
14:50: like different platforms
14:52: and
14:53: um the library like helped quite a bit.
14:55: It has like cool Q mechanism which is
14:58: built on top of shared memory where you
15:00: can put messages into the queue and then
15:02: like them on the other end and it uses
15:04: shared memory. So like you don't um it's
15:08: like really really fast. It also uses
15:09: system level semap force to like signal.
15:12: So you can have like usually one process
15:14: wake up a thread in another process
15:16: which is also like very speedy. Um,
15:21: this Sunday I finally fixed uh one of
15:25: the issues. Um, that that was kind of
15:30: one of those harder ones to diagnose.
15:32: Uh, if you seen the de uh there's been
15:35: this like weird thing that would happen
15:38: where everything like all the particles
15:40: and all the asset updates would just
15:42: freeze out of the everything just stop.
15:46: And I was like, why is that happening?
15:48: And after a bunch of like isolation, I
15:50: managed to find the queueing mechanism.
15:54: It's actually not like it just stops
15:57: like like even if I keep sending
15:59: messages on both ends, they're just not
16:01: coming through. So I was like why is
16:03: that happening? Is there like issue with
16:05: the library itself? Um so I started
16:07: digging through the library code and
16:09: actually found there's like a race
16:10: condition and what it like ended up
16:13: being is
16:16: um so with multi- threading
16:20: one thing that can happen that you have
16:22: to consider is like what operations are
16:24: atomic and which operations are not
16:26: atomic
16:28: because if operation is not atomic the
16:31: operating system can literally even if
16:33: something seems like a small operation
16:35: it can be two different operations at
16:38: the like that are separate and what can
16:41: happen is the timing happens such that
16:44: like the operating system will stop your
16:46: thread literally in middle of that
16:49: operation that you might have thought is
16:51: automic but it's not. So half of it gets
16:53: performed and the other half doesn't and
16:56: then it was another thread around
16:59: which will cause like weirdness
17:01: and specifically what's happening with
17:03: this library uh is when it writes this
17:06: data it writes uh a header for it which
17:09: is two integers and each integer is like
17:12: four bytes. So it writes two integers uh
17:16: which are packed as a strct where one
17:18: integer is a flag saying this data is
17:21: ready to be read. The other integer is
17:23: the size of the message. Um, and if I
17:27: grab a brush.
17:30: So, there's like literally a header.
17:34: There's two integers.
17:37: I'm going to do int
17:41: and int.
17:43: And this is like a strct. And then you
17:46: got like, you know, you got the flag.
17:54: And then you got the size.
18:02: And actually, I'm going to move it just
18:03: a little bit over here.
18:06: So, what normally happens is, you know,
18:09: you have like the writer of the messages
18:12: and you have the reader working with
18:13: this. Actually, it's kind of hard to
18:14: read there. I'm going to move it down.
18:21: So, if like, you know, you've got the
18:23: flag and then got the size and then
18:26: you've got, you know, your thread
18:27: that's, you know, doing stuff and the
18:29: thread like in the code itself, it looks
18:32: like, you know, this happens at once. So
18:34: like you know it runs here and then it's
18:36: done and then the messages are ready to
18:38: be read with like the flexing is ready
18:40: and the size of it.
18:43: However, because these are two integers,
18:46: these are technically two separate
18:48: operations and what would happen now and
18:50: then is the writer would write the first
18:53: integer and then it would freeze like
18:55: you know the operating system pro like
18:58: you know would freeze the thread and it
18:59: would let the reader run. And where did
19:02: my brush go?
19:04: So the reader it would look at this and
19:07: would be hey this says this is a
19:09: readytor you know ready to be read I can
19:12: process the message. So it would read
19:14: the size but because this one hasn't
19:16: finished its job it would read just zero
19:20: or like you know what value there was
19:23: pretty much zero. Um, and now
19:27: you know the message is kind of corrupt
19:29: because like it it didn't read their
19:31: correct size and it would kind of
19:33: explode the cube and things would just
19:35: you know freeze up because like it would
19:36: get into a weird state that it wouldn't
19:38: really properly recover from. Um, there
19:41: was a mechanism that would like auto
19:43: recover things but it would essentially
19:45: miss a messages. So like there was
19:46: probably slight like corruption
19:48: happening as well. Um,
19:52: but pretty much like you know the issue
19:54: was this was like a race condition and
19:56: these are like often times like really
19:58: hard to diagnose, really hard to find.
20:00: Fortunately, this one didn't take me as
20:02: long. Um, as some have like in the past,
20:05: but I still had to kind of like, you
20:06: know, dig a bit to like find this is
20:08: what's happening. Um and pretty much
20:11: like what I had to do is the main change
20:14: to the code is like that before this
20:17: flag gets written the size is actually
20:19: version first. So the size is there and
20:22: then so the size is like you know done
20:25: first it's already staged and only then
20:28: once everything is ready the flag is set
20:30: to like this is ready to be read. So the
20:33: flag is like you know the very last
20:35: thing that happens as the operation
20:38: meaning once this reader is like okay
20:40: because this reader is like you know
20:41: like frequently checking the flag every
20:43: time it tries to read a message it looks
20:45: at the flag it's like is this ready and
20:48: once it's ready you know the size is
20:50: already written and it fixed the problem
20:52: completely
20:54: um but it was definitely one of those
20:56: issues where like you know when it
20:57: started happening I was like this is
20:59: going to be a rabbit hole and it
21:02: actually end up like being a rabbit hole
21:04: in the library itself
21:06: which I'm a little bit surprised like
21:08: nobody has found yet. I think like in
21:10: part it was you know because like like
21:12: for you know the high performance like
21:14: literally have to try like kind of
21:16: hammering it like really hard like you
21:18: know like for the background queue like
21:20: there's just like things being
21:21: constantly written and constantly read
21:24: and it just like end up like hitting
21:25: that race condition frequently. Um, and
21:28: usually like you know when you get a
21:30: library like that you assume like it's
21:31: not going to have issues like that that
21:32: it's like you know it's battle tested
21:34: but in this case it wasn't and it took
21:38: you know I think it took me like 2 days
21:39: to like narrow it down and um then you
21:43: know fix up but now it's fixed so it's
21:46: no longer making me schnop it. It's no
21:48: longer making me disintegrate but I'm
21:51: still a bit surprised it was there.
21:53: [Music]
21:56: Yeah, that's pretty much it. I don't
21:57: know sir if you have any like related to
22:00: this because we also work with library
22:01: quite a bit
22:04: a little bit. there was like some
22:08: weird handling of so so basically what
22:11: uh the library would try to do when like
22:14: let's say you have multiple things
22:17: trying to read one of the memory maps is
22:19: it would uh so it would open the uh the
22:23: memory map file as a uh as like a
22:26: permissive file share mode so like other
22:28: processes could you know read the file
22:31: as well.
22:32: The problem is that it re it relies on
22:36: uh an exception being thrown when you
22:39: try to open a an exclusive file stream
22:42: on the thing. So that's how it checks if
22:45: it's in use is if the file stream does
22:46: an exception. Uh
22:49: that doesn't happen under mono.
22:53: That does not happen because mono
22:54: doesn't implement it that way. And so
22:57: that mechanism is just entirely useless
22:59: under mono especially under Linux. So
23:02: what I had to do was I had to implement
23:04: a secondary file that goes along with
23:06: the memory map but it's also memory
23:09: mapped and then I use interlocked uh
23:12: increment and decrement on it which uh
23:14: makes it a nice atomic counter that I
23:16: can use to determine how many users are
23:18: using the file.
23:20: That was a nice rabbit hole to go down.
23:23: Yeah. And it's unfortunately one of
23:25: those things because you know the big
23:26: part of the split thing is also that
23:28: we're actually you know moving away from
23:31: mono that we don't have to like you know
23:32: worry about it anymore and we have to
23:35: deal with the issues. This is one of the
23:37: few things that where we still need to
23:38: like deal with issue because this is the
23:41: part of the mechanism made for the two
23:43: to communicate.
23:47: So it's um
23:49: yeah but there's only going to be like a
23:52: few now and like most of the code is now
23:53: just going to be like nice and net 9. So
23:55: like that's that's that's going to be
23:57: really nice. But that's that's that's
24:00: for the IP of it.
24:03: Yeah. Like you have like no idea how
24:05: much we are excited for just you know
24:07: the improvement to the tooling and just
24:09: being able to use modern libraries and
24:11: modern language features and stuff like
24:13: it's it's very exciting.
24:15: Um,
24:17: next question. We got the we got typon X
24:20: from Ace on Twitch. Um, actually, wait.
24:23: I'm going to keep the brush around. Um,
24:28: do you have a Tonx?
24:32: I'm I'm actually
24:34: Typox. I don't know how to pronounce it.
24:36: I'm going to say White Tox. You know,
24:38: I'm just going to be lazy again. I'm
24:40: going to make it related to additional
24:41: bit. It's the shared memory. It's such a
24:44: cool concept. Like it's like
24:47: it's literally a mechanism where you
24:50: have two processes share the same chunk
24:52: of physical memory which makes them kind
24:55: of behave as if you know like it makes
24:57: the code not too different
25:00: especially if you structure it right. It
25:02: makes it not too different from just
25:04: working within the same process.
25:07: And that's like super cool that they can
25:09: do that because like it makes
25:10: architectures like this feasible and it
25:13: makes them very speedy because it's it's
25:15: literally like there's no serialization.
25:18: There's no coping like you know like
25:20: from some kind of like communication
25:22: mechanism. You literally write to the
25:24: chunk of memory and the chunk of memory
25:26: is also owned by the other process. So
25:28: like it's the data is just immediately
25:30: there. Um and it allows you know for us
25:34: to like implement this really cool IPC
25:37: architecture that's very very speedy uh
25:40: for like the data exchange because
25:42: literally we have like you know one one
25:44: process stages the data the other one
25:47: just reads it and we don't even need to
25:49: like you know serialize and des
25:50: serialize it because it's it's just we
25:53: just interpret the data like you know
25:55: appropriately
25:56: um and I'm like just l really glad you
26:00: know for this kind of mechanism to
26:01: exist. best.
26:04: Yeah, me too. I'm going to be lazy
26:06: making about that too.
26:09: Thank you. Thank you for the Thanks for
26:10: subscription. Bravo Bravo
26:14: Jules.
26:15: I'm sorry. I don't know how to pronounce
26:17: your name exactly.
26:18: Bravo Jules B34R. Uh, thank you very
26:22: much.
26:22: Thank you.
26:25: Um,
26:27: I'm going to also be lazy and make it
26:29: about uh might as well. I'm actually
26:31: very proud of the
26:33: the uh the atomic like file usage
26:36: capture that I made. So I can just use I
26:39: can just use interlocked increment on
26:41: the memory map.
26:43: That's that's super cool like because
26:45: it's available in C# like you can
26:46: literally use the same mechanisms just
26:48: on the shared memory and also with spans
26:51: like a lot of the code like when I was
26:53: for example making photon dust work like
26:55: with photon does already knew like you
26:57: know we're going to use shared memory to
27:00: put this put the data so the way photon
27:02: dust the way I designed it is so it
27:05: actually all the final data for
27:07: rendering is just put into this big
27:09: memory buffer it's just zero bytes
27:12: And all I have to do is just uh swap the
27:15: mechanism. So instead of just allocating
27:16: array, it actually uses Oh, thank you,
27:18: Jake the Fox for the subscription.
27:22: Get another one.
27:25: Yay. Thank you.
27:29: Um but like like I literally had to just
27:31: replace the allocator.
27:34: So like it allocates it entire memory,
27:36: but most of it it just uses spans. and
27:39: spans they they're like this beautiful
27:41: abscription mechanism where it just it
27:44: doesn't care where the data is. It can
27:46: be managed memory, it can be native
27:47: memory, it can be shared memory which is
27:49: just a form of native memory and like
27:52: pretty much most of the code just stayed
27:53: the same and it just works
27:56: like like like when I go to particles
27:58: like like I remember just being like I
28:00: implemented particles, fixed up a few
28:02: issues and then it just it was just
28:04: working and it was like it just works.
28:07: It's just there the same way, you know,
28:10: it was before, but now it's a completely
28:11: different mechanism. It's going from one
28:13: process to another.
28:15: So,
28:16: so that's um
28:20: that's very it's very cool. I like the
28:22: language and you can use a lot of the
28:23: features like you know regardless where
28:24: the memory is.
28:28: I also realized they might not know what
28:29: I'm saying by interlocked.
28:31: Oh yeah, that's like a Yeah,
28:35: it's it's it's like a nice like set of
28:37: classes for like in C for working like u
28:40: with like doing sort of atomic
28:42: operations and you can like typically
28:44: use them you know on manage memory but
28:46: like it just works the same like
28:47: regardless like even with native memory
28:51: just works DM. So the next question is
28:55: uh from I'm kind of covering up a bit.
28:58: Uh next question is from beauty amazing
29:01: process on the splitting looking forward
29:03: to pre-releases. So far any interesting
29:04: bugs that were accidentally fixed as
29:06: side effect of splittening.
29:10: I don't know if there's been any
29:15: that I noticed. like most of it's been
29:16: kind of focused on just you know fixing
29:19: the bugs of the split caused by the
29:20: split tending just reworking everything.
29:23: Um
29:25: yeah
29:26: um the the hitches from when you first
29:29: like pick up an item from like that boo
29:31: colliders generating legit.
