github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
AlexDaniel who will be doing the moar release this time? 01:16
samcv: ?
samcv AlexDaniel, i could do it. also good morning 06:07
probably. i'm a bit under the weather right now. when does it need to be done 06:08
AlexDaniel samcv: likely not today, maybe tomorrow or in two days 09:58
samcv: fwiw there's also this discussion: github.com/perl6/problem-solving/issues/53
samcv: but in rakudo we'll soon start a release branch I think, so do it when you feel alright 10:00
samcv AlexDaniel, okay. great :) 10:12
MasterDuke hm, when does one want to use MVM_6model_get_debug_name vs MVM_6model_get_stable_debug_name ? 13:20
timotimo jnthn: if you want to hear something fun, the return value decont spesh plugin commit is implicated in a bug where something supply-related accidentally returns a low-level object (i forgot which) into sink context 14:01
that's what dogbert11 talked about yesterday, S17-supply/return-in-tap.t 14:02
jnthn Yeah, saw that; doesn't make terribly much sense... 14:34
jnthn I'm suspecting something race-y, just not sure where it could be. 14:35
timotimo race-y, huh 14:36
i hadn't considered that 14:37
jnthn When compiling MoarVM HEAD I see some warnings... 15:14
src/profiler/configuration.c: In function ‘validate_op’:
src/profiler/configuration.c:273:51: warning: format ‘%d’ expects argument of type ‘int’, but argument 3 has type ‘MVMuint64 {aka long unsigned int}’ [-Wformat=] MVM_exception_throw_adhoc(tc, "STRUCT_SELECT string length %d (index %d) NYI or something", string_length, string_idx);
dogbert11 jnthn, timotimo: bisecting this was quite difficult given that you sometimes have to wait a quite long time for the error to occur. So there's a chance that the last working commit, i.e. 523ffae84 might fail if run for long enough time. 15:27
but I did let it run in one terminal while running a spectest using all cores in another and it didn't break 15:28
and since I'm on AMD Ryzen I'm unable to use rr 15:29
timotimo d'oh, i never get warnings about format strings 16:00
japhb timotimo: Did you see my Terminal::QuickCharts link over on #perl6-dev? You're one of the people I thought might have fun with that (despite that right now it's really incomplete, there's just enough to show the concepts and do basic stuff) 16:01
timotimo that's why i keep pushing commits that have them >_< 16:02
japhb: i have not, and i'm only at the keyboard for a little moment, but it does sound interesting!
japhb Well, if you do get a couple minutes, feedback welcome. It's not ready for ecosystem yet, but early feedback appreciated. :-)
timotimo why are there no screenshots, why do i have to run the code myeslf? :P 16:31
timotimo afk again
Geth MoarVM: d6f9e02bae | (Daniel Green)++ | src/profiler/configuration.c
Fix format string warnings
16:32
MasterDuke timotimo: just fixed the warnings
japhb .ask timotimo What's the typical protocol for screenshots these days? Some static image files in examples/ with links from the README? Or is there something else I should use? 16:54
yoleaux japhb: I'll pass your message to timotimo.
timotimo tbh i don't have a good suggestion :( 17:01
yoleaux 16:54Z <japhb> timotimo: What's the typical protocol for screenshots these days? Some static image files in examples/ with links from the README? Or is there something else I should use?
timotimo it's not great that the screenshots - if they are in the git repository - have to be downloaded every time the module is installed (if it's in the ecosystem rather than cpan)
timotimo the examples do look really nice 17:14