Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
Geth nqp: 825c2182a2 | (Nick Logan)++ (committed using GitHub Web editor) | src/vm/js/nqp-runtime/io.js
Add mostly functional uname op for js backend
 03:23
nqp/ugexe-patch-2: 6fff8dff3e | (Nick Logan)++ (committed using GitHub Web editor) | src/vm/js/nqp-runtime/core.js
[js] Implement cpucores
 03:35
nqp: ugexe++ created pull request #524:
[js] Implement cpucores
06:19 robertle left 07:50 robertle joined 07:51 ufobat joined 08:42 ExtraCrispy left
|Tux| Rakudo version 2019.03.1-47-g76b6416d0 - MoarVM version 2019.03-35-gd15906711
csv-ip5xs0.708 - 0.715
csv-ip5xs-206.215 - 6.234
csv-parser21.843 - 22.621
csv-test-xs-200.434 - 0.440
test6.904 - 7.059
test-t1.718 - 1.756
test-t --race0.774 - 0.869
test-t-2029.823 - 30.705
test-t-20 --race9.240 - 9.578
 09:29
NICE!
10:10 ufobat_ joined 10:13 ufobat left
Geth rakudo: 4825f31e29 | (Elizabeth Mattijsen)++ | t/08-performance/03-corekeys.t
Add test for strays in CORE::

This in relation to R#2773		 10:33
AlexDaniel 6c: .say if .starts-with(‘$’) for CORE::.keys 10:42
committable6 AlexDaniel, gist.github.com/b8a7d21dbcd862299d...8e070507f0
AlexDaniel 6c: .say if .starts-with(‘$’) for CORE::.keys.sort 10:44
committable6 AlexDaniel, gist.github.com/ce0f13234a00ec50cd...59a89c6d86
AlexDaniel lizmat: it's a great test to have 10:46
btw, are these wanted?
m: say $CORE_METAOP_ASSIGN
camelia (Mu)
AlexDaniel m: say $RAKUDO-VERBOSE-STACKFRAME
camelia 0
lizmat yeah... looking at some of those atm
m: dd Signally # say what ?
camelia Signally
lizmat that is *definitely* an implementation detail that doesn't need to be exposed 10:47
Geth rakudo: 45eb1c5f34 | (Elizabeth Mattijsen)++ | 2 files
Simplify "Signally" and move it from CORE::

By moving it into the Signal package		 10:56
11:09 epony left
Geth rakudo: ed074cd17f | (Elizabeth Mattijsen)++ | 2 files
Remove $RAKUDO-VERBOSE-STACKFRAME from CORE::

Keep marker inside the Backtrace class, accessible by a method. This also gets rid of one more race-condition "once". Also internally use the RAKUDO_VERBOSE_STACKFRAME rather than RAKUDO-VERBOSE-STACKFRAME, so that it matches the environment variable's name.		 11:38
moritz Hi there, does everybody get an email from CircleCE for every single push to rakudo? 11:59
I've just opted out on CircleCI, but I think nobody benefits from a CI that is red all the time
lizmat hasn't seen any mail 12:01
jnthn Me either 12:02
moritz maybe I got them because I created the org account 12:09
ugexe: did you build the circleci integration?
Geth rakudo: 6599998c2c | (Elizabeth Mattijsen)++ | src/core/PseudoStash.pm6
Revert PsuedoStash part of 5540212cfc8805f869ce

Fixes R#2775, although it is unclear as to why it fixes that.		 12:15
12:28 AlexDaniel left
lizmat m: my constant Range := Range but role { method foo { "foo" } }; dd Range.foo; dd (^10).foo # somehow I'd expect both to work 12:29
camelia "foo"
No such method 'foo' for invocant of type 'Range'
in block <unit> at <tmp> line 1
12:37 epony joined
jnthn lizmat: prefix:<^> is declared in CORE.setting, so it sees the lexical Range in there. 13:16
lizmat ah, duh :-) 13:20
13:22 lucasb joined
tobs lizmat++ # 2773 13:24
m: $p.Map
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$p' is not declared
at <tmp>:1
------> 3<BOL>7⏏5$p.Map
tobs I think github.com/rakudo/rakudo/issues/2366 can be closed then as well. 13:25
(uhh, maybe I should add this to the actual GH issue)
lizmat closed 2366] 13:26
13:26 AlexDaniel joined
tobs too fast for me :) 13:26
AlexDaniel m: say CLIENT::.keys 13:27
camelia GLOBAL can have no client package
in block <unit> at <tmp> line 1
Geth rakudo: 801850319a | (Elizabeth Mattijsen)++ | 3 files
Remove $CORE_METAOP_ASSIGN from CORE::