29:34: I haven't tested it yet. I mean I don't
29:36: consider it a bug either. Um, that's
29:39: more just, you know, undesirable
29:42: behavior, but like not a bug. Um,
29:48: yeah, like most of us has like really
29:49: been just focusing on the bugs.
29:53: The um the one where our generic typing
29:56: for like casting an unsigned bite to
29:58: like a signed uh type uh it it like
30:03: ignores like the signing. So if you were
30:04: to cast like a bite to an integer and do
30:06: like bite 255, it would do integer 127.
30:11: I haven't tested it either. Like like
30:12: like literally like so far it's just
30:14: been like focusing and splitting just
30:17: getting things to work and fixing
30:18: whatever bugs come like you know along
30:21: with it. Like there's probably going to
30:23: be like you know stuff like that we
30:24: noticed in the past but like I haven't
30:26: really tested any of it like yet. Um,
30:30: and there's like nothing that like would
30:31: like would I consider like accidental
30:34: fix that like wasn't expected. So, right
30:37: now like I don't have any. We might like
30:39: find stuff, you know, during the
30:40: pre-release, which should be coming
30:41: soon. I was kind of hoping like to make
30:43: one today, but it was like kind of
30:46: today's kind of slow. So, um, I still
30:49: need to like fix up the render to
30:50: texture, some of the issues with that.
30:52: Um, but I think there's a very very good
30:54: chance the pre-release is going to open
30:57: tomorrow.
30:59: um with a caveat, you know, that like um
31:02: it's not going to have everything
31:03: implemented yet. There's still like a
31:05: chunk of things that need to be done. Uh
31:07: but the goal is to start testing as soon
31:10: as possible. Um because like when the
31:13: testing is going um you know, it can
31:16: help us out like by narrowing down
31:17: issues and you know and sort of
31:19: isolating them uh while we still kind of
31:22: work on the other things and it just
31:23: kind of it helps speed things up like
31:25: overall. Uh so I think
31:29: again barring something complex
31:32: happening I think it's going to be
31:33: tomorrow at this point. Um
31:36: [Music]
31:38: so yeah but also like you know a bunch
31:40: of stuff missing like for example the
31:41: pool is not going to have VR support yet
31:43: like it won't be able to run in VR. It's
31:45: going to be desktop only.
31:48: uh which is like you know good enough
31:49: for like a lot of the testing but um
31:52: during testing period I'm going to be
31:54: like implementing the remainder of the
31:56: like things which you can already check
31:57: on GitHub as well. I kind of I spent
32:00: like time like yesterday like organizing
32:02: everything. So uh
32:07: you can kind of see you know what needs
32:09: to be done.
32:11: [Music]
32:13: Uh,
32:16: Jagger Fox Arthur is asking, "Uh, should
32:19: I be concerned about textures? They can
32:20: see me. There's no reason to worry.
32:23: Like, no reason to worry at all.
32:27: Just just don't worry. They just they
32:30: just the textures just like to look like
32:32: you like it look. It's only fair because
32:34: like the entire time you're in VR or in
32:37: Resonite, what do you do? You spend time
32:39: looking at textures. You look at them so
32:42: much. It's only fair that one of them
32:44: will get to look back at you.
32:50: It's only fair.
32:54: Lucon's asking, "Fus, why are you
32:56: yellow?" No, it's No, it's just are you
32:59: yellow? I mean, yes,
33:06: nukun. Sir, are you yellow? Um,
33:09: no.
33:11: Yeah, this this is simple questions.
33:14: Bipolar fuzzy bipolar bear is asking who
33:16: ate FX's mic. Um I don't know. It just
33:20: likes to turn off sometimes.
33:25: Uh Nikon's asking, "What world are you
33:27: in and why are you in here?" Uh so this
33:29: world is called Deep Dreams. It's by the
33:32: team vibes. It's one of their MMC
33:33: entries. Actually, I forget which year
33:35: it is. Uh let me check. Uh 2023. And
33:39: it's a beautiful world. Like it's kind
33:41: of it matches like the you know it's
33:43: really hot over here. Um it's very very
33:47: very pretty world.
33:49: Um and it's kind of like you know kind
33:52: of tropical themed which kind of matches
33:54: the weather. So I was like let's let's
33:56: do this one. I really like the visuals
33:58: of it. Um I was actually I'm going to
34:00: switch it to third person. Uh let's look
34:02: around a little. Uh this might be a
34:05: little bit weird like because I'm kind
34:07: of in full body. just going to be kind
34:08: of floating around. But, uh, you got
34:10: this like, you know, you got this cool
34:13: pool. There's like um,
34:16: let's actually go
34:18: back to spawn. So, as you go through
34:23: and let me adjust the camera a little
34:24: bit more. There's like, you know, this
34:27: nice area,
34:31: which I might like actually end up like
34:33: using some of this to like for future
34:35: streams because like this is its own
34:37: vibe. Like I really like this area. This
34:39: one's super cool. And there's like this
34:42: where we can go through
34:47: and it's this very like it's very dark.
34:49: So maybe not this area, but it's very
34:52: very very pretty. And there's like this
34:55: kind of pool where you can kind of see
34:56: inside and you can actually go
35:00: and know there's like fish that that
35:01: like will react to you as well. I forget
35:04: if they're in here or elsewhere.
35:06: I know there are somewhere.
35:09: Um
35:11: this is really really pretty world.
35:17: as one of the reasons I like is because
35:19: the team vibes like they mix some of the
35:21: most beautiful and graphically polished
35:24: worlds um when there is a night. So I
35:28: think they're also like you know a
35:29: really good showcase of like what you
35:31: can do and the other reason is also I
35:34: was testing this world like earlier uh
35:37: on the splitening build
35:40: where um
35:42: there we go let's get outside. I was
35:44: testing to do this build like you know
35:45: on the splittening build um
35:51: just to see like how it looks. I was
35:53: capturing some screenshots and such uh
35:56: because I think like these worlds
35:57: because they're like so complex like
35:58: they make a good test. There's also like
36:01: this area which I jumped into water. Um
36:08: there we go. Let's go. There's like a
36:10: bar and it's like a whole mechanic where
36:13: you can make like shakes and there's
36:15: like, you know, fruit around. Um,
36:18: probably not going to do it because it
36:19: takes a bit, but uh there's there's
36:20: fruit.
36:23: Oh, and we got we got
36:27: there's fruit
36:29: and you can like make things. It's it's
36:31: very very nice. There's a lot of like
36:32: interactivity to it. Um,
36:35: let's see. Who was that?
36:39: We could um
36:42: Oh, the other scrolled by. What was it?
36:46: Was a raid, but the Oh, it's Sky Wolves.
36:48: Thank you for the raid.
36:51: I'm going to position myself back here.
36:53: But yeah, this is a completely cool
36:54: world. I recommend checking it out. The
36:56: other reason is also like you know like
36:58: I feel we have so many cool worlds on
37:00: there night and like often times people
37:03: will go through them and then they will
37:04: kind of forget and this one's like what
37:07: like 2 years old and like I feel like
37:09: people don't really visit this one
37:11: anymore. Um,
37:13: but like it's still like you know it's
37:15: beautiful world. So I like to kind of go
37:18: through the older ones
37:20: and every every like every resonance you
37:22: know I try to like take a different
37:24: world. So we might go like you know kind
37:26: of back in the history through some of
37:28: them. So uh hope that kind of answers
37:31: that question.
37:35: Let me also position myself back. There
37:38: we go.
37:41: Let's see.
37:45: Next questions from Moon Base. Do you
37:47: like the color blue? I say it kind of
37:50: depends. A lot of colors. It's kind of
37:52: like, you know, what specific shade it
37:54: is and how is it used? Because it's got,
37:56: you know, like the deep blue like this,
37:58: but you also got like, you know, sky
38:00: like this one. So, it kind of depends.
38:02: It could be like a really nice color.
38:05: How about this hair?
38:06: I mean, one of my eyes is blue, so I
38:08: guess I like it. Yes,
38:12: you have like it and have green. It's
38:15: also funny because in in some languages
38:17: like Chinese like they didn't
38:18: differentiate between like blue and
38:20: green for like a really long time. They
38:22: just consider them the same color just
38:24: like different shades.
38:28: [Music]
38:30: Uh next questions from checkbox author
38:34: right now. I'm sorry I got a cover for
38:38: the questions.
38:41: Uh, right now the FPS limits when
38:44: unfocused only works if the game was
38:46: started in the sub mode without VR
38:47: runtime. Is there any chance it could
38:48: also work when the game was started with
38:50: VR runtime post split ending? Probably
38:53: not because like when you run it with
38:54: VR, the VR runtime kind of like takes
38:58: control of the frame pacing. So, um that
39:02: would have to be kind of figured out
39:04: like we have to like I think we would
39:05: have to like disengage the VR runtime
39:07: somehow. Uh and right now like I like we
39:12: could look into it like if there's like
39:13: you know a lot of interest but it's also
39:15: like one of those things where it might
39:16: be kind of harder to do or even
39:17: impossible to do depending on how it how
39:19: the runtimes like end up like working
39:21: out. Uh we might need to like unload VR
39:24: completely which might have like you
39:26: know its own kind of issues. So
39:29: not sure about that one.
39:35: Uh next questions from check the fox out
39:38: there. Also any chance the clipboard
39:40: rework fixed losing alpha channel when
39:42: copying to clipboard. So um what I've
39:46: actually ended up like fixing is um when
39:51: you paste something from clipboard it'll
39:53: use it'll check if there's like fancier
39:55: formats available. um
39:58: and use that which actually makes
40:00: copying images with alpha work like when
40:03: you paste them in. Um
40:06: I tried fixing
40:10: I tried fixing the
40:12: like copying but it's a little bit more
40:15: involved to get working and I was like
40:17: this is not a priority right now. So
40:19: right now it doesn't work right. Uh and
40:22: one of the reasons is like there's like
40:23: actually two APIs. There's like the API
40:25: for just working with bit maps in
40:27: clipboard, but it's actually older and
40:29: like the windows will make bunch of
40:30: conversions, but there's a new one newer
40:33: one uh where you put like more complex
40:36: data objects into the clipboard and it's
40:37: like how of the image pasting works and
40:40: how also like you know other formats
40:41: work. Uh the good thing is like we can
40:44: is easier to work with it right now
40:46: because it's you know just kind of part
40:47: of the process. So I do want to take a
40:50: stab at it or maybe have somebody on the
40:51: team like you know take a look at it and
40:53: figure out like you know the right
40:54: formats and how to like format
40:56: everything uh so it works and we could
40:58: even look into like one of the things I
41:00: would want to get work is for example
41:01: copying and pasting audio. So like you
41:03: know we can just be working with
41:05: audacity and like copy audio and just
41:08: paste it in here and make things like
41:10: you know super simple uh and kind of
41:12: like build build on the you know the
41:14: clipboard interactions. Um but that's
41:16: postpl splithning thing. Um because it
41:19: wasn't unfortunately it wasn't trivial
41:21: to fix and I was like you know this and
41:25: I have like other priorities right now.
41:27: So um I need to like hunt that for later
41:31: but it's going to be easier to fix.
41:33: That's the main part.
41:35: Oh no why does Twitch Why is Twitch
41:38: censoring this one?
41:41: Um I don't I imagine it's just numeric.
41:45: Um why did it censor that? Uh shutting
41:49: here is asking when numeric finally
41:52: brings in complex numbers was the first
41:53: node or component you want to extend
41:55: with them as input. Um I mean I don't
41:58: really have like any specific well
42:01: there's assumptions there. Numeric will
42:03: not automatically give complex numbers.
42:06: um we do want to add complex numbers and
42:08: the numeric will make it easier to do
42:11: that but like you know
42:14: it's not going to happen automatically.
42:15: We actually have to we still have to add
42:17: complex numbers. So like you know that's
42:20: not like the numeric the complex numbers
42:22: are not going to happen automatically
42:23: with the system numeric. Um
42:27: however once we do add complex numbers I
42:31: don't really have like any favorite like
42:33: it's mostly just like adding the
42:35: standard set. So like you know anything
42:38: like any like math notes like they'll
42:41: like you know they'll have like versions
42:43: of it but I don't have any specific one.
42:46: [Music]
42:48: I sorry if you got any thoughts on that.
42:53: M
42:56: I mean
42:58: I don't know. I suppose like
43:03: complex numbers what for specifically
43:06: complex numbers I'm not really sure. I
43:08: mean I guess they'll make
43:10: I guess they'll make certain types of
43:11: math easier I suppose. Um but I don't
43:14: have any specific thing. Yeah,
43:16: I know that I'm pretty sure system
43:18: numeric has like a complex number type,
43:20: but I guess we also need to like
43:22: implement it.
43:24: That's what I said like it's not just
43:26: because we switch like you know use to
43:28: system numeric doesn't mean you get it
43:30: automatically
43:32: but yeah like not really specific ones.
43:34: It's just like nice. It'll be nice like
43:36: having it for like simplifying a lot of
43:38: math. So just you know the general
43:40: support with math knows but um that's
43:42: about it.
43:44: So, Glitch is asking, "Is Mayion an
43:47: instrument?"
43:49: Is that a Spongebob reference?
43:52: Yeah, it's a Spongebob reference.
43:53: No, I know. He's um I'm just I'm just We
43:56: have this thing like where Glitch is
43:58: like
43:59: he's always like there's a Spongebob
44:01: reference and there's like so many
44:02: things that are Spongebob reference that
44:04: I don't know are a Spongebob reference.
