00:11 vrurg_ left, vrurg joined 02:34 vrurg_ joined 02:37 vrurg left 03:00 camelCaser left 03:01 camelCaser joined 03:09 vrurg joined 03:13 vrurg_ left 03:17 vrurg left, vrurg_ joined 06:44 patrickb joined 07:27 sena_kun joined
lizmat t/spec/S07-hyperrace/basics.t has hung on test #7 for me last night 08:06
bisectable6: my Num $target = 9 08:19
bisectable6 lizmat, Will bisect the whole range automagically because no endpoints were provided, hang tight
lizmat, Nothing to bisect!
lizmat, Output on all releases: gist.github.com/44bc00bf14074b7625...f35d643ba2
MasterDuke a spectest just completed fine for me 08:20
lizmat yeah, it did for me just now as well
sena_kun releasable6, status 08:21
releasable6 sena_kun, Next release in ≈10 hours. 3 blockers. Changelog for this release was not started yet
sena_kun, Details: gist.github.com/dc4d5d64fadd1ec907...0bbc9556ac
sena_kun No in time release this time, I'm afraid. 08:22
MasterDuke hm. did just get t/spec/S07-hyperrace/basics.t to die at 80 (it's been running in a loop with a spectest running in another tab) 08:36
valgrind got me this gist.github.com/MasterDuke17/c2ad2...06f26106c9 09:03
ah, and now a hang 09:05
gist updated with backtrace at the hang 09:14
anything else of interest while i'm in gdb? 09:15
added output of MVM_dump_bytecode 09:18
lizmat I wonder if that could be related to the new hash implementation 09:52
I haven't seen this hang / fail in a *looong* time 09:53
MasterDuke dunno. doesn't look like hash-related stuff in the backtraces though 10:23
haven't gotten it to hand/die yet with spesh disabled 10:58
11:02 Altai-man joined 11:04 sena_kun left
Geth rakudo/rakuast: a4e61e89e0 | (Elizabeth Mattijsen)++ | 5 files
Add DEPARSEing to statement prefixes

  - adjust tests for them
  - some other DEPARSE tweaks
MasterDuke got the same `Died at: emit without supply or react` even with spesh disabled 11:46
11:46 squashable6 left 11:47 squashable6 joined
Geth nqp-configure: patrickbkr++ created pull request #24:
Add `sq_escape()` macro
nqp-configure: 85ecd2d9ff | (Patrick Böker)++ | 4 files
Add `sq_escape()` macro

Helpful when filling in Perl script templates. This is required by rakudo/rakudo#3918.
nqp-configure: 48d07aa78b | (Patrick Böker)++ (committed using GitHub Web editor) | 4 files
Merge pull request #24 from patrickbkr/sq-escape

Add `sq_escape()` macro
linkable6 RAKUDO#3918 [open]: github.com/rakudo/rakudo/pull/3918 [BLOCKER][WIP] Build System: Fix RUN_RAKUDO to use the dummy Rakudo home
12:09 squashable6 left 12:10 squashable6 joined 12:55 patrickb left 13:46 patrickb joined
patrickb releasable6: status 14:53
releasable6 patrickb, Next release in ≈4 hours. 3 blockers. Changelog for this release was not started yet
patrickb, Details: gist.github.com/198a76b73f3fbac30f...f6ecdcaf7e
patrickb Altai-man: Can you have a look at github.com/rakudo/rakudo/pull/3918...-695232134 ? I value your opinion... 14:58
15:03 sena_kun joined 15:04 Altai-man left
sena_kun patrickb, replied. 15:09
patrickb sena_kun: OK. So shall I'll merge all the things now?
sena_kun patrickb, I'd say "yes" given the CI is clean. 15:10
patrickb Will do. :-)
Geth rakudo: d40586de7d | (Patrick Böker)++ | 3rdparty/nqp-configure
Bump nqp-configure

This pulls in the new `sq-escape()` macro required by a follow up commit.
rakudo: a56ce834e8 | (Patrick Böker)++ | 8 files
BUILD: Fix RUN_RAKUDO to use the dummy Rakudo home

