!MljVigmyqZqNMImsdX:matrix.org

Rust Gamedev WG

45 Members
The Rust Gamedev Working Group3 Servers

Load older messages


SenderMessageTime
27 Jul 2020
@vancha112:matrix.orgvancha112if i need to mutate it in the update function, how can i show it in the render function? :(21:07:43
@vancha112:matrix.orgvancha112seems fixed..21:28:56
@ralith:ralith.comRalithmostly people don't use piston these days21:59:44
31 Jul 2020
@zen3ger:matrix.orgzen3ger joined the room.06:59:47
2 Aug 2020
@msavoritias:privacytools.iomsavoritias left the room.00:05:30
@cata0309:matrix.orgcmarincia changed their display name from cata0309 to cmarincia.15:47:39
@cata0309:matrix.orgcmarincia set a profile picture.15:59:26
5 Aug 2020
@ozkriff:matrix.orgAndrey @ozkriff Lesnikov https://github.com/rust-gamedev/rust-gamedev.github.io/pull/229 - r? 22:59:55
9 Aug 2020
@ozkriff:matrix.orgAndrey @ozkriff Lesnikov https://github.com/rust-gamedev/rust-gamedev.github.io/issues/233 what do you think, folks? 19:56:03
18 Aug 2020
@17cupsofcoffee:matrix.org17cupsofcoffee joined the room.15:23:04
19 Aug 2020
@senzaki:matrix.orgsenzaki left the room.06:10:09
27 Aug 2020
@costalfy:matrix.orgcostalfy__ joined the room.22:01:25
31 Aug 2020
@scoopr:matrix.orgscoopr set a profile picture.10:48:04
7 Sep 2020
@ozkriff:matrix.orgAndrey @ozkriff Lesnikovhttps://github.com/rust-gamedev/rust-gamedev.github.io/pull/274 - final draft, everyone is invited to review! :) Hope to merge this today in a few hours if there're no objections.18:21:12
@kvark:matrix.orgkvarkhttps://rust-gamedev-preview.netlify.app/posts/newsletter-013/#bevy-engine has way to many lists18:24:15
10 Sep 2020
@iamdejan:matrix.orgiamdejan left the room.09:21:30
18 Sep 2020
@ralith:ralith.comRalithsebcrozet is going to extremes to refactor the nphysics stack to avoid performing monomorphization in downstream code02:49:25
@kvark:matrix.orgkvarkhiding generics?02:50:08
@ralith:ralith.comRalithIMO this is a laborious workaround for a shortcoming in rust, i.e. the absence of explicit generic instantiation to allow generic code to be pre-compiled for common usecases02:50:08
@ralith:ralith.comRalith(compare C++ explicit template instantiation, not to be confused with template specialization)02:50:33
@ralith:ralith.comRalithremoving generics and publishing multiple crates for different variations of what would otherwise be a generic parameter02:50:57
@kvark:matrix.orgkvark dumb question. What if you do pub type Foo = MyGeneric<Bar> in the library. Does it not affect how Rust generates the library code in any way? 02:51:42
@ralith:ralith.comRalithafaik it is impossible to suppress monomorphization by any means02:52:07
@ralith:ralith.comRalithfairly sure type aliases are erased very early on, in any case02:52:37
@ralith:ralith.comRalithanyway, this seems like a shame; I wonder if there's any way the wg could direct resources to improve the situation?02:55:08
@aclysma:matrix.orgaclysmaWhat is the intended effect of avoiding monomorohization in downstream code? I can think of a few but am curious to know if there is a particular reason subcrozet has in mind05:54:26
@aclysma:matrix.orgaclysma I know for me I find it annoying that much of the nphysics stack ignores cargo overrides because it uses optimization settings at the callsite 05:56:51
@wodannson:matrix.orgWodann I think if you export a type like kvark proposes in an upstream crate and use that type in downstream code, it should no longer perform monomorphisation. I could be mistaken though. 09:39:24
@ralith:ralith.comRalith
In reply to @aclysma:matrix.org
I know for me I find it annoying that much of the nphysics stack ignores cargo overrides because it uses optimization settings at the callsite
that's one benefit he's called out; the other is compile times
16:26:01
@ralith:ralith.comRalith
In reply to @wodannson:matrix.org
I think if you export a type like kvark proposes in an upstream crate and use that type in downstream code, it should no longer perform monomorphisation. I could be mistaken though.
fairly certain no such behavior exists; type aliases are just cosmetic shorthand, they have no effect on codegen
16:26:55

There are no newer messages yet.


Back to Room List