44:06: apparently even a lot of stuff in on the
44:08: resume that happens is a Spongebob
44:10: reference and I'm like oblivious to it
44:11: because I never watched Spongebob and
44:14: then glitch made me watch a bunch of
44:16: Spongebob
44:17: so I get some of the references
44:20: and also yes the well you see that's the
44:23: power of VR we can make mayonnaise that
44:27: actually is an instrument on here so the
44:30: answer could very well be yes
44:34: I know if anybody's watching
44:37: Somebody please make a mayonnaise
44:38: instrument.
44:40: Just just make make it happen. Make make
44:43: a mayonnaise instrument so we can then
44:46: be so we can just be you know when
44:50: somebody asks is mayonnaise an
44:51: instrument we can answer yes it is an
44:53: instrument. Here it is. You can play it.
44:59: Please make my an instrument.
45:02: Please don't.
45:03: Please. Yeah. to
45:05: No, just just just if you need
45:07: motivation, think how much Sire will
45:09: suffer.
45:14: I'm going to cry now.
45:15: I don't want to cry.
45:19: Make it making an instrument.
45:21: Think of the memes.
45:28: Uh, next question is from uh, Cos Cosmic
45:33: Waffle. Roughly how long do you estimate
45:35: it will be before Splatooning hits
45:37: stable branch? Uh, this kind of depends,
45:39: you know, how many issues there are, but
45:41: I think like it's very likely going to
45:43: happen by the end of this month. Um, you
45:45: know, things are can always happen so
45:47: you can like, you know, if you run into
45:49: complications, but so far right now
45:52: things feel pretty good. Like I've I've
45:54: run like you know the builds where
45:56: um I pretty much run through a bunch of
45:59: worlds and it would run for like you
46:00: know 40 minutes. I went through the
46:01: entire Black Mesa like right and things
46:04: are pretty stable. There's like some
46:05: issues but like it's gotten to the point
46:09: where
46:12: we'll see. We'll see what happens with
46:13: pre-release testing. But right now I do
46:15: feel like it's going to be by the end of
46:17: the month.
46:21: And like now now now I'm worried because
46:23: you know like I'm saying that like and
46:24: it's like
46:27: you know Morph's law will try to make it
46:30: not so.
46:33: Oh boy.
46:36: And next question is from Jack the Fox
46:38: author. Uh currently it's pretty hard
46:40: for users to identify sources of lag
46:42: because there has only limited profiling
46:44: capabilities. What kinds of thing do you
46:45: have planned to help creators identify
46:46: sources of lag and optimize content
46:48: easier for the future? So Sarah's
46:51: already being activated. Um
46:54: do do you want do do you want to answer
46:56: that that part?
46:58: Do you want to talk about Tracy?
47:00: Yes. I want to put Tracy profiler in FX
47:03: engine really badly. Basic like
47:06: basically you like instrument the
47:09: functions that you want to like profile
47:12: and as you start like instrumenting more
47:14: of your codebase and adding like
47:16: profiling zones it'll uh it'll you can
47:21: connect to the actual profiler app and
47:22: it will start building out like a flame
47:24: graph of like this function and then all
47:27: the functions that make up the time that
47:29: that function took and then like down
47:31: the tree and you can see exactly how
47:34: long something took. And you can even
47:36: mark locks as well, so you can diagnose
47:38: lock contention as like too and see
47:40: which areas of your hood base are being
47:42: like contended for resources and stuff.
47:45: And it's just it's super lightweight and
47:47: it's super cool. And there are ways to
47:49: build it in where it doesn't even have
47:50: any impact on like the the runtime
47:53: performance until like you turn it on.
47:54: And it's it's amazing. I love it. I want
47:57: to put it everywhere in the codebase. I
47:59: want to smear it all over Fruits Engine.
48:01: It is going to be insane. and it's going
48:04: to be cool and it's going to show us
48:06: what the biggest like things uh what the
48:10: biggest things are that are actually
48:12: like hitting the engine and we'll be
48:13: able to use it now because the library
48:16: for it like the C# bindings are only
48:18: like .NET 7 and up. Um, so we can
48:21: actually use that library and we can
48:23: like integrate it and then we can
48:25: profile it and then we can whatever we
48:28: want we can see how what parts of engine
48:30: are the heck and we can do it now and
48:33: it's kind of cool because like it like
48:35: comes like a really cool visualization
48:36: tools as well because like flame graphs
48:38: like a lot of flame flame graphs
48:40: and it's like one of those libraries
48:42: like we really really wanted to use and
48:44: we're not able to you know because it's
48:46: like requires the modern net runtime and
48:49: now we'll be able to um however the
48:53: there's also like other thing because
48:54: like performance profiling you know is
48:56: complex and flame guys develop a lot but
48:58: there's also going to be more tools
48:59: where you can sort of measure you know
49:01: how many assets are on the like you know
49:03: how much texture data is being loaded
49:04: how much of mesh data is being loaded so
49:07: there's also going to be like other
49:08: tools because usually even like for
49:11: myself when I'm like diagnosing
49:12: performance flame graphs are like flame
49:14: graphs are like excellent but also they
49:17: don't always give you the full story for
49:19: things. Um, so there's likely going to
49:21: be more tools as well on on top of that.
49:24: But I think like the Tracy one, it's
49:25: like a good like
49:28: good one like to like even start because
49:29: like it's like existing library.
49:33: You can do fancy stuff like that.
49:36: This is very fancy.
49:39: Yeah. You can like see all of like the
49:41: times that things take up. Like you can
49:43: see like this thing right here like the
49:46: it shows like what functions make up the
49:48: function above it and then the function
49:50: above that and it's I could talk about
49:52: it for hours like oh my god I'm like I'm
49:55: going to like explode. I'm going to
49:57: explode and soon we'll be able to like
49:59: shove it in. We're getting there. This
50:02: is this is one of the reasons it's so
50:04: like exciting for like you know 9. So
50:07: should you give me a second? I need to
50:08: wipe my face because like it's really
50:11: hot and it's itching my forehead and I
50:14: can't
50:16: what?
50:16: Okay,
50:18: Puks is sweaty. He's uh going to
50:21: desatify, I guess.
50:25: [Music]
50:34: There we go. I wish I keep like a a
50:37: small towel nearby because now it's
50:40: getting so hot that like my f face just
50:43: gets sweaty in the VR headset and I just
50:46: need to like take the headset off and
50:47: wipe it for a bit. It's like the worst
50:50: thing is like when it like you know you
50:51: got like sweat like rolling down and you
50:53: can't reach into the headset to scratch
50:55: it and it's like itching it. It's awful.
50:59: Yeah, it's very awful.
51:00: I'm I'm very excited for this like
51:02: having tooling like this. This is going
51:04: to be huge. And it's also like we still
51:07: probably have to figure out how to kind
51:08: of like, you know, expose it for things.
51:10: But like
51:13: once that's done, you know, like that's
51:15: going to be I think excellent tooling.
51:21: [Music]
51:22: Uh, next question is from Rasmos0211.
51:28: Uh, is it possible to run the game and
51:29: headless entire offline? examples like
51:32: setting up a land party of sorts
51:34: demonstration without internet access.
51:36: So right now there's some issues that I
51:38: know like where it won't run offline. Uh
51:41: we kind of want to fix those up at some
51:43: point but I don't think it works right
51:46: now. Um so yeah like right now you
51:50: probably not going to be able to.
51:55: If it's like something you're interested
51:56: in then I think there is an issue on
51:59: GitHub. Um
52:02: there's um
52:05: uh check if get a hard giving up if it
52:08: is something you know like you would
52:09: like to do.
52:11: Uh next question is uh from cosmic
52:14: waffle. Uh also will the splitting
52:16: improve status freezes when users join
52:18: wars and large items of theirs are
52:19: loaded or is it separate issue in my
52:21: experience that the single worst instant
52:23: of B performance currently in Ronite? So
52:25: I'm curious to hear. Um it should help a
52:28: fair bit. Um the problem with like you
52:31: know stutters is there's like not a
52:32: single cause of them. So whatever the co
52:36: like you know whatever the specific
52:37: cause is of a given stutter
52:39: it will depend you know on that. Um I
52:43: did notice like uh you know with net 9 a
52:47: lot of it like a lot of the stutters are
52:48: are either going to be solo like you're
52:50: not going to notice them or they're
52:51: shorter. Uh but there's also like you
52:54: know starters that are for example with
52:56: asset loading where you know they need
52:58: to be uploaded into the GPU and it's
53:01: probably something I think we'll have to
53:03: kind of fine-tune. Uh so it kind of
53:05: balances you know how fast things load
53:07: versus how much frame time they take. Uh
53:10: also like the mechanism that we're using
53:12: right now is actually um it helps with
53:15: loading performance a lot because uh
53:17: previously what would happen is when fru
53:19: engine is you know doing its job unit is
53:24: like not able to do anything um and then
53:28: like once fision finishes its uh you
53:30: know finishes its job it's because like
53:32: you know Unity was actually running the
53:34: fision code then it does you know bit of
53:36: the asset uploading and then it renders
53:38: a frame. What happens with the split
53:40: setting is actually when Unity is like,
53:42: "Okay, I'm ready for the frame. Here's
53:44: the data. Proxension is doing its
53:47: updates while Unity waits on for
53:49: extension to send it a frame, it's going
53:51: to keep uploading assets to, you know,
53:53: memory. So, it's kind of doing that in
53:55: parallel as Fenion is like computing
53:57: everything. And once it's ready with a
53:59: frame, it sends the frame and then unit,
54:01: you know, renders it out. Uh, and that
54:04: like helps a lot with loading speed
54:06: because like now it doesn't have to, you
54:08: know, do it in sequence, but it can be
54:09: doing that in parallel. Uh, but that
54:11: mechanism might require some tuning. Um,
54:14: I think I have to kind of touch it up.
54:15: But overall it depends you know a lot
54:19: like like overall the net 9 it's gonna
54:22: it's it speeds things up
54:25: and also like you know some some uh some
54:27: of the hes were caused like by the
54:28: garbage collector which should be like
54:31: largely gone uh because the garbage
54:33: collector with run time is you know much
54:36: much much more smarter. Um but it
54:39: depends you know there's still going to
54:41: be some hitches. Um and like you know we
54:45: can look like but the good thing is you
54:46: know with the profile and things we'll
54:48: be able to like you know diagnose more
54:49: of them and figure out whatever
54:52: solutions are necessary you know
54:55: it it it's like you know
54:59: there there'll still be some stutters
55:01: like you
55:03: shouldn't expect you know they're just
55:05: going to be gone completely because
55:07: there's always going to be something
55:08: that might end up like you know taking
55:09: too long. what the goal is, you know, is
55:11: like make them lower and also like, you
55:13: know, eliminate the most common ones so
55:15: you don't encounter them as often. Um,
55:17: but I don't think like we will ever be
55:19: able to eliminate every single source of
55:21: stutter.
55:23: Um, you know,
55:26: that could happen because like it can
55:28: happen for lots of reasons and a lot of
55:30: it can also be user content related too.
55:33: And there is one thing I do want to do
55:35: that's already I have again a issue for
55:36: is like if the main engine kind of
55:39: stutters for too long
55:42: uh will actually keep like the previous
55:44: state of the scene meaning like you'll
55:46: be able to like you know look around but
55:48: your avatar and stuff like will be
55:50: frozen uh because it's not going to have
55:52: like new data on it. Uh and that should
55:55: also help a bit like you know with some
55:56: of the stuff especially for example if
55:58: you're like watching a video. So say
56:00: like you're watching a video right now
56:02: if it freezes everything freezes. The
56:03: video freezes you know that kind of
56:05: disrupts you. Um
56:09: with this you know like when it happens
56:11: you will be able to see the video
56:12: continue playing. So that should also
56:15: help like make things feel much better
56:17: but I don't think we'll ever you know
56:20: eliminate like 100% of it. It should
56:22: just get like, you know, a lot better
56:24: across the board, but
56:28: unlikely ever completely gone.
56:34: Uh, Satan uh is asking any question you
56:38: have for chat.
56:41: H, what's going to what will be the
56:45: first thing you're going to do with with
56:46: a split ending? I don't know.
56:50: I I I I I I'll have to think over this.
56:52: I
56:54: [Music]
56:55: s do I have any questions?
57:00: Let me see here. Do I have any
57:02: questions?
57:08: Yeah, I guess mine is like what's what's
57:10: what's what's going to be the first
57:12: thing you want to do thing when it
57:15: releases. That's I mean that's kind of
57:17: like on the top of my head right now. So
57:19: that's kind of like you know what I'm
57:20: asking.
57:22: Cosmic is asking first thing I'll do
57:24: post buttoning is spawn the 1.6 million
57:26: polygon copyart and see if it crashes.
57:29: I mean I don't think it's going to help
57:31: that much with that because it's mostly
57:33: GPU side. Actually it helps with the
57:35: things for the collider.
57:37: Well yeah like if it's a mesh collider.
57:39: Yeah, maybe. I don't know. That's
57:42: Nick's asking.
57:43: First thing that I will do with
57:44: splitting is go through my wars and test
57:46: optimize them in testing phase video.
57:49: First thing after splitting is released,
57:51: benchmark some flux contraptions that
57:52: I've been worried about. Overhead of if
57:55: they're light enough, I'll be using them
57:56: in mod attacks. Nice. That's going to
58:00: make more systems feasible that might
58:01: like, you know, might have been too
58:02: heavy. Actually, I do have another
58:05: question.