Basically make &METAOP_ASSIGN a front for R:I.METAOP_ASSIGN, on the assumption that current optimizing techniques will inline that call pretty quickly when needed. Also, maybe we should codegen a call to R:I.METAOP_ASSIGN directly, rather than to &METAOP_ASSIGN.		 14:03
lizmat m: dd CORE::<&await>.WHICH; dd SETTING::<&await>.WHICH # looks like we have an &await in CORE:: as well as in SETTING::, and they're different 14:09
camelia ObjAt.new("Sub|38266400")
ObjAt.new("Sub|35297312")
lizmat the core one I understand, the SETTING:: one I don't
14:39 mgoebel_ joined 14:46 brrt joined
brrt ohai #perl6-dev 14:46
ugexe moritz: i created the circle ci config file. thats it 14:56
as far as circleci being red -- maybe its a sign that things are broke. i mean its almost like after a tests flaps for so long we are just pretending its not broke anymore 14:58
*** Error in `/home/circleci/project/install/bin/moar': free(): invalid pointer: 0x000000000789cd64 ***
otherwise nobody is benefiting from those flapping tests period -- CI be damned 15:00
15:00 MasterDuke left
ugexe i agree the emails are annoying though -- i just dont know anything about it 15:00
there are organizational settings but i am not a rakudo org admin and thus cannot view the on circleci 15:03
ok i think it might just be you and me moritz. go to Workflows > (settings gear icon by rakudo row) > Overview > Unfollow Project 15:04
or you can go under your account settings and change the general Notification Settings 15:05
moritz ugexe: I've unsubscribed from the notifications; I just don't see the value of a CI if the results are unreliable, nobody seems to investigate them, and nobody can use it as signal for anything 15:09
we're just wasting energy
ugexe then the test itself is unreliable
*** Error in `/home/circleci/project/install/bin/moar': free(): invalid pointer: 0x000000000789cd64 *** 15:10
moritz but is anybody doing anything about it?
ugexe well i assume no. so our differences are around how flappers are seen period. 15:11
but if something is allowed to fail constantly in a given situation, why have the test
moritz my question is more: is anybody using the pipeline, at all?
if not, it's just carbon emissions without any reason 15:12
ugexe travisci is essentially dead. so the question is do we want CI at all? i would assume yes
moritz I'd assume so as well, but the current state of the pipeline is an indication of the contrary 15:13
ugexe its especially useful on windows (which everyone also ignores but me I think) to find when exactly someone broken it
moritz but we don't run any windows tests with circleci 15:14
ugexe ...correct 15:15
you are saying travis is good because it passes. im telling you circleci is good because its failing.
moritz I'm not saying travis is good
ugexe well, the alternative is no CI? i would say that is bad
moritz I'm saying that I don't see anybody using the results of circleci, so it's a waste of resources 15:16
ugexe i dont see anyone using any of the CI we have honestly
i know i do, but thats me
moritz so, you *do* use it
that's the first piece of new information from this conversation 15:17
then we have one user, and I'll shut up
15:18 patrickb joined
patrickb o/ 15:32
I'd like to get my relocatability branches merged. It's these: 15:33
github.com/MoarVM/MoarVM/pull/1053
github.com/perl6/nqp/pull/520
github.com/rakudo/rakudo/pull/2693
I have some free time in the coming days to do fixes for whatever might pop up as a result of the changes.
I can do a rebase if that makes a difference. (Can I do that without trashing the PR?)
Are there any objections? What can I do to make the merge happen?
ugexe do whatever makes it easiest on me and i'll do it tonight if no one else does or objects 15:34
well, i cant do the moarvm one 15:35
AlexDaniel we should fix every single flapper 15:37
currently it's hard to see when we introduce new flappers because we have so many
ugexe yeah i've ignored new legit test failures because they happened inside a test that usually flappers 15:38
flaps rather
AlexDaniel I've been marking the most annoying ones as blockers because they get in the way of the release process both for us and others (e.g. debian)
but there are too many and I shouldn't be abusing the label… 15:39
patrickb ugexe: Was that directed at me?
ugexe patrickb: yeah
patrickb OK. I'll see if doing a rebase is possible. 15:40
In the worst case I'll have to reopen the PR, but I guess we can live with that
