ab5tract It occurs to me that once RakuAST is shipped as the default, we will probably want to make RAKUDO_RAKUAST=0 actually disable the new frontend. or else introduce a new environment variable eg RAKUDO_NO_RAKUAST, which feels a bit more awkward somehow but is probably equally acceptable in the end 04:04
actually the latter may be preferable, as it's weird to have an environment flag which is considered "on" by default 04:07
japhb ab5tract: I suspect that will rot in a damn hurry unless we make sure to run all CI with and without it. 04:46
ab5tract It can be temporary, but some people will prefer speed 04:47
And I don’t think we should block on that
But it’s a fair point. We can always refer such folk to an earlier release 04:48
japhb nodnod 05:09
07:43 [Tux] left 07:47 [Tux] joined 08:26 camelia left 08:43 camelia joined
nine gist.github.com/niner/af4abb8ab488...fc789ac8f4 10:27
11:20 finanalyst joined
timo nine: can you describe what issue exactly that addresses? the issue liz had with assuming not getting everything into the correct serialization context and exploding with "missing static coderef for closure blabla (blabla)' 13:42
13:42 finanalyst left 13:58 guifa_ left
nine Ah, no, that's just a fix for something MasterDuke was working on yesterday. Accidentally reposted due to input lag 14:02
14:02 guifa joined
timo oh, OK! 14:05
is there a simple way to cause the entire compilation to "skip to the end" from inside a module that gets "use"d? (in this case from a -M switch) 14:24
Geth rakudo/main: 3a86fd2d23 | timo++ (committed using GitHub Web editor) | docs/release-guide-binary.md
When a release happens, remind timo about compiler-explorer
14:29
timo i guess i can literally "exit(0)" haha 14:39
i don't really need for the rest of the normal compiler stuff to run probably
ab5tract timo: ooc, what’s the use case? 14:53
timo ehehehehe. 15:01
15:02 guifa left
timo i'll show something soon I hope 15:03
15:03 guifa joined
timo here's a fun one: when there's a -e on the commandline, and we "use" a module, the code inside the module seems to be attributed to "-e" instead of the actual filename 15:03
actually, not sure if that"s also for explicitly putting a "use" statement, i'm still woking with -M
it may be required to put "no precompilation" in that used modul as well 15:05
maybe important that i'm using -Ilib since i didn't write a META6 yet 15:08
ab5tract hmmm... yeah there's a few moving parts there, I guess 15:45
But ISTR that there is a relatively naive check related to `-e` when determining the file name / line number 15:47
How was the RSC? I'm bummed to have missed it, though I had a good weekend as well 15:55
timo it was good, i enjoyed it 16:25
ab5tract \o/ 17:13
timo maybe the next summit we'll have a chance to meet in person, that would be great 17:21
putting a $?FILES in github.com/rakudo/rakudo/blob/3dad...kumod#L127 is at least where i can give it a file that shows up, but $name is just the name of the compunit, so not a path and not a filename since it has the extension stripped 17:50
[Coke] waves from the atlantic 17:57
japhb waves back from near the Pacific 17:58
ab5tract timo: I'd love for these summits to happen regularly, as often as possible really :) 18:34
18:59 lizmat left 19:01 lizmat joined
timo the natural name for a blog engine (not static site generator per se) would be MyCroBlog 19:10
or i guess Cro::Blog so you can `my Cro::Blog $foo .= new;`
lizmat waves from Echt, really! 19:18
timo you made it back!
lizmat yeah, we did a bit of a visit to Amsterdam and then to Nijmegen, and now home again 19:28
pretty tired though :-)
19:30 librasteve_ joined
timo my neck still feels stiff, my body did not agree with the mattress i think :( 19:35
lizmat timo: sorry to hear that... :-(