58:08: Um,
58:10: shenanigans will happen when the
58:12: spliting happens. Was that is that I
58:14: think that's I think it's ages because
58:16: of the shenanigans work. Then I get like
58:19: a lot of people in our team is
58:20: shenanigans.
58:22: Um, I think that's ages.
58:26: I think that's ages. Is it ages? Is it
58:29: you? Is it you?
58:34: Cuz yeah. Um
58:38: the question I actually have like this
58:40: one's actually like a more practical one
58:42: is like what would um so like one thing
58:46: I kind of feel like we need for resite
58:48: to grow is get more people to talk about
58:51: it on social media because there's like
58:53: a lot of cool stuff that happens there's
58:54: a lot of reasons you know people like
58:57: being on night and they're like having a
58:59: lot of fun but I feel like a lot of
59:01: people that are outside of Resonite they
59:03: don't know about it like you know
59:04: they're like It's almost like it's more
59:06: of a black box to them. And what I feel,
59:10: you know, would help us a lot is like if
59:12: more people kind of shared, you know,
59:15: clips from Resonite, like all the cool
59:16: things on social media to like, you
59:19: know, showcase
59:20: what can be done here, what kind of fun
59:22: you can have. Is there anything that
59:25: would make you want to share it more?
59:28: like like what what could we do on our
59:30: end that would make you like you know
59:32: share more stuff on here and kind of
59:34: help like you know do spread the word of
59:37: like um what's the word like like do
59:40: more like you know word of mouth kind of
59:42: well it's is it word of mouth when it's
59:43: social media
59:45: just kind of like you know essentially
59:46: talk about it more on the social media
59:54: let's see
59:55: yeah that's a good question
59:56: we got Uh
59:59: we also get more questions. Uh not
01:00:02: questions. Uh
01:00:04: as TW I want to invite friends there
01:00:07: post splitending. Oh it's scrolling but
01:00:09: I might wait until post spliting when
01:00:11: book gets updated. Yeah. Uh spawn a
01:00:15: bunch of a a ward hop when it becomes
01:00:18: out of stable. I'm going to try pop into
01:00:20: more public worlds like camp for Friday
01:00:22: and stuff. Oh yeah. Come for come for
01:00:23: Friday. Like I'm looking forward how
01:00:25: that feels like you know when bit better
01:00:27: performance because that usual has a lot
01:00:29: of people looking yes I'll invite more
01:00:31: people have groups that are waiting to
01:00:33: join most is saying yeah try and pile
01:00:35: more people on performance is improved
01:00:37: that is one of the biggest complaints
01:00:38: from arrivals sand makes sense navy
01:00:41: first thing I'll test a guardian oh
01:00:44: that's going to be good one isn't that
01:00:45: one of the actually wait is that that's
01:00:47: one of the Japanese game worlds isn't it
01:00:50: that was a heavy one but cool but heavy
01:00:53: Okay. Uh, cosmic waffle. I had a social
01:00:56: media presence. I bree here and
01:00:58: appreciate it.
01:01:00: Maybe make a social media presence.
01:01:05: How like creator jam and stuff will
01:01:06: feel.
01:01:07: Yeah. Yeah. The one is also I'm really
01:01:09: curious about creator jam. I want to
01:01:12: moon base my also try a few friends who
01:01:14: I've tried before to try again. Cosmic
01:01:15: resing every friend I have about game.
01:01:17: Thank you. Make it easier to record
01:01:20: video for insiders on an official
01:01:21: interactive video camera. Yeah, that's
01:01:23: like this one thing I've been kind of
01:01:25: really thinking about because like
01:01:28: I've really wanted to like you know add
01:01:30: something like that because if it's if
01:01:32: it's as easy to record videos you know
01:01:34: like because sometimes I see like
01:01:36: there's something fun happening I'm like
01:01:37: I wish I could just go like and record
01:01:40: clip share done like I think like making
01:01:44: it simpler like that would definitely
01:01:45: like help and that's like it's been on
01:01:47: my mind quite a bit. Um, Epic is if
01:01:50: you're talking Richard. Uh,
01:01:52: as F public service announcements,
01:01:54: everyone falls. There's an access more
01:01:56: people can play it. Uh, if you want more
01:01:59: people to talk about and I just released
01:02:00: the split ending there. Yeah. Like
01:02:02: that's also like you know what we're
01:02:03: doing. Uh, people are excited so be
01:02:06: talking about it.
01:02:08: Yeah.
01:02:11: Okay. Anyway, that's that's the question
01:02:12: for chat. Like we should do this more. I
01:02:14: kind of want to do more stuff like where
01:02:16: we get where we get to ask you the
01:02:18: questions.
01:02:20: I want rigid body physics.
01:02:22: Oh my god. I want rigid body physics
01:02:24: too. I feel it's going to be another one
01:02:25: of those things that's just going to
01:02:26: lead to infinite shenanigans. that's
01:02:28: going to be like like that's going to
01:02:29: bring it to you know Gary's level kind
01:02:31: of shenanigans where they just have like
01:02:33: things like eating everywhere and like
01:02:34: I'm like somebody will make like they'll
01:02:36: make like you know like a rocket like
01:02:38: you can stick on something that's going
01:02:39: to make it fly around they're going to
01:02:41: make like a spring constraint put it on
01:02:43: their leg and they're going to be
01:02:44: flailing behind it and you know stuff
01:02:46: like that just happen
01:02:48: like like it's that's I feel that's
01:02:50: going to make leggers and I feel like a
01:02:52: very different game for the shenanigans
01:02:53: because like there are a lot of
01:02:55: shenanigans but like Richard B physics
01:02:58: there going to be lots of games because
01:02:59: like there's like there's a whole bunch
01:03:02: of games that are like 90% just physics
01:03:05: and people already making some it's just
01:03:07: that's been kind of difficult to make
01:03:08: them you know like the pool table or the
01:03:10: bowling like Ka made like a really good
01:03:12: one there that's also the bowling world
01:03:13: but there's so many games where it's
01:03:15: just literally
01:03:17: physics like eating things around
01:03:19: vehicles yeah like people have been
01:03:20: doing vehicles but those also kind of
01:03:22: you know more difficult to do um and
01:03:25: when you can just have stuff like
01:03:28: beating around. It's going to be a lot
01:03:30: of fun.
01:03:32: Yeah. It's like one of the hallmarks of
01:03:33: like a game like Gary's pod.
01:03:35: Yeah. You can, you know, if if you were
01:03:37: if you had any more dabs that it's ages
01:03:41: ages activated.
01:03:43: Yes.
01:03:43: Like there's there's literally been like
01:03:45: number of times where like I'm just in a
01:03:47: session and just some vehicle comes in.
01:03:49: I'm like is that ages? And it just comes
01:03:51: out. I'm like it is ages.
01:03:54: this there there's like 50% chance he's
01:03:56: going to arrive you know your session in
01:03:58: some kind of uh vehicle
01:04:03: the next questions is from uh also Sen
01:04:07: ask what's your favorite brush question
01:04:10: to both I kind of used the yellow glow
01:04:13: one because I made it like forever ago
01:04:14: and I just I don't know I kind of
01:04:17: I like it's I I I
01:04:21: feel like I could like make a better one
01:04:22: but also like I don't better enough like
01:04:24: you know this works okay enough
01:04:27: so it's more of like a brush hole
01:04:30: oh that's fun
01:04:33: I remember when I had a convex hole I
01:04:34: was like like I saw it in the tilt brush
01:04:36: and I was like I'm going to steal that
01:04:38: and and I was like I made like a post
01:04:41: about it I was like I stole stole the
01:04:43: thbrush convex hull and it's a little
01:04:46: like my idea was like you know because
01:04:47: people say steal like in sort of tongue
01:04:50: and cheek like mode but some people got
01:04:52: so upset They were like, "Oh, like I see
01:04:54: a lawsuit coming." And I was like,
01:04:57: "What?"
01:04:58: Like, it's it's it's like,
01:05:01: you know, I was like,
01:05:04: it wasn't like actual like, you know,
01:05:06: steal steel. It's like the tongue like,
01:05:08: you know, tongue and cheek when you see
01:05:09: like that's a cool idea. I'm going to,
01:05:11: you know, steal that and they don't
01:05:14: you're not actually stealing it. But I
01:05:15: was I was a little bit taken back by
01:05:19: that.
01:05:21: Um,
01:05:22: you're stealing it.
01:05:23: You're stealing it the same way, you
01:05:25: know, I steal a reality when I 3D scan.
01:05:28: It's just like, you know, tug and cheek.
01:05:32: Yeah, I like the wireframe like convex
01:05:34: holes. It's a free rock generator.
01:05:35: Oh, yeah. You can Yeah, there's like so
01:05:37: much you can do with it. It's just It's
01:05:39: satisfying to work with.
01:05:42: Uh,
01:05:46: next question is from Kite VT. How long
01:05:48: did it take you to stare at the rain in
01:05:50: total for that one delog? Can't remember
01:05:52: which one it was. Yeah, that was
01:05:53: actually the IPC shop. That's how I was
01:05:55: diagnosing it. Like I found the easiest
01:05:58: way to have that happen is just open the
01:06:00: shadow cove and just stare at the rain.
01:06:04: And I was think like overall it took me
01:06:08: I think it's like just total like about
01:06:10: an hour because like when the bug was
01:06:12: still present like it would happen like
01:06:14: in the first 3 minutes or something like
01:06:16: that like the rain would just freeze
01:06:18: like it would be very obvious. Um so I
01:06:22: would like you know like and every time
01:06:23: it happened I would just be like okay
01:06:25: happened shut it down look at the logs
01:06:27: diagnose what's happening. So, I wasn't
01:06:29: staring at like the rain for too long.
01:06:32: And it wasn't until it was actually
01:06:34: fixed
01:06:35: that like, you know, I had like I was
01:06:37: like staring at it for a long time. Like
01:06:38: I literally I had it open for like I
01:06:41: think like 10, 15 minutes, 20 minutes,
01:06:44: something like that. Was just like
01:06:46: staring at it and I was like, "Okay,
01:06:47: it's not happening. I'm just going to
01:06:49: I'm going to record a delog and talk
01:06:50: about it and, you know, use it to stare
01:06:52: at it some more." Um,
01:06:56: so yeah, like the longest the longest
01:06:59: part of that was like once it was once
01:07:00: it was actually fixed is just kind of
01:07:02: start staring at it long enough to be
01:07:04: like, okay, yeah, this is fixed. Like
01:07:05: it's not happening anymore. Uh, and it's
01:07:07: not like, you know, not a fluke that
01:07:09: it's not happening in the first 3
01:07:10: minutes. So
01:07:13: I would say my estimate is like about an
01:07:15: hour.
01:07:19: Um, Jack Fox is asking, "Are you worried
01:07:22: the performance improvis
01:07:24: run too fast and actually make other
01:07:26: social platforms look bad? So, you have
01:07:28: to arbitrary dirt sleeps are slowing
01:07:30: down again, not to make other platforms
01:07:32: look bad. I'm not worried about that.
01:07:35: I'm more worried that like it's not
01:07:37: going to help as much as some people
01:07:39: like will want this to." And right now,
01:07:41: like I don't know how much it's like
01:07:43: going to help like in some cases like
01:07:45: I'm but I'm more worried like I know
01:07:46: that people might be too overhyped.
01:07:49: Um
01:07:51: [Music]
01:07:53: you know like so I'm more worried about
01:07:55: that. I'm I'm more worried like it's not
01:07:57: going to be like that's some like I mean
01:08:01: like give example is another question
01:08:03: about the hitches is because like it's
01:08:06: going to help with hitches like hitches
01:08:08: are going to be better but I'm worried
01:08:10: that some people have you know
01:08:11: expectation that now all hitches will
01:08:14: just magically be gone and there will be
01:08:16: no more hitches ever again and that's
01:08:19: not going to be the case like there's
01:08:21: still going to be hitches like um
01:08:23: there's still going to be things that
01:08:24: will drag the performance down. It's
01:08:26: just like it should help like you know
01:08:28: overall bring everything up
01:08:31: but it's not going to you know it's not
01:08:33: going to make like everything immune to
01:08:34: lack. It's also going to depend on
01:08:36: scenario because in some cases you're
01:08:37: just going to have lots of geometry to
01:08:39: render um and if like if you become GPU
01:08:43: bound you know there's a different thing
01:08:46: and that maybe that's bringing your
01:08:48: performance down. So there's
01:08:52: it's not a universal fix. It's just it
01:08:54: it should help a lot across the board.
01:08:56: It's going to help in some cases more
01:08:57: than in others. Um but it's also not you
01:09:01: know the be all end all of
01:09:03: optimizations. There's going to be more
01:09:05: optimizations we'll do in future. Uh to
01:09:08: make things even faster
01:09:11: and you know smooth out like some of the
01:09:13: hook you know small some of the hitches
01:09:15: and things but performance is a
01:09:18: neverending problem.
01:09:20: So, I'm more like, you know, just like
01:09:24: hoping like all this work like it's
01:09:26: going to be substantial improvement that
01:09:28: and I hope like, you know, people will
01:09:29: be like happy with it. Um, but I'm more
01:09:32: worried like, you know, they'll be like
01:09:35: overhyped and maybe expect more than
01:09:37: it's going to be. So, that's my main
01:09:39: kind of worry with it.
01:09:42: Like if it's if it's like a if it's like
01:09:46: you know super fast like and like better
01:09:48: than like other places then I'm like you
01:09:50: know like that's a win for me.
01:09:54: Uh I mean it's always going to be user
01:09:56: generated content so like
01:09:58: it will always be variable. Um but it
01:10:01: should be it should be much easier to
01:10:04: make worlds that run very well now.