The command to execute Rakudo during the early stages of the build process needs to have a Rakudo home set to an empty dummy directory to make sure it does not try to access the installation directory. If it would, it would create directories out side of the build directory during the compilation phase and result in failures if a previous installation is already located in the install dir. The `--rakudo-home` Configure parameter doesn't exist
  (any more?). Replace it with a perl one-liner that sets the Rakudo home via
an environment variable.
rakudo: b857aee9d6 | (Patrick Böker)++ (committed using GitHub Web editor) | 9 files
Merge pull request #3918 from patrickbkr/fix-build-rakudo-cmd-rakudo-home

Build System: Fix RUN_RAKUDO to use the dummy Rakudo home
sena_kun patrickb, also, about testing: I think El_Che had issues with relocatable builds last release and it means 1)our azure tests somewhere somehow differ from how is it built there. If possible, it will be great to investigate this case as well.
sena_kun So we will be having a bit more coverage there. 15:13
Geth rakudo: patrickbkr++ created pull request #3924:
Re-Revert PR #3913 from patrickbkr/fix-static-rakudo-home
linkable6 RAKUDO#3913 [closed]: github.com/rakudo/rakudo/pull/3913 [BLOCKER] Fix `--rakudo-home` Configure.pl parameter
patrickb sena_kun: The failures El_Che observed were the same ones I already had the PR created for. My takeaway from that incident is: If you fix something and don't intend to merge the fix right away, do create an issue as well. Then people will know that the error is known and a PR ready. 15:25
sena_kun: The difference in CI and build pipeline were also already accounted for in github.com/rakudo/rakudo/commit/f2...ae1531db96 15:27
sena_kun Oh, silly me! You noted that and I forgot, sorry to double-ask.
Well, given core devs can confirm their re-building works after update, debian ticket resolved plus profile ticket resolved and then it stays calm for a week, I'd say it's safe enuogh. 15:31
lizmat confirms it builds cleanly on MacOS Mojave and also tests clean 16:58
sena_kun patrickb++
releasable6, status 17:03
releasable6 sena_kun, Next release in ≈1 hour. 3 blockers. Changelog for this release was not started yet
sena_kun, Details: gist.github.com/248559d3494f20a55e...40fb41ec83
patrickb lizmat: Sadly that 17:28
's not a useful test yet for the error you observed last time. The merged change, was the fix, but the upcoming PR is the change that reveals the currently masked problem. 17:29
Merging now. We 17:30
'll see how it goes. :-)
Geth rakudo: 98b35f463b | (Patrick Böker)++ | 5 files
Revert "Revert "Merge pull request #3913 from patrickbkr/fix-static-rakudo-home""

Now that the `RUN_RAKUDO` build runner is fixed to no longer use the install Rakudo-home (which results in "missing or wrong version of dependency" errors during the build), we can now fix Rakudo-home handling and get the `--rakudo-home` build option working again.
This reverts commit f352b6d3309e7a957fcb7a3bf4da73ef4fbe383b.
rakudo: 8e6f7f0d2d | (Patrick Böker)++ (committed using GitHub Web editor) | 5 files
Merge pull request #3924 from patrickbkr/fix-static-rakudo-home-2

Re-Revert PR #3913 from patrickbkr/fix-static-rakudo-home
linkable6 RAKUDO#3913 [closed]: github.com/rakudo/rakudo/pull/3913 [BLOCKER] Fix `--rakudo-home` Configure.pl parameter
patrickb lizmat: Thanks for testing though!
lizmat: Can you test after this merge? 17:31
17:46 jmerelo joined
jmerelo Just closet the user survey. Three days without a new answer, so I guess enough is enough 17:50
Going to share results in no time. 17:51
sena_kun jmerelo++ 17:57
jmerelo Here's the release, where I've attached the PDF file with the summary of answers github.com/Raku/survey/releases/tag/v2020.09 18:02
First take away is: 30 answers less than last survey. So there's that
18:30 jmerelo left 18:31 travis-ci joined
travis-ci Rakudo build failed. Patrick Böker 'Merge pull request #3924 from patrickbkr/fix-static-rakudo-home-2 18:31
travis-ci.org/rakudo/rakudo/builds/728609474 github.com/rakudo/rakudo/compare/b...6f7f0d2d1f
18:31 travis-ci left
lizmat patrickb sena_kun confirmed build and test ok on MacOS Mojave 18:33
nine build.opensuse.org/package/live_bu...eed/x86_64 18:47
patrickb: ^^^
19:02 Altai-man joined 19:04 sena_kun left 20:17 patrickb left 21:50 Altai-man left