ugexe patrickb: i dont have commit privs to moarvm though so it'd probably end up being someoen with a commit bit to all 3 repos
AlexDaniel I can merge all 3
this is the perfect time to merge this
releasable6: status
releasable6 AlexDaniel, Release manager is not specified yet.
AlexDaniel, Next release in ≈31 days and ≈3 hours. 0 blockers. 0 out of 52 commits logged
AlexDaniel, Details: gist.github.com/68aefd45b693e3525a...fa3e9cc3e4
patrickb AlexDaniel: I can do the rebase in ~ 3h, will shout once I'm done... 15:41
AlexDaniel patrickb: ok, sounds like a plan!
15:42 kawaii joined
ugexe fwiw though most projects have *some* flappers. so i wonder if make test couldn't be smart enough to retry a failed test file (since there should be no failures there should be very few retries for a given run) 15:43
thats not to say we couldn't fix the flappers we do have 15:44
AlexDaniel btw we don't even have a list of flappers anywhere 15:45
Zoffix used to send a message here whenever they stumbled upon one, and quotable6 was able to grep them all into a nice list
but this process no longer works because quotable6 doesn't know how to work with colabti 15:46
ugexe it'd be nice to have something like todo/skip but for flap
AlexDaniel quotable6: ZOFF?LOP 15:47
quotable6 AlexDaniel, OK, working on it! This may take up to three minutes (4582161 messages to process)
AlexDaniel, 224 messages (2017-04-14⌁2018-04-28): gist.github.com/61e361a16d7bdcbda1...1231503389 15:48
ugexe i haven't seen S10 or S11 flops in a long time. that like half of them that might no longer be an issue 15:50
AlexDaniel yes
I think some were fixed intentionally 15:51
so maybe we should file a ticket for everything that we stumble upon
some issues get much more fixable when there's a golf, so… 15:52
jnthn: can you add kawaii to the rakudo org? CLA was received 0 minutes ago 16:04
jnthn AlexDaniel: Yeah 16:11
AlexDaniel kawaii: please add yourself to release_guide.pod and CREDITS (once you have access) 16:12
kawaii AlexDaniel: will do as soon at github email me about access to the organisation
as *
jnthn Done 16:13
ugexe m: role T { has $.a }; my $x = 1 does T(:z(:y(5))); $x.a.perl.say; # why isn't $.a = `z => y => 5` ? 16:24
camelia :y(5)
jnthn ugexe: oops, that looks like a bug 16:25
ugexe: Maybe an accident in which it processed it thinking it was going to be an arglist, but then it wasn't one.
Then it was sufficiently rare to do that, that nobody noticed 16:26
16:26 patrickb left
ugexe ah. someone in #perl6 is curious why they couldnt do Foo(:$a, :$b) and it lead to this 16:27
jnthn Yeah, so I see 16:29
That's worth an issue; it's likely an easy fix 16:30
16:42 mgoebel_ left 16:48 robertle left
Geth rakudo: 67594c0b39 | (Kane Valentine)++ (committed using GitHub Web editor) | CREDITS
Update CREDITS
 16:53
kawaii AlexDaniel: CREDITS done, where in `release_guide.pod` am I editing? 16:54
lizmat kawaii++ # welcome to the club! 17:03
17:05 brrt left
kawaii thank you! 17:07
samcv [Tux], I'm curious if you can try out this and see what the speed results are github.com/MoarVM/MoarVM/pull/1072 17:11
after i do a forced push since i forgot some files that is 17:14
17:20 ufobat_ left
samcv done 17:36
17:45 brrt joined 17:51 vrurg left 18:04 brrt left
ugexe github.com/rakudo/rakudo/blob/mast...c.pm6#L178 18:14
should spawning a blocking process use a special scheduler that has no max thread limit? 18:15
seems like that would solve stackoverflow.com/questions/552651...blocks-run
because by saying i want to spawn a new blocking process i don't care about *thread* limits 18:16
or something
jnthn Uh, if you're using `start` blocks you are already dealing in threads? 18:17
But seriously, if juggling so many that the limit is reached, it's way past time to be using Proc::Async instead. 18:19
dinner & 18:20
|Tux| samv, run started 18:23
18:30 lucasb left 18:31 patrickb joined
|Tux| Rakudo version 2019.03.1-53-g67594c0b3 - MoarVM version 2019.03-35-gd15906711
csv-ip5xs0.716 - 0.720
csv-ip5xs-206.267 - 6.437
csv-parser22.994 - 23.290
csv-test-xs-200.433 - 0.437
test6.964 - 7.431
test-t1.770 - 1.889
test-t --race0.795 - 0.915
test-t-2029.747 - 31.861
test-t-20 --race9.213 - 10.268
 18:39
18:40 leont joined
and I am *not* working from home. I'll be at $work till ± 01:30
AlexDaniel kawaii: like this: github.com/rakudo/rakudo/commit/c0...2282598a8f 18:45
Geth rakudo: 0cc563c870 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/releasable/Sakefile
Add releasable Sakefile