01:10:06: Yeah. It it brings up like you know it
01:10:08: makes your stuff just run like way
01:10:10: faster.
01:10:12: Um, so like you have a lot more headroom
01:10:14: and like you know this it brings up the
01:10:16: ceiling for things but there's still
01:10:18: going to be you know lots of stuff we'll
01:10:20: do like in the future.
01:10:23: It's also like it will probably bring
01:10:25: like other issues because once like you
01:10:27: know we have bigger sessions some other
01:10:28: stuff might become more of a problem is
01:10:30: like networking. So we might need to
01:10:31: like you know implement network
01:10:33: optimizations like mesh networking and
01:10:35: like you know stream calling and other
01:10:37: stuff like that.
01:10:42: Um,
01:10:44: Cosmic Waffle is asking, so I'm just
01:10:46: checking time. Uh,
01:10:50: Cosmic Waffle is asking, uh, what's your
01:10:53: second favorite color? Either or both of
01:10:55: you. I mean, I've got yellow and orange.
01:10:59: It's orange.
01:11:01: I've got blue and green. So, there's
01:11:04: your answer. I would also say if you
01:11:07: wanted from me, if you wanted like one
01:11:08: that's not like, you know, related to
01:11:10: this, I would say purple. I kind of
01:11:12: like, you know, I like combination of
01:11:14: yellow and purple. It looks kind of
01:11:16: cool. Yeah. Kind of like pink, too.
01:11:20: Actually used to like pink like when I
01:11:22: was a kid, like a lot. And my parents
01:11:23: would get me like pink stuff. And I
01:11:25: still kind of like the pink color, but
01:11:27: uh it's no longer my top color.
01:11:30: Strawberry lemonade.
01:11:33: I like them like really hot pink like
01:11:35: the like you know
01:11:43: uh BD is asking uh did you ever confirm
01:11:46: whether the 9 was magically using high
01:11:48: HTTP3 or was it just generally faster
01:11:50: without HTTP3? I haven't confirmed that.
01:11:52: Um I mean it's it's fast enough so I'm
01:11:55: like like you know it's like I'm like I
01:11:58: don't care why is it happening. I'm just
01:11:59: like like it's working fast but like I
01:12:02: happen
01:12:04: it helps. So I'm like it's not a
01:12:08: priority right now.
01:12:12: Like we'll probably check eventually if
01:12:13: it's like using that because like you
01:12:15: know it can be additional performance
01:12:16: gains but like I'm like it's super fast
01:12:19: good good enough for me.
01:12:22: Uh and also for the subscription uh I
01:12:25: don't actually remember the context uh
01:12:27: but uh Bravo Shar is saying symbiotic
01:12:32: interesting I don't remember what that
01:12:34: was in context of
01:12:38: you wanted to like uh we can provide
01:12:40: more context for this
01:12:44: uh chicken fox was also had a
01:12:46: subscription thank you
01:12:49: also made like the the the Twitch thing
01:12:51: you were using which Super cool.
01:12:58: Will uh be the other asking will the
01:13:00: elders textures be implemented in
01:13:02: initial pre-release or will this be only
01:13:04: brief window of time that we can have
01:13:05: solace from them? The elders textures
01:13:08: will come whenever they're ready to see
01:13:10: you.
01:13:16: Platypical 7 uh7744
01:13:19: is asking uh just a fun alter in
01:13:22: history. What if the iicare work was
01:13:23: more popular in the community poll last
01:13:25: year and was made focus? What position
01:13:27: would platform be in now and near
01:13:29: future? I mean that's kind of like I
01:13:33: don't know like this one of those things
01:13:35: is like
01:13:38: what would it take you know for it to be
01:13:40: more popular?
01:13:43: I don't know like I don't I don't know
01:13:46: like you know the hypotheticals because
01:13:48: like there's a lot in that is you know
01:13:51: because like if the I carry work was
01:13:54: more popular across the board what would
01:13:57: lead to that kind of situation to happen
01:13:59: because like you know would people's
01:14:01: mindsets have to be very different for
01:14:03: them to prefer IK over performance or
01:14:06: would it be that performance as was much
01:14:08: better than it you know was in that
01:14:10: author history Um, you know, like like
01:14:14: it always makes me think like what does
01:14:15: this what does this alternate reality
01:14:18: where people
01:14:20: preferred IK more than performance like
01:14:23: look like in that alternate you know if
01:14:26: if we assume that like Resonite is the
01:14:28: same in that alternate you know history
01:14:33: and like you know the performance is the
01:14:34: same way it was when we made that poll
01:14:37: and Ike was the same and people chose
01:14:39: Ike over performance. It means that
01:14:41: people, it's the people in that
01:14:44: alternate history scenario who have a
01:14:46: different mindset who for some reason,
01:14:48: you know, prefer IK over performance
01:14:52: and
01:14:55: you know it's it's it's the question of
01:14:57: like you know how would a different
01:14:59: mindset like how how would a mindset be
01:15:01: different like why did everyone prefer
01:15:04: IK over performance
01:15:08: and that's you know why it's kind of
01:15:09: really hard for me to answer because I'm
01:15:11: like I I don't know what that mindset
01:15:13: like looks like.
01:15:16: I guess the part of it would be that
01:15:18: like we might have like if if people
01:15:20: would have preferred that and like it
01:15:22: would have brought the same amount of
01:15:23: people that performance update will
01:15:25: bring uh we maybe might have been in a
01:15:27: better spot because like I think the I
01:15:31: carry would have been done already like
01:15:34: that one is not as a big of a chunk of
01:15:36: work as performance
01:15:39: um because the performance like that
01:15:41: it's is huge like there's lot of
01:15:43: preparatory work, you know, for things.
01:15:47: So maybe like we would have been like in
01:15:48: a better spot because we would have
01:15:50: gotten there sooner,
01:15:52: but also it's kind of mood because, you
01:15:54: know, it wasn't the thing that most
01:15:56: people like preferred. The biggest part
01:15:59: was the performance. So I guess it would
01:16:03: have made things like it would have made
01:16:06: things easier for us to get into a
01:16:09: better spot where like we removed like
01:16:12: you know major sort of blocker for a lot
01:16:15: of people. Um
01:16:18: so it's um I think that that's probably
01:16:21: the main part is like you know if if
01:16:23: people if the major broker for most
01:16:26: people was okay I we would have sol and
01:16:29: maybe we would be like in a better spot
01:16:31: like we would have like more users uh
01:16:33: would have like more funding by this
01:16:36: point but that's not you know it's not
01:16:38: what it was uh so it's kind of like
01:16:42: wood in a Okay.
01:16:50: Uh, next question.
01:16:54: Uh, let's put this here. And so I'm
01:16:57: going to keep going with these.
01:17:00: I can put it. Um, as on Twitch 17 is
01:17:04: asking, "Last week you talk about images
01:17:06: and having overlay on images that allows
01:17:08: you to share them on social media. A
01:17:10: while back you also talked about posting
01:17:12: system to add more social aspect. This
01:17:13: might be a no-brainer. Will there be a
01:17:15: button most likely the first that allows
01:17:17: you to use that image as a post that
01:17:19: will then let you be able to add caption
01:17:21: or mention other people? I can see this
01:17:23: being a good feature like if you want to
01:17:24: share an event where you want to
01:17:26: understand another stuff. Yeah, that
01:17:28: would be like really cool like like I
01:17:30: want to like I think one of the things
01:17:31: that would help us a lot socially is
01:17:33: like you know um having a system where
01:17:37: you can like you know make arbitrary
01:17:38: posts and the post can be you know
01:17:40: something like I just released a new
01:17:42: tool or I released a new world or
01:17:43: released new gadget or maybe I'm working
01:17:45: on this and you show like you know
01:17:47: progress uh whatever you want to post
01:17:50: and
01:17:52: what we really want to do with that
01:17:54: system is make it so you can easily link
01:17:56: it to other social media like Blue Sky
01:17:58: or YouTube and other things. And if you
01:18:00: make a post, we would have Resonite auto
01:18:03: post it to other platforms and also be
01:18:06: able to do like you know import other
01:18:09: platforms like into like you know post
01:18:10: here. So you can you could use Resonite
01:18:12: as sort of like your hub for social
01:18:14: posting and one of the reasons is
01:18:16: because like I wanted a tool for like
01:18:18: that forever and there's like bunch that
01:18:21: are paid and I'm like I'm not paying you
01:18:23: know however many dollars for like the
01:18:25: service every month. like I'm just like
01:18:28: I feel like it's going to help like you
01:18:29: know help us having that system natively
01:18:31: and we just we build our own um and get
01:18:35: you know maybe it'll get more people to
01:18:37: use Resonite as well like just by making
01:18:39: this more versatile hab which is one of
01:18:42: the big parts of our vision is you know
01:18:43: that the Resonite is this sort of
01:18:46: everything app and you can you know
01:18:48: socialize here you can like you know use
01:18:50: it for work you can use it for education
01:18:52: you can whatever you want to do you can
01:18:55: do it in here. And the other part is
01:18:57: like you know we could integrate that
01:18:59: system with ProRox. So you could like
01:19:01: build stuff like you know social flows
01:19:03: because like one of the things I want to
01:19:05: do like every time for example I start a
01:19:07: stream I want to be I have to post you
01:19:10: know Discord to post there I post on
01:19:13: blue sky I need to post update this
01:19:15: thing and this thing and what I want to
01:19:17: do is just have like a a like a script
01:19:20: that's like a social flow for this thing
01:19:22: like I'm doing a resonance I click
01:19:24: button maybe I fill some basic
01:19:26: information and it makes Discord post it
01:19:28: makes a you know um blue sky post, it
01:19:31: updates to Twitch, you know, maybe it
01:19:33: posts like on Telegram, maybe it posts
01:19:35: like, you know, in other places and it
01:19:37: just does it for me and all I have to do
01:19:38: is just click a button, maybe fill some
01:19:40: basic information that then auto
01:19:42: propagates and I can set up my script to
01:19:44: follow whatever logic I want. Uh and I
01:19:48: think that could be like a really really
01:19:49: powerful tooling and we already have
01:19:51: like you know with our stuff we have
01:19:52: perlux uh so we can kind of build around
01:19:56: that and I feel the more we build around
01:19:58: the tools we have the more powerful we
01:20:00: make them the more appealing they become
01:20:02: to like you know lots of different
01:20:03: people for doing all kinds of stuff and
01:20:07: it's sort of like you know just it kind
01:20:08: of enhances everything and that's like
01:20:11: what I would really want to do is like
01:20:12: you know make it all sort of
01:20:14: interconnected and
01:20:17: powerful, you know, like where you can
01:20:18: kind of interface different things.
01:20:27: Uh, next question is from Septe
01:20:32: Seepkite. I'm sorry, I don't know how to
01:20:34: pronounce that. Um, uh, now that the
01:20:38: splitting is more clear, uh, do you
01:20:40: think performance is going to increase
01:20:42: by two times, three times, or four
01:20:43: times? It's going to depend a lot like
01:20:47: um so the thing with performance is like
01:20:50: you know it's not a single thing it
01:20:52: depends on what you're doing and there's
01:20:54: like some things where there's like 10
01:20:57: times you know increase like for example
01:20:59: particle systems and I've already
01:21:01: actually done a test I've done a test
01:21:02: like where I had like you know like
01:21:04: 30,000 particles all with collisions
01:21:08: and like it was just running super fast
01:21:10: where like you know now with mono it
01:21:12: kind of ends up lagging quite a
01:21:14: So that had like a really substantial
01:21:16: performance boost. Loading is you know
01:21:18: significantly faster like that's
01:21:20: multiple times. Um
01:21:25: it's uh
01:21:27: like you know but there's also going to
01:21:29: be places like where it's not going to
01:21:31: help us much. You know there might be
01:21:33: places where it's just two times for
01:21:34: example but it's hard like we have to
01:21:37: measure it. we have to like you know get
01:21:39: to the point like where we run
01:21:41: benchmarks and there's also some things
01:21:42: we'll need to tune
01:21:44: so like I really don't have like a
01:21:47: specific number for you like that's
01:21:48: something that's going to be
01:21:51: revealed like you know during the
01:21:53: testing phase um and also during the
01:21:56: test phase you know we might need to
01:21:57: tune some things as well so it's going
01:21:59: to change up a bit um but also like you
01:22:03: know it's going to like it's going to
01:22:05: depend on a scenario
01:22:07: So some things will see like higher
01:22:09: boost, some more like lower. It's not
01:22:11: going to be, you know, that everything
01:22:12: it gets the same amount of improvement.
01:22:21: Uh, next question is from Epic Eston.
01:22:25: Uh, I feel like worlds get lost due to
01:22:27: time leaving many cool worlds like this
01:22:29: one basically abandoned. This will be a
01:22:31: real difficult problem to solve, but
01:22:32: there are any plans to potentially solve
01:22:34: this. Um, so I think it's like like a
01:22:37: lot like you know like for example the
01:22:39: social mechanisms I think that could
01:22:41: help with some of that because people
01:22:44: you know if you discover a core world
01:22:46: you can be like you know I found this
01:22:48: cool world and like you post about it
01:22:50: you know and like maybe more people will
01:22:52: see it. Um, we could also add more
01:22:54: mechanisms to promoting world. So, we
01:22:56: can be like, you know, check out this
01:22:58: world. We haven't like people haven't
01:23:00: visited this world in like a long while.
