nwc10 good *, bots 07:59
brrt good * #moarvm 08:28
nwc10 good *, brrt 08:29
brrt can't wait for the christmas holiday
samcv good * all- 13:51
lizmat samcv o/ 13:52
lizmat hopes samcv has been synchronized :-)
samcv yeah doing pretty good now
lizmat :-) 13:53
samcv was up at 8am
lizmat that's early for me :-)
nwc10 you get more morning that way 13:54
samcv well that's 11pm in california
brrt ohai samcv 13:55
samcv working on getting us able to test every unicode codepoint
lizmat whee! :-)
.oO( test all the things )
samcv trying to get it to work with Test. since not like i'm going to have 800,000 is-deeply's
and it would slow it down. i already made many things be in nqp so it doesn't take minutes to work 13:56
lizmat some files in roast are generated
perhaps that could be an inspiration ?
samcv well. it is going to read the unicode files themselves. that is much faster than generating a million line file
nwc10 IIRC the way I'd (tried to) things like that before (in Perl 5) was to do the actual test in regular code, then if it failed re-test it with Test::More, so that it fails, generates diagnostics, and gets counted 13:57
samcv but i'm trying to get a series of todo's working so that you set in the file how many failures you expect. and it will pass or fail or todo things depending on such
lizmat well, that sounds ok, but that doesn't guarantee that parsing of every unicode char actually works
samcv so if you pass more than it thought you were going to, then the todo'd thing will pass. and if not then it will not be todo'd
nwc10 and then have something that also outputs the number of implicitly passing tests that weren't reported
samcv and it will be a raw failure
nwc10 OK, what I just said doesn't sound like it helps totally here 13:58
samcv not ok 2 - Property General_Category from file extracted/DerivedGeneralCategory.txt General_Category=Cn got <= 97 failures # TODO Expecting 97 failures from General_Category=Cn
nwc10 er, however, one internal optimisation that IIRC Test::More in Perl 5 now has is to assume that most tests pass most of the time
and not store the array of passing test numbers if they all have
but just the highest pass
samcv should probably change the wording a bit it makes it sound like the test tests something different
nwc10 and extrapolate that as 1 .. $n 13:59
whenever asked about it

Powered by Cro and the Raku® Programming Language.

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

Thank you!