[00:23] *** leppard left [01:18] *** leah2 left [01:24] *** hulk joined [01:25] *** kylese left [01:29] <[Coke]> on rakudo, I agree. on NQP, there's just one failure in make test [01:31] *** leah2 joined [01:44] <[Coke]> ... looks like we're already just testing the build for JVM, not the tests. [01:47] <[Coke]> Looks like we're not doing win or mac rakuast by default? [01:52] <[Coke]> I assume relocatable doesn't work on mac? [02:15] *** hulk left [02:15] *** kylese joined [02:39] *** wayland left [03:09] *** human-blip left [03:12] *** human-blip joined [03:55] timo: if the rule contains a rule, that would propagate as well ? [03:55] *** human-blip left [03:56] m: grammar G { rule TOP { :i 'Foo' }; rule bar {'bar'} }; say G.parse("FOO bar") [03:56] simon_sibl, rakudo-moar 177d09d2c: OUTPUT: «｢FOO bar｣␤ bar => ｢bar｣␤» [03:56] m: grammar G { rule TOP { :i 'Foo' }; rule bar {'bar'} }; say G.parse("FOO BAR") [03:56] simon_sibl, rakudo-moar 177d09d2c: OUTPUT: «Nil␤» [03:56] ok ok [03:57] m: grammar G { rule TOP { :i 'Foo' 'Foo' }; rule bar {'bar'} }; say G.parse("FOO foo bar") [03:57] simon_sibl, rakudo-moar 177d09d2c: OUTPUT: «｢FOO foo bar｣␤ bar => ｢bar｣␤» [03:57] and it is indeed for the whole rule, still fine for what I wrote ok ! thanks [03:58] *** human-blip joined [04:46] *** leppard joined [04:56] *** xinming joined [05:32] *** hwj joined [05:34] *** lucerne90 left [05:36] *** Sgeo_ joined [05:38] *** Sgeo left [06:06] *** Sgeo_ left [07:13] *** rnddim joined [07:14] *** rnddim left [07:14] *** ShimmerFairy left [07:14] *** rnddim joined [07:15] *** rnddim left [07:16] *** rnddim joined [07:29] *** rnddim left [07:29] *** rnddim joined [07:33] *** leppard|2 joined [07:33] *** leppard|2 is now known as Inline [07:33] *** leppard left [07:33] *** Inline is now known as leppard [08:05] *** lucerne90 joined [08:17] *** dakkar joined [08:22] *** spa1983 joined [08:22] *** spa1983 left [08:22] *** sp1983 joined [08:23] . [08:23] voldenet: I have decided to start my own git server from the scratch and incorporate dsci into it ) [08:24] let's play big, ha-ha ) [08:32] *** sp1983 left [09:56] *** sp1983 joined [09:58] By starting from the scratch I mean I am going to develop my own gitlab/forgejo alternative to host git repos . In mvp version I am thinking git pull / push and merge request . Plus of course dsci ci which will be part of the solution … [09:59] *** sp1983 left [12:17] *** El_Che_ left [12:18] *** El_Che joined [13:19] *** leppard|2 joined [13:22] *** leppard left [14:05] *** vrurg_ joined [14:08] [Coke]: if there is a limit, I have missed it. Also worth mentioning is the tap vs act distinction [14:08] <[Coke]> ? [14:08] [Coke]: eek, sorry, it looks like I responded to a quite earlier message from you :/ [14:08] <[Coke]> Is that a throwback to an older discussion? [14:09] *** vrurg left [14:09] You had asked about whether there is a limit to the number of taps on a supply [14:09] hi, is it possible to have a published package with an independent sub package which can be used alone? for example, use A; or use A::B [14:10] I need to remember that Halloy doesn't scroll forward automatically [14:12] the purpose is to use common constant resources. [14:13] [Coke]: Since I can no longer even find the message in the scrollback, were you able to successfully do a bunch of taps on a supply? [14:13] Because I don't think there is any set limit, but it would be interesting to have some details from a stress test [14:14] <[Coke]> ab5tract: i think I ended up completely giving up on that approach. [14:14] <[Coke]> instead of having chained supplies, I did a race on the first set to get a second set, then did a race on that. [14:15] <[Coke]> And then went further and added a skiplist so I only had a small number to process (theoretically I should be checking everything, but it was just WAAAAAY too slow) [14:15] <[Coke]> s/set/group/, not Set [14:15] ah, makes sense. I took a similar approach in some local code I inted to share eventually for converting cbr files to pdf [14:16] (minus the skiplist) [14:17] you could "shake" the skiplist so that each run excludes a random subset, in case that feels like more "honest" coverage [14:18] Am I ass'ming myself with the assumption that this is blin related? [14:25] *** hwj left [14:27] <[Coke]> nope, $dayjob [14:27] <[Coke]> trying to find all instances of "X" across hundreds of git repos with maaaaany matching file types. [14:31] oh, slick! sounds like a perfect task for Raku [14:39] <[Coke]> s/file types/files/ [14:39] *** Sgeo joined [14:39] <[Coke]> I also thought slow and then didn't want to wait hours for it to finish. :| [14:39] <[Coke]> HEH [14:39] <[Coke]> thought *so* [14:42] :| indeed [14:51] <[Coke]> It's to grab a metric that we only need to look at once a quarter anyway. [14:56] <[Coke]> ... so I should care less about how long it takes. [14:58] weekly: https://5ab5traction5.bearblog.dev/shelling-in-raku-ne-is-unequaled-for-log-lines/ [14:58] ab5tract, Noted! (weekly) [15:04] So I haven't really done much with it before, but a single supplier can have multiple supplies [15:04] so if there were issues with having many taps, I wonder if they could be resolved by fanning out at the supply level instead of the tap level [15:07] m: dd my $supplier = Supplier.new [15:07] rakudo-moar 8eee82394: OUTPUT: «$supplier = Supplier.new(taplist => Supplier::TapList.new)␤» [15:07] the presence of taplist at the Supplier level makes me think not [15:15] forget my question about class and independent subs. [15:20] new question: didn't the default META6.json used to have an api value? if not, shoiuld i add it for a major incompatibility revision? if so should i start at 1 or 2? [15:20] m: my @foo = ; @foo[1,2,*.raku.say]; [15:20] rakudo-moar 8eee82394: OUTPUT: «6␤» [15:20] ab5tract: the callable in postcircumfix:<[ ]> gets only the number of elements, not the entire list [15:28] also: cool post! [16:29] timo: I assumed that -1 put it into numeric context [16:29] *** dakkar left [16:29] well, first I remembered it to be as you said [16:31] fixed. and thanks for reading! [16:32] "it sends the size" --> "but what about @a[*]?" --> "ah, but that's not a callable" [16:32] ^ the way my thinking _should_ have gone [16:54] *** leppard|2 left [17:23] . [17:34] *** lizmat left [17:45] *** lizmat joined [17:52] *** leppard|2 joined [17:54] *** wayland joined [17:54] *** belluzj joined [17:57] *** leppard|2 left [17:57] *** lizmat left [17:58] *** sergot left [17:59] *** lizmat joined [18:01] *** sergot joined [18:02] *** belluzj left [18:41] *** hwj joined [18:58] *** silug2 joined [18:59] *** silug left [18:59] *** silug2 is now known as silug [19:31] wow git server is a big goal [20:37] *** hwj left [21:13] . [21:57] *** [Coke] left [22:13] *** [Coke] joined [22:21] *** erooke left [22:27] *** silug left [22:49] *** erooke joined [22:50] *** silug joined [23:01] *** silug left [23:10] *** silug joined