01:23:02: Maybe it's worth a visit. You know,
01:23:04: adding more kind of social aspects where
01:23:07: things can sort of like be promoted and
01:23:09: people can be encouraged to like share
01:23:11: things and explore. Um, I do think like
01:23:13: having some kind of exploration like
01:23:15: system as well will help. Having some
01:23:17: system to rate things, you know, like
01:23:19: for example with a workshop, you'll be
01:23:20: like this is a really good one. Um that
01:23:24: could help like just kind of like
01:23:26: encouraging people to explore. Um
01:23:33: next question is from Cosmic Waffle. Uh
01:23:35: do you have Phantom Sense in VR? And if
01:23:37: so, do you think it affects how you
01:23:39: design things for VR? Uh I don't have
01:23:41: much. I have like bit on my nose uh like
01:23:43: on my face. Um I would kind of want more
01:23:46: but I don't have more than that. Um,
01:23:50: one one of the principles though I think
01:23:53: is like good for designing for VR and
01:23:55: this is something like you know I've had
01:23:56: like for a long time is
01:23:59: design things that like would be
01:24:01: comfortable to do you if if they were
01:24:04: real. Like if it's going to if it will
01:24:06: make you uncomfortable in real life then
01:24:08: it's probably going to make you
01:24:09: uncomfortable in VR. And that's like a
01:24:11: good rule of thumb.
01:24:14: Um once people get in more violics, you
01:24:17: kind of get used to things and you're
01:24:18: going to be comfortable with things you
01:24:20: wouldn't be in real life. Uh but it it
01:24:22: is a good kind of, you know, starting
01:24:24: point for a lot of things. Like if
01:24:26: something has, you know, if you had like
01:24:28: a fast moving things or some interactive
01:24:30: tool and has a lot of sharp edges, you
01:24:32: know, or even spikes, you'll be like,
01:24:34: you know, I don't want to touch that,
01:24:35: you know, I don't want to have that near
01:24:37: my face. I don't want to have that near
01:24:38: my body. Um, you know,
01:24:42: you might want to avoid like those
01:24:44: things for the design and make it like,
01:24:46: you know, like something that be comfy.
01:24:47: Like if you if you made this item as a
01:24:50: real thing, would you be comfortable
01:24:52: like, you know, playing with it?
01:24:58: Um,
01:25:00: Beard of Borg is asking, uh, what would
01:25:03: you say is a good active hangout world?
01:25:06: Um I mean the ones like uh it kind of
01:25:09: depends like there's uh there's a whole
01:25:11: bunch like there's like you know some
01:25:12: events like on Friday but they usually
01:25:14: rotate different worlds. Um ones that
01:25:18: people like to hang out is like very an
01:25:20: island like I've seen bunch of people
01:25:22: kind of hosting you know different
01:25:23: sessions of that one. Uh there's quite a
01:25:26: bunch like I think ultimately depends
01:25:28: you know on the group of people. uh like
01:25:30: I I don't think the world matters
01:25:33: as much.
01:25:36: Um there's al like you know the recent
01:25:38: ones from MMC that have like some really
01:25:40: good social worlds. Uh like this one
01:25:42: like this has been designed as a social
01:25:44: world and it's also kind of cool one. It
01:25:46: depends what you're doing. Like if
01:25:47: you're like just hanging out like you
01:25:48: probably want something you know kind of
01:25:50: more quieter I guess but it depends what
01:25:53: the vibe you want. Like you know there's
01:25:54: wars that are bigger like if you want to
01:25:56: like you know hang out like more small
01:25:58: like more groups. Uh there's ones that
01:26:00: are very cozy that are like you know
01:26:02: small there literally like a room. Um so
01:26:06: there's there's a lot. It depends on
01:26:07: your mood.
01:26:12: Uh next questions from Cosmic Waffle. Uh
01:26:16: Cosmograph is asking is there a
01:26:18: particular point at which you would
01:26:19: consider as a feature complete and would
01:26:21: have nothing else major you want to add
01:26:24: and if so what do you think what do you
01:26:26: think you would do once you reach that
01:26:28: point? I can pretty much straight up say
01:26:31: no. Um Resonite is pretty much like you
01:26:33: know this like project which can just go
01:26:36: on forever. There's always going to be
01:26:39: millions of things to add. Some of them
01:26:41: can be like you know very specific like
01:26:43: you know like earlier today we were like
01:26:45: discussing like what about like set of
01:26:47: procedural meshes for graphs you know so
01:26:50: you can make excel you know in there and
01:26:52: you can make visualizations of things
01:26:54: and you can use it for science and
01:26:56: education so there's like you know stuff
01:26:57: like that we could go even deeper maybe
01:27:00: add a whole bunch of stuff you know say
01:27:02: for astronomy or for like you know
01:27:04: chemistry like adding components you
01:27:06: know for like doing like chemical kind
01:27:09: of like interactions and like you
01:27:11: putting molecules together and doing
01:27:12: stuff like it could get like you know
01:27:14: very specific. Like one thing I would
01:27:16: want to kind of get back to is actually
01:27:17: adding electronic simulation in here
01:27:20: because that's something I kind of
01:27:21: experimented with in the past and that's
01:27:23: kind of fun to me and being able to like
01:27:25: simulate circuits like that would be
01:27:27: super cool. Um,
01:27:29: so there's like
01:27:32: I think there's like, you know, just no
01:27:35: matter how like this goes on, there's
01:27:37: going to be stuff to add and stuff to
01:27:39: build and that's kind of like, you know,
01:27:40: by design. Like this is designed to be
01:27:42: like
01:27:44: you can't do anything kind of app, you
01:27:47: know?
01:27:48: So there I don't think there's ever
01:27:51: going to be a point like, you know,
01:27:52: where I'm like there's nothing more to
01:27:54: add. There's always going to be more.
01:27:57: So, and that's kind of the thing, you
01:27:59: know, it's like one of the points of
01:28:01: Resonate, one of the reasons I made it
01:28:02: is because I wanted to like be like I
01:28:04: want to all the cool stuff I ever want
01:28:06: to be working on. I just want to make it
01:28:09: part of the same shared universe. And
01:28:11: that's like, you know, the big kind of
01:28:14: fundamental idea behind this is like
01:28:16: everything I could want to work on. Even
01:28:18: there's like games I want to build.
01:28:20: Like, I haven't done game design in so
01:28:23: long, but it's kind of how it started.
01:28:25: Um, and there's like games and stories I
01:28:28: want to build and I would want to build
01:28:30: them on here.
01:28:32: Um, you know, so that's like that's
01:28:35: that's another thing. So there's so much
01:28:38: to do and everything everything I kind
01:28:40: of want to do I want to you know make it
01:28:41: part of Resonite because the more stuff
01:28:44: is added to Resonite
01:28:46: the more stuff is added in the CS engine
01:28:50: the more powerful this place becomes
01:28:51: because whatever is added now works with
01:28:54: everything else it's part of the same
01:28:56: universe and it just makes the universe
01:28:57: like you know richer and more more
01:29:00: capable.
01:29:06: Uh, Cosmic Waffle is uh asking, "Also,
01:29:08: the character controller component
01:29:09: mentions prepared body support is
01:29:11: planned. Is that on the radar somewhat
01:29:13: soon or more of a long-term goal?" Um,
01:29:16: depends what you mean by soon. Um,
01:29:20: I would say there is non zero chance it
01:29:23: might happen this year, but also maybe
01:29:26: not. So, like I don't know. It depends
01:29:29: whatever the prices end up like being
01:29:31: and how it ends up like shaking out. But
01:29:32: like I haven't like really thought ahead
01:29:35: like for some of the things. So I'll say
01:29:38: maybe no, not this year. I don't know.
01:29:41: We'll see. We'll see. Like usually like
01:29:43: once I'm kind of like finished with like
01:29:45: a big task that's kind of like where
01:29:47: things become a little bit clearer or
01:29:49: what the next goal is going to be. Um
01:29:54: but yeah, like I would say kind of more
01:29:57: long term I guess. We'll we'll see.
01:30:02: like I haven't I haven't really made up
01:30:03: my mind.
01:30:06: Uh Resonator is asking is the other
01:30:09: texture the secret to the splitting
01:30:11: ritual?
01:30:15: I think that's how you pronounce that.
01:30:18: Yes, I think I think you got the dialect
01:30:20: right.
01:30:23: Also, uh dancer is uh
01:30:29: that's what he said.
01:30:34: Uh next question is from cosmic buff.
01:30:38: Base game matrix support when uh why
01:30:40: isn't currently supported? Um
01:30:43: what
01:30:44: you can use
01:30:44: matrices.
01:30:46: We have matrices.
01:30:47: you might need to elaborate a little bit
01:30:50: because like they're like we do support
01:30:52: matrices they're like fixed size. I
01:30:54: don't know if you're asking for like
01:30:57: you know variable size matrices but you
01:31:00: we do have matrices you can do matrix
01:31:03: math
01:31:07: I would say like um give us a lot of
01:31:09: more context on that question.
01:31:14: So generally the answer for why isn't
01:31:18: something currently supported is because
01:31:20: like we haven't prioritized it yet.
01:31:24: Yeah.
01:31:27: Uh next question is from BD if you add
01:31:30: Tracy profile to FX engine will it be
01:31:32: available for users to use profile their
01:31:34: flux contraptions and mods? And I mean I
01:31:37: figure it would. Um I guess the main
01:31:39: thing is you know like because we made
01:31:41: sure like there's like not really an
01:31:42: overhead until it's used. So like if
01:31:44: that's the case like we could probably
01:31:45: just have it in the base and then like
01:31:48: you know be like use this tool but
01:31:51: there's a there's a way you can
01:31:53: structure it that will like make it so
01:31:55: that uh the code for the profiling will
01:31:58: beed until uh you enable like a flag or
01:32:02: something.
01:32:03: Yeah. Well it might require a
01:32:04: compilation maybe.
01:32:07: I mean worst case like you know we'll
01:32:09: have like I think we can have like you
01:32:12: know two builds one that's instrumented
01:32:14: ones that's not where it's like compiled
01:32:16: away and if you want to do profiling
01:32:18: we'll be like okay run this version um
01:32:21: connect this tool you know and it's like
01:32:22: available but I feel we will be we will
01:32:25: make it available somehow.
01:32:28: Yeah like it's it'll it'll
01:32:31: it will be the trace profiler. I don't
01:32:33: know if there's anything specific in
01:32:34: regards to that. Yeah, I don't know what
01:32:37: you're Yeah.
01:32:40: [Music]
01:32:44: Uh, Nikon's asking, I know you are
01:32:46: interested in worlds which connect to
01:32:47: each other and I have been working on
01:32:48: one Jupyter station with its airlocks
01:32:50: and space fairies. How do you recommend
01:32:52: people do things like that? What do you
01:32:54: like to see in those sorts of
01:32:55: transitions? What sort of features would
01:32:57: you like to implement for it in the
01:32:58: future? I mean,
01:33:01: I don't really know. Uh,
01:33:03: it kind of depends how you structure
01:33:05: things. I don't have like any specific
01:33:07: things for that.
01:33:09: [Music]
01:33:16: Uh, next question is also from Nikon.
01:33:19: Uh,
01:33:22: uh, I may ask you before, but what is
01:33:24: would you like to see in sci-fi setting?
01:33:26: What kind of sci-fi worlds excite you?
01:33:27: What sort of sci-fi content would you
01:33:29: like to see more of? I really like like
01:33:31: hard sci-fi stuff that sort of like
01:33:33: makes that's very far future kind of
01:33:36: thing like like stuff that has to do
01:33:37: with mind upload uh you know similar to
01:33:40: the realities the also that kind of
01:33:42: somehow messes with like the
01:33:44: fundamentals of laws of physics uh like
01:33:46: for example a lot of novels by Greg Egan
01:33:49: like that's like some of my favorite
01:33:50: sci-fi I would love to see more of that
01:33:52: kind of thing um the expans also like a
01:33:56: great series I don't want to give too
01:33:57: many spoilers but some of the stuff that
01:33:59: happens especially in the last three
01:34:01: books. Like I really love that kind of
01:34:03: stuff.
01:34:05: [Music]
01:34:07: Sorry about like anything on this one.
01:34:11: I like um uh Foundation.
01:34:15: Oh, Foundation's also good. Oh, there's
01:34:17: a new season coming out.
01:34:19: Yeah, I like the black hole ships. Those
01:34:21: are really cool.
01:34:25: Also, I wanted to like speed through the
01:34:26: questions a bit because like we have 25
01:34:28: minutes left and there's still like a
01:34:30: quite bit. Um, beers is asking
01:34:34: summernit. Yeah, it's it's very hard.
01:34:37: Uh, is face with your summer snapp kind
01:34:40: of. Yeah, like I like it's it's horrible
01:34:42: just feeling it on your face and it's
01:34:44: like you can't stretch yourself.
01:34:48: Um,
01:34:51: anyway, what's your favorite board game?
01:34:53: I mean, I don't play it as a board game
01:34:56: usually. Uh, but it is a board game.
01:34:58: It's blood on clock as the social
01:35:00: deduction game. So, I would say that
01:35:06: uh I don't think I have a favorite board
01:35:07: game. At least I don't have any that I
01:35:10: play recently.
01:35:12: There's also one I don't know. Does it
01:35:13: count as a board game? This card spins a
01:35:15: human.
01:35:18: I guess you could consider that like the
01:35:21: same realm. It's It's like a It's like a
01:35:23: sit down and play it kind of game around
01:35:25: a table.