It is ugly and somewhat unfinished, but it does the job. In fact, I've used it for every single release that I did. It is flexible enough to cut point releases.
Use with care, and tweak it to your liking.		 18:56
AlexDaniel kawaii: see this ↑ 18:57
kawaii: it's a weird piece of work… Once it was good enough to cut releases I stopped working on it, so… uh… yeah, it's ugly 19:00
but it worked great, at least for me. You decide what to do with it 19:01
patrickb: so? :)
patrickb I'm on it.
kawaii AlexDaniel: alright alright, just heading out for dinner right now but I shall take a look over later :) 19:07
AlexDaniel kawaii: cool. I can explain what that thing does whenever you feel ready
19:11 vrurg joined
tbrowder AlexDaniel: (1) thnx for all your work as release manager, (2) i wish for you smooth sailing with thesis, and (3) sorry i didn't notice impact of doc portion of recent grant request 19:22
kawaii AlexDaniel: what timezone are you by the way? I need to know when I can pester you with questions :)
AlexDaniel tbrowder: thank you, thank you! Don't worry about (3), maybe I'm wrong after all 19:23
kawaii: I don't know
kawaii: see this github.com/AlexDaniel/orgsleep
my body has a timezone of its own
tbrowder no, i think that is a valid concern given how much work by the community, esp. jj, to contunually improve the whole doc system and content. 19:25
we don't need indepent competition 19:26
*independent
AlexDaniel kawaii: and if you think that there's a stable pattern somewhere, here's the latest graph: imgur.com/a/SErYWu7 19:27
patrickb AlexDaniel: I rebased all three branches.
The PRs seem to have understood it right
AlexDaniel kawaii: so please just bug me 24/7 :)
patrickb AlexDaniel: I also did another compile and short sanity test. All seems well. 19:28
So pulling can proceed from my side. 19:29
AlexDaniel pretty sure we want to have a bump before merging it, so… 19:30
ah, already bumped, huh?
patrickb I rebased all three branches onto current master 19:31
AlexDaniel yeah, but here's what I'm thinking… chances are that most of these commits will be unbisectable, so we probably want to do bumps before the merge too, right? 19:32
patrickb Ah. Yes, a bump would be a good idea.
AlexDaniel ah, now that I checked, the last one was 4 days ago and there are no important commits after that 19:33
I think we're good
Geth nqp: e3bc18d291 | (Patrick Sebastian Zimmermann)++ (committed by Patrick Böker) | tools/build/install-moar-runner.pl
Make nqp startup script relocatable on linux.
 19:36
nqp: c5aec55873 | (Patrick Böker)++ | tools/build/install-moar-runner.pl
Remove an unused argument.
nqp: 2dad1fcbbf | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | tools/build/install-moar-runner.pl
Merge pull request #520 from patzim/reloc

