github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
Geth MoarVM: vrurg++ created pull request #1204:
Use `make` exit status as own		 01:38
Geth MoarVM: bb239a91df | (Stefan Seifert)++ | src/types.h
Fix MVMSpeshPluginState type defined twice

Fixes part of GH #3266		 06:43
travis-ci MoarVM build errored. Stefan Seifert 'Fix MVMSpeshPluginState type defined twice 07:04
travis-ci.org/MoarVM/MoarVM/builds/602174175 github.com/MoarVM/MoarVM/compare/b...239a91df14
nine What I would give for a reliable CI... 07:11
tadzik heh 07:23
I have mostly good experience with gitlab-ci fwiw
but travis... for some reason, whenever something travis-backed shows CI failures, it's always travis' fault rather than the code's
brrt \o 07:51
tellable6 2019-10-22T20:24:05Z #moarvm <lizmat> brrt please do write better notes, or better commit messages :-)
brrt I'll try :-D
nwc10 o/ 07:55
Geth MoarVM: 32003eecfd | (Vadim Belman)++ | Configure.pl
Use `make` exit status as own

Propagade `make install` exit code for upstream NQP `Configure.pl`. Otherwise NQP continues to build as if nothing happened and would confuse a user with bizzare _'too old moar'_ error message if older moar is already instaleld.
Resolves rakudo/rakudo#3266		 08:36
synopsebot RAKUDO#3266 [closed]: github.com/rakudo/rakudo/issues/3266 [build] Rebuilding fails on older git
MoarVM: 67b6acf71e | (Jonathan Worthington)++ (committed using GitHub Web editor) | Configure.pl
Merge pull request #1204 from vrurg/rakudo_3266

Use `make` exit status as own
Kaiepi i suspect the way async sockets deal with their handles may be prone to memory leaks 14:55
it frees the handles whenever they get closed, rather than from the gc free function. it's not guaranteed there will be a socket that needs closing though, like if uv_tcp_connect or uv_tcp_bind fail, since uv_tcp_create doesn't actually create any kind of socket until another function that requires it to exist is used 14:58
is this something worth looking into before the release or should i just focus on reverting the changes to dns resolution that are breaking? 15:01
jnthn No, I'd rather that such changes would have time for more testing 15:03
nine Kaiepi: the important part of releases is to not introduce new errors. As long as we ensure that, we can say that every release is better than the previous one and we're on the right track. Requiring a release to fix all known errors would just lead to never releasing at all. 15:04
That's also why reverting is often preferable to last minute fixing frencies - you can't say what new issues you introduce with your fixes. 15:06
s/frencies/frenzies/
Kaiepi aight 15:12
timotimo Stage start : 0.000 17:41
Malformed UTF-8 near bytes 00 00 80
at SETTING::src/core.c/Encoding/Decoder/Builtin.pm6:44 (/home/timo/perl6/install/share/perl6/runtime/CORE.c.setting.moarvm:consume-line-chars)
how did this even happen
i'll try rebuilding nqp and rakudo first 17:43
with latest master
ok the error is gone 17:49
Geth MoarVM: Kaiepi++ created pull request #1205:
Revert to only using the first address received from a DNS resolution		 18:01
AlexDaniel Kaiepi: wow, well. Is it just a revert or are there any extra changes? 18:10
Kaiepi i'll need to double check AlexDaniel, this was something i needed to do manually 18:28
actually AlexDaniel, would reverting this whole mess and remaking the changes that are worthwhile to have so the commits are possible to work with after the release be ok? 18:36
rather than just the commit that broke tests 18:39
AlexDaniel Kaiepi: If you can do that, then probably yes 19:03
Kaiepi: but it sounds like a huge pain in the ass, so I'm definitely not insisting on that 19:04
Kaiepi: also, is anybody from the moarvm team helping you out with the grant stuff? They might have a better opinion than I have 19:05
by helping out I mean like supervising and giving advice 19:06
Kaiepi i'm supposed to have a supervisor, but we haven't been in contact since we first got introduced, which was at least a couple months ago by this point AlexDaniel 19:14
not sure if they're on the moarvm team or not
AlexDaniel, the changes themselves didn't take terribly long to make, just i had a lot of trouble getting anyone to review them before they could be merged, so it's something i can do 19:18
AlexDaniel Kaiepi: then you're probably the only person who can do a good decision :) 19:24
Kaiepi: As long as I'm relatively sure that the changes are not breaking existing code in the wild, I'll be fine with it
Geth MoarVM: Kaiepi++ created pull request #1207:
Revert "Merge pull request #1166 from Kaiepi/ipv6-unsupported"		 19:34
brrt .tell Kaiepi - I'm not aware of the details of your grant, so I'd need some time to get up to speed, but I'll volunteer as a 'moarvm-side' mentor, if that'd be of any help 20:09
tellable6 brrt, I'll pass your message to Kaiepi

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!