01:35:26: I like that one, too. And there's
01:35:27: actually sessions like every Wednesday.
01:35:29: That's quite fun.
01:35:33: So, what's your least favorite board
01:35:35: game? I don't know that many board
01:35:36: games. So, like I don't really have one
01:35:39: that'd be like least favorite.
01:35:41: Yeah.
01:35:47: Yes. Both of you tell us about Cy games.
01:35:49: What do you have? What do you hate? I
01:35:51: mean, we cannot just answer that.
01:35:53: We just didn't.
01:35:56: Uh, okay. Volops. I'm curious. I don't
01:35:59: remember if this was answered. What is
01:36:00: the next biggest goal after a split
01:36:02: thing? Is it like tons of bug fixes,
01:36:03: working on your own render, or just
01:36:05: implementing tons of features? The next
01:36:07: haven't really been decided yet. There's
01:36:08: like a bunch of things that are floating
01:36:10: around, but like right now the focus is
01:36:12: spliting, spliting, splitting. So, I'm
01:36:14: like that's that's my brain is full of
01:36:16: that and like I don't have too much
01:36:18: space for deciding on what the next gun
01:36:20: is going to be.
01:36:23: Uh, okay. Come on. First thing I Oh,
01:36:25: this is when we asked the question. Oh,
01:36:27: boy. We're kind of behind.
01:36:30: Um,
01:36:33: oh, they're talking about uh
01:36:35: Yeah, that's when we asked the question.
01:36:38: Uh, yeah, we talked about that. We
01:36:41: talked about that.
01:36:44: questions. If you were talking within
01:36:45: the chat where please ask like, you
01:36:47: know,
01:36:50: Yeah, please don't do the thing where
01:36:51: you add like a question mark at the
01:36:52: beginning of your thing and then just
01:36:54: say something because that like clogs up
01:36:56: the question cue.
01:36:57: Yeah, but we don't know like how much is
01:36:59: uh
01:37:01: uh next questions from Nikon. How well
01:37:05: will rigid body work over slow
01:37:06: connections? Uh that was my main issue
01:37:09: with my flying machine engine. So the
01:37:12: big part like the biggest part of rigid
01:37:14: body physics is actually the networking
01:37:16: because we have to make it like be
01:37:18: pretty efficient and it can be made like
01:37:20: pretty efficient but it also is going to
01:37:22: depend how many rigid bodies you
01:37:23: actually have. Um
01:37:27: we can have
01:37:30: you know
01:37:32: maybe have it like adapt so like it's
01:37:34: going to be able to scale and you're
01:37:36: going to maybe have more like glitches
01:37:37: with the physics. So it's going to
01:37:39: depend but uh um like that you know the
01:37:43: specific answer is going to depend once
01:37:45: it's actually implemented.
01:37:47: We have to test it.
01:37:50: Uh as on Twitch 17 can we have next
01:37:52: week's residence on net 9. Uh that
01:37:55: depends on how much implementation is
01:37:57: done over the next week. So like I don't
01:37:59: know for sure yet. If if it works out
01:38:01: then yes but we'll see.
01:38:04: We'll see like what state is it in.
01:38:07: We shall see.
01:38:10: Um, next question is from Moonbase. In
01:38:12: response to concerns on performance and
01:38:13: improving as much as people would want,
01:38:15: do you still have data model work which
01:38:17: will bring more optimizations and
01:38:18: performance? I mean, yes, there's more
01:38:20: stuff we have to do, but also like I'm
01:38:23: in a state where I'm kind of like we
01:38:25: need to do some other student
01:38:26: performance right now and we've at this
01:38:29: for a while. So I hope like you know
01:38:31: this is the final phase of this push and
01:38:35: that it's going to be like enough
01:38:37: um because there's more stuff we want to
01:38:40: do in the future but also like you know
01:38:42: I don't want to focus on other stuff for
01:38:44: for a bit before going back to
01:38:46: performance
01:38:49: and that's kind of you know the part is
01:38:50: like you know I don't want to like end
01:38:53: up being like okay like we need to focus
01:38:54: on performance for like you know next
01:38:57: several months more.
01:39:00: Yeah.
01:39:03: Um I was getting like a little bit like
01:39:05: worn out with it.
01:39:08: Um and I guess like you know that's the
01:39:10: part is like you know where I'm like
01:39:11: worried is like
01:39:15: that like you know people are not going
01:39:17: to like let me kind of like you know
01:39:19: focus on more things like that people
01:39:21: want like you know more performance
01:39:23: work.
01:39:24: But I do think it's going to be like
01:39:26: substantial. So I think like it will be
01:39:28: good to like move to other things but
01:39:31: that is my worry you know that like
01:39:33: people will be like no this is not
01:39:34: enough we need more
01:39:37: and will like spend like more time and
01:39:41: we'll see um
01:39:44: next questions from stage dub uh I mean
01:39:48: it's not often now I think but I
01:39:50: remember seeing the streams that the the
01:39:52: session was getting heavy dynamics
01:39:54: started flopping around very fast why
01:39:55: does that happen any tips on avoiding
01:39:57: that um it pretty much happens because
01:39:59: the simulation time the simulation steps
01:40:01: become so big the simulation becomes
01:40:04: unstable because like um the simulation
01:40:07: essentially compensates based on the
01:40:08: frame rate and this simulation step is
01:40:11: too long it essentially moves too fast
01:40:13: so it kind of overshoots and then
01:40:15: overshoots back and back and forth and
01:40:18: it just becomes unstable. Um, it there
01:40:20: was supposed to be a system that
01:40:21: prevents that from happening by limiting
01:40:23: the simulation time step, but it's not
01:40:25: working right. I haven't really focused
01:40:28: on fixing that because like main focus
01:40:31: is just improving the performance and
01:40:33: making sure it doesn't get to that
01:40:34: state. Uh, but it's something, you know,
01:40:36: like we might like touch again at some
01:40:38: point depending.
01:40:43: Uh again, if you if you're having a
01:40:44: conversation in the chat, please like
01:40:46: don't include like the question mark so
01:40:49: it doesn't like you know pull out chat
01:40:51: with it.
01:40:55: Uh
01:40:57: question from epic east question about
01:40:58: social posting system but is also
01:41:00: support software must shy peer tube go
01:41:03: to social atma etc. Yeah, like the goal
01:41:06: would be pretty much like you know have
01:41:08: support for as many platforms as we can
01:41:10: because the more it supports the more
01:41:12: powerful it is and it's pretty much you
01:41:14: know what I would want is like you know
01:41:16: I make a single post and it just posted
01:41:17: like you know to everything I can be
01:41:20: like have presence on lots of like media
01:41:22: and have like you know there's an IB the
01:41:24: hub of it and the more we support the
01:41:26: better it is.
01:41:30: Uh, Protrusion Flex is asking, uh, we've
01:41:33: seen how much Net 9 speeds up word
01:41:34: loading, but how notable do you think
01:41:36: will improve harmon downloads will be
01:41:38: for things like inventory, contacts,
01:41:40: world browser, all the things that
01:41:42: require a lot of individual small asset
01:41:43: downloads? I think pretty much the same.
01:41:45: Like all of that is using the same
01:41:47: system. So, and when I've been like
01:41:50: using it, it does feel snappier. So, I
01:41:53: think like, you know, it's just it's
01:41:54: just improvement across the board. uh
01:41:57: and those like I said like those systems
01:41:59: they're not using you know anything
01:42:00: different. So
01:42:03: um it's pretty much like you know the
01:42:05: same improvement
01:42:08: which also like one of the benefits you
01:42:10: know of like like having unified system
01:42:12: for handling this stuff because if if
01:42:14: the unified system you know gets
01:42:16: improvement then everything using it
01:42:18: gets the improvement as well.
01:42:21: Uh there's a PSA support group. Uh, I
01:42:25: guess we can like bring this one up. Um,
01:42:27: they g us a PSA, but uh it's also been
01:42:31: in the in the chat.
01:42:33: [Music]
01:42:37: Uh, this is actually hard sci-fi
01:42:40: physics. Okay.
01:42:42: Um,
01:42:44: all the questions now. Okay. So, we
01:42:46: sorted through them. We've got uh
01:42:49: 17 minutes left. So, if you got any more
01:42:51: questions, feel free to ask. Um, I would
01:42:54: probably suggest not like not like long
01:42:57: like like super longwitted ones at this
01:42:58: point because like we're, you know, near
01:43:00: the end of the stream. Um, so we might
01:43:03: not be able to get into depth of some
01:43:05: things, but uh, if they're like long
01:43:09: ones, we might not be able to get to all
01:43:10: the questions, but uh, you can still
01:43:13: ask. Still got like, you know, time.
01:43:21: Cosmic Waffle is asking and are you
01:43:24: interesting working on the splitening? I
01:43:26: mean both yes and no.
01:43:30: It it is a lot of it is kind of fun like
01:43:33: you know being able to finally work on
01:43:35: this because I've been imagining working
01:43:38: on this for years at this point like
01:43:40: I've been like I want to get to this
01:43:42: point where we get to use all these cool
01:43:43: tools and just feel it make you know
01:43:47: everything kind of feel better across
01:43:49: the board. Um, and it's been a lot of
01:43:52: preparatory work. Like I've literally
01:43:55: been thinking about this
01:43:58: for years and now to finally be able to
01:44:01: work on it. I'm like like you know it's
01:44:03: here. It's happening. And then I have
01:44:04: the build like if you imagine like the
01:44:06: build like how it's going to feel like
01:44:09: you know having the build actually run
01:44:11: on net 9 and have it render across IPC
01:44:13: mechanism and have all of it working
01:44:16: like I was like looking forward to that
01:44:19: like you know for years and now like I
01:44:21: actually have it working on my machine
01:44:23: and like sometimes I'm work playing with
01:44:25: the build and I'm like this is real now
01:44:28: like I'm literally working on it right
01:44:30: now. I'm making it happen is the thing
01:44:33: is the thing I wanted to work on. I
01:44:35: wanted to make happen for so long and
01:44:37: it's now I'm like I'm finally at the
01:44:40: point I'm finally done all the peritive
01:44:44: work, you know. I'm done with photos.
01:44:45: I'm done with audio. I I'm finally
01:44:48: working on this like you know big goal
01:44:51: of the Switch and it's going to make
01:44:53: things feel much better going forward.
01:44:55: So there's that element to it like you
01:44:58: know where I'm like it's happening.
01:45:01: I wanted this to happen for so long. The
01:45:04: other side of it though is because I'm
01:45:06: like I need this to happen as fast as
01:45:09: possible like you know we want to do a
01:45:11: big marketing push. Uh we like you know
01:45:14: have a lot of depending on this. We want
01:45:15: to like you know we want to have like
01:45:18: you know
01:45:19: a lot more kind of community events. you
01:45:21: know,
01:45:23: I need it I need to make it happen as
01:45:25: fast as possible. So, it's also kind of
01:45:27: a crunch which becomes like tiring and
01:45:32: you know wears me out. So, it's kind of
01:45:34: like and also I don't have as much time
01:45:36: like you know I wanted to play some
01:45:38: games. I don't have time for them right
01:45:40: now. Um I wanted to like you know spend
01:45:44: time like doing other things but I don't
01:45:45: have that time right now because it's
01:45:47: just crunch is largely crunching.
01:45:51: So, and then there's like you know the
01:45:53: aspect the worries where it's like what
01:45:56: if it's not going to be enough for
01:45:57: people like you know like what if all
01:45:59: this work will be for you know
01:46:02: and people will be unhappy with it. Um
01:46:05: and that's like you know brings some
01:46:07: anxiety to it and then the other part is
01:46:09: you know what if what if we do all of
01:46:11: this all of this work and then people
01:46:14: aren't going to care like you know maybe
01:46:17: people will not come back to the
01:46:19: platform and you know
01:46:22: all of this like you know which we were
01:46:24: like thinking you know this is like
01:46:26: people said a lot of people were like
01:46:28: you know this is one of the major
01:46:29: reasons we don't play is the
01:46:32: performance. So our thinking is
01:46:36: this is going to help us bring more
01:46:38: people. It's going to bring more
01:46:39: support. It's going to bring more
01:46:40: funding. But then there's also the
01:46:42: anxiety of like you know being like what
01:46:44: if it doesn't? And I think like in big
01:46:47: part is kind of irrational because like
01:46:50: you know the data we have like you know
01:46:52: switches otherwise the data we have with
01:46:55: the headless when we did the net 9
01:46:58: switch went headless. Well, it was
01:47:00: actually net 8 back then, but when we
01:47:02: did a switch, we saw substantial
01:47:04: performance improvement across the
01:47:06: board. And that was the reason we, you
01:47:08: know, like I wanted to do it for
01:47:10: headless first because it runs pretty
01:47:12: much the same code except you know the
01:47:14: rendering bits
01:47:16: and I was like, you know, how
01:47:19: much is it going to help? and it showed
01:47:21: it helps a lot and that's like a strong
01:47:24: data point is you know showing okay this
01:47:27: runtime actually helps run FRS engine
01:47:29: way faster.
01:47:32: So there's data there, you know, that's
01:47:33: showing like it should make things way
01:47:35: better.
