github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
00:08 sena_kun left 00:35 konvertex left 02:03 Voldenet left 02:09 Voldenet joined, Voldenet left, Voldenet joined 04:47 leont joined
samcv are we moving everything to azure? 06:29
lizmat samcv: I believe that's the plan for CI, yes 08:46
08:53 sena_kun joined 09:12 patrickb joined
patrickb o/ 09:13
nwc10 \o
patrickb samcv: The plan is to only have AzureCI.
nwc10 I assume that this is because "it's better", but if you said more, I forgot the "why" 09:16
patrickb I think Azure does everything we need. So once Azure is in shape and our setup does the same things the others do we can shut down travis, appveyor and circleci. 09:17
nwc10 aha. in that, none of those three provide everything we need? (clearly, if I have this right, not even any two of them offer enough) 09:18
patrickb nwc10: Better in this case means: Win/Mac/Linux support, No monthly usage limits.
none of the others did sadly.
I actually tried CircleCI first but then hit a bug in their billing that prevented us from actually doing any Windows builds. 09:19
nwc10 OK, good :-) I was going to guess that it was something about (also) having Mac and also Win32
forgotten about usage limits
patrickb Travis is bad with Windows, AppVeyor is bad with non-Windows.
All the others have some form of monthly rate limits.
nwc10 Microsoft has deep pockets and wants everyone to be friends with them, so is throwing CPU at the problem? (also, if one *owns* the cloud service, it's less cost. Or even "a different business model") 09:20
much like (effectively) they are sponsoring github as a loss-leader, for other things they do want to sell to folks.
patrickb nwc10: Might be so. But the why doesn't really matter as long as we can assume it stays that way.
nwc10 I'm not too worried about the "why" on this, as long as it stays that way for "long enough" 09:21
patrickb I have to leave again. o/
nwc10 but the sale of github was an eye-opener. In one sense, github had the entire open source world hostage (or at least, a hostage to fortune) and it's lucky that they didn't go bust, or get bought by a (subtly) unfriendly firm
have fun \o 09:22
09:22 patrickb left 09:35 Altai-man_ joined 09:38 sena_kun left 11:36 sena_kun joined 11:38 Altai-man_ left
Geth_ MoarVM: de8955822c | (Stefan Seifert)++ | src/core/bytecodedump.c
Fix segfault in bytecode dumper when local_types is NULL
12:57
MoarVM: 9e12424fbe | (Stefan Seifert)++ | src/core/bytecodedump.c
Show index of outer lexials in bytecode dump instead of ??
13:35 Altai-man_ joined 13:38 sena_kun left 14:06 zakharyas joined 15:36 sena_kun joined 15:38 Altai-man_ left 16:00 zakharyas left 17:35 Altai-man_ joined 17:38 sena_kun left 18:15 lucasb joined 19:04 brrt joined 19:36 sena_kun joined 19:38 Altai-man_ left
MasterDuke two weeks until a release. anyone object to merging github.com/MoarVM/MoarVM/pull/1281 ? 19:55
nine certainly not me :) 20:06
MasterDuke cool, i'll merge now, that'll give it time to bake before the release 20:08
Geth_ MoarVM: 9b60b37ca9 | (Daniel Green)++ | 9 files
Spesh and JIT get(cur)?hllsym

Introduce a new spesh op `sp_gethashentryvalue` that has an object destination and the literal address of an MVMHashEntry. Then the interpreter and JIT just stick the `->value` of that MVMHashEntry in the destination.
... (26 more lines)
20:09
MoarVM: 5f0f7919d1 | MasterDuke17++ (committed using GitHub Web editor) | 9 files
Merge pull request #1281 from MasterDuke17/spesh_getcurhllsym

Spesh and JIT get(cur)?hllsym
20:13 zakharyas joined 20:22 travis-ci joined
travis-ci MoarVM build failed. MasterDuke17 'Merge pull request #1281 from MasterDuke17/spesh_getcurhllsym 20:22
travis-ci.org/MoarVM/MoarVM/builds/695492850 github.com/MoarVM/MoarVM/compare/9...0f7919d104
20:22 travis-ci left 21:09 zakharyas left 21:35 Altai-man_ joined 21:38 sena_kun left 22:16 brrt left 23:17 leont left 23:36 sena_kun joined 23:38 Altai-man_ left 23:42 sena_kun left