Relocatable NQP
rakudo/master: 37 commits pushed by (Patrick Böker)++, (Aleks-Daniel Jakimenko-Aleksejev)++
review: github.com/rakudo/rakudo/compare/0...3f468d2c48
AlexDaniel and a bump is coming too… 19:37
weekly: Relocatable! patrickb++'s work was merged github.com/MoarVM/MoarVM/pull/1053 github.com/perl6/nqp/pull/520 github.com/rakudo/rakudo/pull/2693 19:39
notable6 AlexDaniel, Noted!
patrickb Whoot!
Now hoping the fallout won't be too big
AlexDaniel kawaii: good luck! 19:40
patrickb: now that I think about it… are there any tests? 19:41
patrickb None I wrote. 19:42
Not sure how to test this actually.
One would have to move the entire installation...
AlexDaniel I understand it may be hard to test, but then what is preventing me from breaking it tomorrow… :)
patrickb Hm... 19:44
AlexDaniel Geth: HELLO? 19:45
19:47 Geth left, Geth joined, ChanServ sets mode: +v Geth
AlexDaniel ok, Geth is refusing to announce the bump, for whatever reason 19:47
20:06 robertle joined
DrForr m: grammar foo { token A { 'a' }; token TOP { <A > }; }; say foo.parse("a"); 20:20
yoleaux 19 Mar 2019 22:14Z <jnthn> DrForr: 'fraid it'll be at least a week before I've time to look at that
camelia 5===SORRY!5=== Error while compiling <tmp>
Null regex not allowed
at <tmp>:1
------> 3r foo { token A { 'a' }; token TOP { <A 7⏏5> }; }; say foo.parse("a");
DrForr .tell jnthn No worries at all, it's by no means crucial and I might even be able to provide you a patch when I've got some spare tuits myself. 20:22
yoleaux DrForr: I'll pass your message to jnthn.
DrForr <'A'> does work and I might just use that formulation, but this feels a touch misleading. 20:25
20:27 ggoebel left
DrForr m: grammar foo { token A { 'a' }; token TOP { <'A' #`{..}> }; }; say foo.parse("a"); 20:31
camelia 5===SORRY!5===
Unrecognized regex metacharacter < (must be quoted to match literally)
at <tmp>:1
------> 3mar foo { token A { 'a' }; token TOP { <7⏏5'A' #`{..}> }; }; say foo.parse("a");
Unrecognized regex metacharacter ' (must be qu…
DrForr m: grammar foo { token A { 'a' }; token TOP { <'A'> }; }; say foo.parse("a");
camelia 5===SORRY!5===
Unrecognized regex metacharacter < (must be quoted to match literally)
at <tmp>:1
------> 3mar foo { token A { 'a' }; token TOP { <7⏏5'A'> }; }; say foo.parse("a");
Unrecognized regex metacharacter ' (must be quoted to…
DrForr Huh, the last one behaves differently on my system... 20:32
Oh, no, it doesn't, my bad.
Sorry 'bout the noise. 20:33
I still think it's LTA that <A > throws a "Null regex" error rather than "failed to find regex 'foo '" or something similar. 20:34
(this is probably my misunderstanding in general, but the error confused me, even knowing the rules surrounding the fact that <..> is for barewords, it's probably just a weird consequence of those rules. 20:44
)
21:04 travis-ci joined
travis-ci Rakudo build failed. Aleks-Daniel Jakimenko-Aleksejev '[NQP Bump] Brings 5 commits 21:04
travis-ci.org/rakudo/rakudo/builds/509107828 github.com/rakudo/rakudo/compare/7...0dcc4e9892
21:04 travis-ci left
lizmat hmmm... rakudo doesn't build on MacOS 21:22
clang: error: invalid argument '-install_name @rpath/libmoar.dylib' only allowed with '-dynamiclib'
adding "-dynamiclib" manually seems to create a perl6, but "make test" completely fails 21:27
however, running test-files separately with perl6, they seem to work 21:28
vrurg lizmat: are you compiling with XCode or GNU?
lizmat XCode
clang
vrurg hm, same here. Just finished with moar. 21:29
It compiles for me. Did you re-run Configure.pl? 21:30
lizmat yes 21:31
perl Configure.pl --gen-moar --make-install
vrurg: are you running on Mojave ? 21:34
vrurg lizmat: yes
lizmat weird
10.14.3 I presume ?
ugexe do you not need to --gen-moar=master --gen-nqp=master ? 21:35
vrurg Ok, here it goes. Yes, perl6 fails.
lizmat didn't need to before just nonw
ugexe oh nqp was already bumped 21:36
vrurg You probably already have master because the problem is likely to be caused by recent merge of relocatable moar.
lizmat ugexe: yeah, the problem is in the final ste[
*step
ugexe broke on windows too it seems -- ci.appveyor.com/project/rakudo/rak...waug#L1096 21:38
vrurg It is probably a discrepancy in options used to build moar and the debug runner.
patrickb ugh
ugexe patrickb: you were asking how to create a test -- perhaps adding the neccesary `mv rakudo/ && perl6 -e 'say 1'` to .travis.yml .appveyor and .circleci/ 21:41