01:47:37: And same, you know, with the survey, we
01:47:39: ask people
01:47:41: what's the main reason you don't play,
01:47:42: you know, or what would you make what
01:47:44: would make you play more? And a lot of
01:47:46: people said is the performance and
01:47:48: people said it across different
01:47:50: demographics. Like people who play on it
01:47:52: a lot had the same answer. People who
01:47:55: play on it regularly had the same
01:47:56: answer. People who don't play had the
01:48:00: same answer. All of them said they will
01:48:02: play more or they will want to play once
01:48:04: the performance is better. We hear it
01:48:06: from a lot of people too like you know
01:48:07: where you say you know like you're going
01:48:09: to bring more people onto the platform
01:48:11: once the performance is better. Um we
01:48:14: see a lot of comments where people are
01:48:15: like I don't play because you know my
01:48:17: passion cannot handle it or because the
01:48:19: performance or the frame rate is really
01:48:20: low. So a lot of the data is like you
01:48:23: know saying
01:48:26: it should help a lot like so like you
01:48:29: know the rational part of me is like
01:48:31: like this is the right thing to do. We
01:48:33: should be like working on this. This is
01:48:34: the best approach we can take.
01:48:38: It should help a lot with performance.
01:48:40: It should help bring a lot more people.
01:48:41: It should bring a lot more support. But
01:48:43: then the irrational part of the brain is
01:48:45: like what if it doesn't? And that's kind
01:48:49: of, you know, hard to deal with. And the
01:48:52: closer
01:48:54: the closer I get, you know, to having
01:48:57: the split ending done or even having the
01:48:59: pre-release, the more I'm like, now it's
01:49:02: going to, you know, now it's like the
01:49:04: moment of truth. Like what what is not
01:49:06: going to be enough? And the more the
01:49:08: closer it gets, you know, the more
01:49:09: anxious I get. So like it's kind of this
01:49:10: weird case where like you know I'm both
01:49:14: like getting excited and also getting
01:49:16: more anxious as it gets kind of closer.
01:49:18: So like
01:49:21: yes I kind of I both enjoy it and I also
01:49:23: don't enjoy it and it's kind of you know
01:49:25: conflicting in my brain with that.
01:49:30: [Music]
01:49:38: Uh K WS is asking why does there's a
01:49:41: night save world if you close program
01:49:43: externally I find it pretty annoying. Uh
01:49:45: the main reason is like we want to
01:49:48: prevent data loss when possible. uh we
01:49:50: had actually it like with early versions
01:49:53: of engine it didn't used to do that and
01:49:55: what would happen is
01:49:58: people would like you know happen to
01:50:00: close it sometimes by accident and they
01:50:02: would lose all their changes.
01:50:04: Um so figured like you know it's better
01:50:07: to like save things rather than not.
01:50:11: uh because like
01:50:15: data loss you know is like something
01:50:17: like we want to avoid whenever possible
01:50:19: and like I get like you know worst cases
01:50:20: where you don't want to like save it but
01:50:23: you know the flip side is people will
01:50:24: lose their data and I don't think that's
01:50:27: a that's a good tradeoff
01:50:32: would ideally make it like potentially
01:50:33: configurable so you could like request
01:50:35: you know a toggle for it I think it
01:50:37: might be kind of best solution there uh
01:50:40: but We are not going to like I don't
01:50:42: think we will change the default because
01:50:44: the default is there to prevent data
01:50:46: loss.
01:50:50: Uh Vex 120 is asking are there any
01:50:54: updates planned for in-game tutorials to
01:50:56: help guide new players in a better
01:50:58: creative flow? Yes. Uh the uh the art
01:51:01: team they're actually working on a new
01:51:02: help tab and they showcase some of it on
01:51:05: the last uh recap stream we had this
01:51:07: Friday. Uh so that should help like
01:51:10: people a lot. Um if you want to see more
01:51:12: uh I recommend checking out their office
01:51:14: hours uh on Wednesday. Uh they might be
01:51:17: able to kind of show you a little bit
01:51:18: more about it.
01:51:22: So N is saying you can do its thank you
01:51:29: which we believe in FS. Thank you.
01:51:33: [Music]
01:51:34: Thank you. You guys are sweet.
01:51:37: Yeah. Yeah, I just saw.
01:51:40: Never mind. That's unrelated.
01:51:44: Thank you. I mean, for me, it's like not
01:51:46: like, you know, a thing of killing
01:51:49: myself. It's more like
01:51:53: like it's more like I'm going to like,
01:51:57: you know, work through it and get
01:51:58: through it.
01:52:00: But I'm like,
01:52:02: I hope like it works out. And like, you
01:52:05: know, like I don't have like every
01:52:06: reason to believe it will, but my brain
01:52:09: is like, you know, like
01:52:11: if there's like 0.1% chance something
01:52:14: will go wrong. I'll be like
01:52:17: I be like like I'll be like 90% focused
01:52:20: on that 0.1% and it's like annoying and
01:52:24: makes things hard, but that's what it
01:52:26: is.
01:52:28: Uh, we got 7 minutes left, so we still
01:52:31: got time for a few questions. Uh,
01:52:33: probably short ones at this point. We
01:52:36: might also like end a few minutes early
01:52:37: if it's um if there's not too much.
01:52:41: I need to get some water.
01:52:44: I might have some water here.
01:52:47: I have water here, but it's not the VR
01:52:49: friendly bottle, so I'm not going to do
01:52:51: it.
01:52:53: I'm actually Why does my necklace have a
01:52:55: G on it?
01:52:57: Uh, well, I mean, first of all, it's a
01:52:59: collar. Um, it's a necklace. Uh, but the
01:53:02: reason it has a G on it. Um, is because
01:53:05: it's from, uh, I used to play Gary's mod
01:53:08: a lot back in the day. Uh, that's where
01:53:10: I got a lot of like my, uh, programming
01:53:12: knowledge and intuition from. It's where
01:53:15: I learned math. I literally learned
01:53:17: trigonometry with Dark RP because I
01:53:19: wanted to make things fly around in my
01:53:21: head or fly around my head in a circle.
01:53:23: That's how I learned trigonometry. Um,
01:53:26: which is kind of funny. Um,
01:53:30: yeah. I mean, like that's pretty much
01:53:32: it. I played like tw,000 hours for Gmod
01:53:35: or something crazy like that. Uh,
01:53:39: so it's I it's it's there to stay
01:53:42: probably.
01:53:44: It's a permanent edition.
01:53:52: Yes. Expression two.
01:53:56: The next question is from AS on Twitch
01:53:58: 17. What's your favorite mango?
01:54:01: I
01:54:04: an edible one. The perfectly ripe one
01:54:08: or maybe in the form of ice cream or a
01:54:10: smoothie
01:54:12: actually
01:54:15: sherber. I don't I don't have a
01:54:17: sherbber. I may not have mine going in
01:54:19: lots of forms.
01:54:21: Freeze dy is good too in chocolate.
01:54:27: Not a question, but BL mentions the
01:54:30: amount of transparency and communication
01:54:31: you and the Resate team do about ongoing
01:54:33: technical work, the road map, etc. make
01:54:35: me super hopeful in Resonate's future.
01:54:37: Thank you so much for that.
01:54:38: Hey, thank you. I do like Vana being
01:54:41: like very open with the development like
01:54:43: it's kind of like
01:54:45: kind of like fill your face.
01:54:48: Yeah. And it's like
01:54:49: it can
01:54:54: it can be a double-edged sword
01:54:56: sometimes, but I think it's something
01:54:57: that we'd prefer to do because not a lot
01:55:00: of not a lot of companies do it and
01:55:05: stand out. A lot of companies will just
01:55:07: kind of be like very like hush hush with
01:55:09: stuff they do and like not do this kind
01:55:11: of stuff and like you know I do want us
01:55:13: to be like different and also like
01:55:15: because it kind of like it feels like
01:55:17: it's better that way and it's like you
01:55:19: know like often times like seeing other
01:55:22: companies and I'm like like I don't want
01:55:24: us to be like that you know I want us to
01:55:26: like be different and it's definitely
01:55:27: there's negatives to it and there's like
01:55:29: cases where I'm like I understand why
01:55:32: the other companies don't do it.
01:55:35: Uh but also like you know like we I feel
01:55:37: like we should be different. We should
01:55:38: be you know we should be the change we
01:55:41: want to see in the world as the saying
01:55:44: goes.
01:55:45: Yeah.
01:55:47: Also like it's also just I enjoy like
01:55:49: you know talking about technical stuff
01:55:51: and so on. It's one of the reasons like
01:55:53: started like these like um started like
01:55:57: you know these uh resonances as well. It
01:56:00: helps um be able to ramble about it a
01:56:03: fair bit. and talking about like you
01:56:06: know the reason and the motivations and
01:56:07: things in part because like for me like
01:56:10: you know I'm like I'm very like
01:56:13: technical person but also like I like
01:56:16: like you know a lot of like kind of
01:56:17: broad design and architectural design
01:56:19: and sort of like philosophy of like how
01:56:22: you approach things and how you solve
01:56:23: problems on a high level and like to
01:56:25: kind of combine these two because I can
01:56:27: be like you know this is a big sort of
01:56:31: philosophy of how this should be
01:56:32: designed because like you know these
01:56:35: kind of like things like lead you to
01:56:36: this thing and this thing and it can be
01:56:38: like you know this is how we this is
01:56:40: exactly how we achieve it on you know
01:56:43: technical level and I feel with a lot of
01:56:45: companies like you know it also is a
01:56:48: thing where like you know those things
01:56:49: are done by two different people
01:56:52: and like that like leads to like
01:56:54: inconsistencies because like you know
01:56:56: they need to communicate and
01:56:57: communication is ultimately lossy
01:57:01: um and you're not able to like you know
01:57:03: really give the vision you have in your
01:57:05: head for something and have it like
01:57:07: translate like fully.
01:57:10: So like I think like you know there's
01:57:11: kind of benefit and like in being able
01:57:13: to talk about it like you know it's I
01:57:16: enjoy like kind of process like sharing.
01:57:21: Okay, we got about two and a half
01:57:23: minutes left. We have maybe time for one
01:57:25: question but I think there's too many
01:57:27: more questions coming
01:57:29: so we'll probably end this at this
01:57:30: point. Um, I should not forget uh let's
01:57:34: check first we actually if we have
01:57:36: anybody to raid because if anybody's
01:57:38: like streaming right now we do want to
01:57:40: send
01:57:43: um we want to send everyone
01:57:46: to them because we want like you know we
01:57:48: want more people to stream and showcase
01:57:50: and um so if you if you're a streamer
01:57:55: like this is this is a really good time
01:57:57: to like be streaming Resonite because
01:57:59: we'll send people your Okay.
01:58:02: But it doesn't seem that anybody's
01:58:05: streaming, unfortunately. Not even
01:58:07: Creator Jam.
01:58:10: Yeah, I don't see anyone. So,
01:58:11: unfortunately, I think this will end
01:58:13: here. I could do a funny and just send
01:58:15: people like someone random, but I don't
01:58:17: know who. Um,
01:58:21: yeah, I think I might just like end it
01:58:22: here, unfortunately.
01:58:31: Yeah, I don't see anyone. So, uh, with
01:58:34: that, uh, thank you everyone, you know,
01:58:35: for watching. Thank you, Syro. Uh,
01:58:38: yeah,
01:58:40: on Twitch about Resonite. That's what
01:58:42: we're streaming on.
01:58:45: We're going to raid ourselves.
01:58:48: Um,
01:58:50: yeah. Thank you everyone for like, you
01:58:51: know, watching. Thank you for all the
01:58:52: questions. Uh, thank you for Syra for
01:58:54: like, you know, being here helping
01:58:55: answer things.
01:58:56: Um, and thank you everyone for, you
01:58:58: know, for your support, whether it's
01:58:59: like, you know, just playing as night,
01:59:01: being part of our platform, uh, you
01:59:03: know, talking about it, showing people
01:59:04: around, building cool content,
01:59:07: um, or supporting us, you know,
01:59:08: financially, whether it's on like
01:59:10: Patreon or Stripe. Uh, if you do support
01:59:13: us on Patreon, please consider switching
01:59:16: to Stripe because we get about 10% more
01:59:19: funds for the same amount you give,
01:59:22: which actually helps us a lot. Um,
01:59:26: and like if you in case you also missed
01:59:28: it, strap is now out of beta. So the the
01:59:31: major bug with it like you know it
01:59:34: wasn't even that ma like it wasn't a
01:59:35: major bug. Uh, but the last kind of bug
01:59:37: we consider a blocker was fixed. Uh, so
01:59:41: it's pretty much stable. It actually
01:59:42: gives better experience at Patreon. Um,
01:59:45: so please consider like switching
01:59:46: because that will help us quite a bit.
01:59:49: Um, so thank you very much. Uh, and
01:59:52: we'll see you for the next one. And
01:59:56: hopefully splittening pre-release
01:59:59: tomorrow.
02:00:01: Hopefully, assuming things don't
02:00:02: explode.
02:00:03: Fingers crossed.
02:00:04: Fingers crossed. You might be able to
02:00:06: get your hands
02:00:07: your toes and your legs.
02:00:08: Yeah, just crissc cross everything to
02:00:10: become become a pretzel.
02:00:12: I don't I don't know what I'm doing.
02:00:15: If you all of your pretzel energy
02:00:16: combined might make it happen, maybe.
02:00:18: Yes.
02:00:21: So, I'm going to check if there's nobody
02:00:24: to stream in case like somebody started
02:00:26: streaming.
02:00:31: Nope.
02:00:33: Okay. Unfortunately, this ends here.
02:00:36: Nobody to send you to. So, thank you
02:00:38: very much. Uh, thank you very much again
02:00:40: and we'll see you soon. Bye.
02:00:43: Bye, fellas.
02:00:45: Borf.