| lizmat | Files=1207, Tests=62761, 219 wallclock secs (13.15 usr 5.02 sys + 1326.96 cusr 134.70 csys = 1479.83 CPU) | 08:27 | |
| [Tux] | This is Rakudo version 2017.06-64-gfe7ea124c built on MoarVM version 2017.06-14-g888b33d1 | 09:04 | |
| csv-ip5xs 2.631 | |||
| test 13.068 | |||
| test-t 4.091 - 4.154 | |||
| csv-parser 12.586 | |||
| Geth | rakudo/nom: 1f5bed4b4b | (Elizabeth Mattijsen)++ | src/core/Rakudo/QuantHash.pm Introducing R:Q.BAGGY-CLONE Create a deep clone of any Baggy/Mixy |
10:37 | |
| rakudo/nom: a2a0d5c622 | (Elizabeth Mattijsen)++ | 8 files Tighten up all inter-QuantHash coercions/clone There were quite some cases where coercing a mutable to an immutable version or clone would keep the underlying values shared, potentially causing quite some action at a distance. |
|||
| rakudo/nom: 19a1caa330 | (Elizabeth Mattijsen)++ | 2 files Fix various coercion related issues with (-) |
12:03 | ||
| dogbert17 | .seen lollercopter | 12:13 | |
| yoleaux | I saw lollercopter 24 Jun 2017 17:00Z in #perl6: <lollercopter> \o | ||
| dogbert17 | .tell lollercopter hasn't IOninja's recent IO work made github.com/perl6/doc/issues/852 closable? | 12:15 | |
| yoleaux | dogbert17: I'll pass your message to lollercopter. | ||
| Geth | roast: 274766007d | (Elizabeth Mattijsen)++ | S03-operators/difference.t Add quite a few (-) coercion tests |
13:03 | |
| dogbert17 | I'm trying to fix github.com/perl6/doc/issues/858. I have tried to improve the text but there's always a chance that it has become even worse than before. Grateful if anyone would like to look it over | 14:05 | |
| and here's the gist: gist.github.com/dogbert17/1674bbf7...b5cbc300a4 | 14:06 | ||
| lizmat | dogbert17: looks like an improvement to me :-) | 14:08 | |
| dogbert17 | lizmat, thx. If no one else comments I'll add the text and close the issue :) | 14:12 | |
| Geth | rakudo/nom: f3b2d952d6 | (Elizabeth Mattijsen)++ | src/core/Rakudo/QuantHash.pm Fix issue with Baggy (-) Setty Caught while writing more tests |
14:14 | |
| rakudo/nom: b66d87833d | (Elizabeth Mattijsen)++ | 2 files Make Setty (-) Iterable about 3.5x faster - actually infinitely faster when left side is empty Set, as we don't iterate at all in that case - also stops iterating as soon as result is the empty Set, which in some case may give significantly better results |
|||
| rakudo/nom: 36ca2fb4a5 | (Elizabeth Mattijsen)++ | 2 files Make Setty (-) Setty stop when result is empty Set Also abstract the logic into R:Q.SUB-SET-FROM-SET |
14:33 | ||
| dogbert17 | m: say 2.5.^lookup(‘sqrt’).perl | 14:42 | |
| camelia | method sqrt (Rat $: *%_) { #`(Method|52406416) ... } | ||
| dogbert17 | when doing the above is it method 'lookup' hidden in 'role Metamodel::MethodContainer' which is called? | 14:45 | |
| looking ar github.com/perl6/doc/issues/875 | |||
| *at | |||
| lizmat | dogbert17: any foo.^method is really foo.HOW(foo).method | 14:52 | |
| is that an answer ? | |||
| timotimo | since 2.5 has the ClassHOW HOW you can find what roles it looks through in the ClassHOW file | 14:59 | |
| and yeah, i'd expect this to be MethodContainer's lookup method | |||
| Geth | roast: 3a994f69e3 | (Elizabeth Mattijsen)++ | S03-operators/difference.t Add two coercions, fix 2 coercions for (-) |
||
| rakudo/nom: a2133dbc6a | (Elizabeth Mattijsen)++ | 2 files Fix Baggy/Mixy (-) Any/Iterable Any/Iterable should be coerced to Set, not Bag/Mix |
15:01 | ||
| dogbert17 | lizmat, timotimo, thx, if we're to add docs for lookup it's good to know where :) | 15:31 | |
| m: say 2.5.^lookup(‘sqrt’).perl | 15:39 | ||
| camelia | method sqrt (Rat $: *%_) { #`(Method|42670208) ... } | ||
| dogbert17 | m: say 2.5.^lookup(‘not-exist’).perl | ||
| camelia | Mu | ||
| dogbert17 | so if no method object can be found it returns a type object? | 15:41 | |
| lizmat | dogbert17: nqp land doesn't know about Nil | 15:42 | |
| dogbert17 | lizmat thx ahain | 15:53 | |
| *again | |||
| timotimo, lizmat: here's my first attempt, rip it to shreads :) gist.github.com/dogbert17/4bb9490a...aad381925c | 16:06 | ||
| I'm a bit unsure if I should write 'method lookup(Metamodel::MethodContainer: $obj, $name)' when the code explicitly says 'method lookup($obj, $name)' | 16:09 | ||
| Geth | rakudo/nom: 9936a3be28 | (Elizabeth Mattijsen)++ | 2 files Make Setty (-) Map between 2x and 3x faster - 3x for objecthashes - 2x for normal hashes and Maps |
16:21 | |
| lizmat | afk& | ||
| dogbert17 | I've now updated the gist, grateful for feedback | 16:39 | |
| masterduke++, fixed | 16:51 | ||
| MasterDuke | an updated example of coverable's output if anybody is interested: gist.github.com/Whateverable/2065c...4159bb3fca | 20:20 | |
| thoughts/suggestions welcome | |||
| timotimo: ^^^ | |||
| AlexDaniel | more suggestions? gist.github.com/Whateverable/a65c0...3d5135d3c6 | 22:59 | |
| timotimo: ↑ | |||
| timotimo | lovely! | 23:01 | |
| Geth | rakudo/nom: e4f3358fa8 | (Elizabeth Mattijsen)++ | src/core/set_operators.pm Make Any (-) Map|Iterable about 2x faster - because the right hand side no longer gets coerced to a Set first - but instead is handled on-the-fly - possibly without any iteration if the left is equivalent to an empty Set |
23:06 | |
| lizmat | which concludes my hacking for today | 23:07 | |
| good night! |