patrickb I will try to reproduce the windows build error (I do have a windows available). 21:46
21:47 leont left
vrurg patrickb: wild guess: rakudo's Configure.pl must include moar::ldshared and moar::ldrpath_relocatable at line 322. 21:54
vrurg is gonna test it.
patrickb vrurg did you test with a completely clean build? - You actually managed to build and run the branch successfully some time ago. I'm unsure what changed until then... 21:58
vrurg patrickb: the point is that I was only testing moar, not the whole chain. :(
patrickb hm...
vrurg lack of time, you know... 21:59
I'm testing my guess now.
patrickb Actually I tested it on the mac of a friend. I don't have it quickly available though. :-)
:-(
vrurg NP. And done. It compiles. 22:00
lizmat: did you try just make test or make spectest? 22:01
lizmat both... but lemme double check to be sure 22:02
vrurg make test passed
lizmat aahhh I know what I did wrong 22:03
after hacking in "-dynamiclib" into the Makefile, I didn't do a "make install" but just a "make" 22:04
vrurg I think I have it fixed. Waiting for the spectest to pass.
lizmat with the "-dynamiclib" hacked in, and "make install" run, it all seems to work ok
patrickb lizmat did you change Configure.pl in the rakudo repo the way vrurg did? 22:07
lizmat no, I hacked the generated Makefile
basically I did what the error message told me :-)
vrurg Gimme a couple of minutes to complete the spectest and I'll generate a PR 22:08
lizmat vrurg++ 22:09
vrurg It would take longer – need to test norelocatable too. 22:17
patrickb vrurg: The problem is backwards.
the problem isn't the -dynamiclib being missing 22:18
it's that "-install_name @rpath/libmoar.dylib" shouldn't be there
lizmat tries that 22:20
22:20 km1 joined
vrurg patrickb: why so? I would expect all binaries to be compiled using same or similar options. 22:21
Otherwise there be incompatibilities lurking.
patrickb that flag is necessary for compiling the shared library
22:21 km1 left
patrickb but not to compile the executable needing that library 22:21
vrurg: Could you compare the build line for "moar" in the moar repo with the one building the perl6 runner? 22:22
vrurg didn't check what exactly is being compiled.
lizmat confirms that perl6 builds ok without "-install_name @rpath/libmoar.dylib" 22:23
patrickb They should be basically identical
vrurg -O3 -DNDEBUG -Wl,-rpath,@executable_path/../lib – this is for moar. 22:25
Basically, @moar::ccshared@ @rakudoshared@ must be removed from the Makefile-Moar.in then. 22:26
This basically obsoletes rakudoshared bacuse it is only used for the runners. 22:28
patrickb huh?
It's needed for libmoar
lizmat afk& 22:29
patrickb It's not needed for the moar *executable* and the perl6 runners
vrurg: Could you try reverting efb35b003fab51a74f71592729761df8ab61010f in the rakudo repo and changing nothing else? 22:30
vrurg patrickb: Makefile-Moar.in is rakudo's template, not moar's. The template is only using rakudoshared to build runners. 22:32
patrickb ah 22:33
then we are talking about the same thing
vrurg Gonna try reverting the commit now.
patrickb that commit introduced that
vrurg Comipled, testing now. 22:36
Looks ok.
22:37 leont joined
patrickb Great! 22:37
vrurg: Thanks for your quick rescue! 22:38
lizmat: You too!
vrurg patrickb: welcome! Let me test norelocatable too.
I won't be completing all the spectests. Most of them have passed ok. 22:40
patrickb: non-relocatable passed tests too. I've started spectest but isn't foreseeing any problems with it. 22:47
patrickb Great
Can you commit this?
vrurg patrickb: no. I don't have permissions. 22:48
Lizmat can. Shall I create a PR? Or you will?
patrickb If you do it'd be great. 22:49
vrurg Ok. 22:51
Geth rakudo: vrurg++ created pull request #2777:
Fix for failing build on macOS		 22:54
patrickb Thank you! 22:59
vrurg patrickb: you too! 23:00
patrickb I'm off to bed. Will look at the windows failure tomorrow. 23:05
o/
23:05 patrickb left
Geth rakudo: f9aebab539 | (Vadim Belman)++ | 2 files
Revert "First try on fixing the build on Mac."

This reverts commit efb35b003fab51a74f71592729761df8ab61010f.		 23:12
rakudo: 68c2c9fea1 | (Nick Logan)++ (committed using GitHub Web editor) | 2 files
Merge pull request #2777 from vrurg/fix_Configure_for_relocatable

Fix for failing build on macOS

Powered by Cro and the Raku® Programming Language.

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

Thank you!