[01:56] *** vrurg_ joined
[01:58] *** vrurg left
[05:38] *** xiaomiao left
[05:54] *** xiaomiao joined
[06:13] *** [Coke] left
[06:13] *** [Coke] joined
[08:26] <Geth> ¦ MoarVM/main: c68094f8b7 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/io/asyncsocketudp.c

[08:26] <Geth> ¦ MoarVM/main: Fix per-datagram memory leak in async UDP reads (#2017)

[08:26] <Geth> ¦ MoarVM/main: 

[08:26] <Geth> ¦ MoarVM/main: As described in: https://github.com/MoarVM/MoarVM/issues/2016

[08:26] <Geth> ¦ MoarVM/main: review: https://github.com/MoarVM/MoarVM/commit/c68094f8b7

[12:21] *** librasteve_ left
[12:21] *** ab5tract left
[12:21] *** Techcable left
[12:21] *** leont left
[12:22] *** SmokeMachine left
[12:22] *** tbrowder left
[12:23] *** Techcable joined
[12:23] *** ab5tract joined
[12:23] *** SmokeMachine joined
[12:24] *** leont joined
[12:25] *** tbrowder joined
[12:36] *** librasteve_ joined
[13:12] <[Coke]> Can someone update the changelog for that?

[16:29] <lizmat> done

[16:46] <[Coke]> lizmat++

[19:35] <timo> moved the num_strands member of MVMStringBody into the storage union and increased the in situ 8 storage up to 10 slots

[19:36] <timo> unlikely to be a earth-shattering performance improvement or anything, but it's always nice to have fewer allocations

[19:36] <lizmat> so what sizes of strings are we talking abour ?

[19:37] <timo> 10 graphemes if they all fit into 8bit, up from 8

[19:37] <timo> but i made the number a #define

[19:38] <lizmat> so it's adding memory for all strings, but 9/10 char strings don't need to be on the heap ?

[19:39] <timo> the 10 slots one doesn't add memory for strings because the num_strands member of the struct that is only used when a string has the strand storage type is now part of that same area

[19:39] <lizmat> ah... so always a net bonus!

[19:39] <timo> if you increase it above 10 it will make all strings larger