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.
00:03 Xliff joined
Xliff . 00:03
tellable6 2019-09-09T22:42:56Z #perl6-dev <SmokeMachine> Xliff: your model has no columns...
Xliff SmokeMachine: After adding traits, I'm still getting the same error. 00:08
Geth rakudo: vrurg++ created pull request #3177:
Try not to attempt removing stale precompiled files when not needed		 00:25
rakudo: c9428a5fa0 | (Vadim Belman)++ | tools/lib/NQP/Config/Rakudo.pm
Try not to attempt removing stale precompiled files when not needed

Hopefully, fix rakudo/rakudo#3090, second issue.
synopsebot RAKUDO#3090 [closed]: github.com/rakudo/rakudo/issues/3090 [BLOCKER][build] Installer does not play well with Arch Linux’s makepkg
Geth rakudo: fce9642319 | (Vadim Belman)++ (committed using GitHub Web editor) | tools/lib/NQP/Config/Rakudo.pm
Merge pull request #3177 from vrurg/rakudo_3151

Try not to attempt removing stale precompiled files when not needed
00:29 reportable6 joined 01:31 MasterDuke left 03:19 Xliff left 03:20 Kaypie left 03:22 Kaiepi joined 04:48 Kaiepi left 05:23 squashable6 left 05:28 squashable6 joined 06:18 ufobat joined 07:04 robertle_ joined 07:23 anatofuz joined
lizmat Files=1274, Tests=109496, 204 wallclock secs (26.42 usr 7.69 sys + 2868.71 cusr 257.75 csys = 3160.57 CPU) 07:50
lizmat is at a bit of a conundrum 08:57
t/spec/*/sprintf-*.t contains tests for sprintf that I created earlier this year 08:59
it codifies errors in the nqp implementation of sprintf
I'm now working on a new Perl 6 based sprintf implementation, which should be according to the POSIX book
this now sits in a branch and I plan to make that part of 6.e 09:00
the new implementation is both more correct and more extensive than before 09:01
m: dd sprintf(q/%2$4d/,42) # this should throw, not produce wrong data 09:02
evalable6 " 0"
lizmat since these new tests were not part of 6.d, it feels I should just change them 09:03
but the new implementation is only active with "use v6.e.PREVIEW"
should I just put that in the test files until the release ??
nine I'd say do so and yes 09:04
jnthn Yes; we previously had test files with `use v6.d.PREVIEW` for non-blocking await and similar, then tossed it when we did the next language release. 09:07
lizmat ok, will take that route then 09:08
jnthn: also ok on making that 6.e specific?
the new sprintf implementation, I mean ? 09:09
jnthn lizmat: I don't know. :)
lizmat: I guess there's enough breaking changes you feel that we can't just put it in place of the original impl?
lizmat yeah, the nqp implementation has quite some quirks when mixing parameters 09:10
which make it non-POSIX compliant 09:11
jnthn OK 09:13
lizmat m: dd sprintf("%-08.2f",2.71)
evalable6 "0002.710"
jnthn Then yeah, it's probably sensible
lizmat "fatal: cannot rebase with locally recorded submodule modifications" WTF ? 09:38
ARGH 09:40
seems I've gotten myself into git submodule / branch hell 09:41
so I have a branch "new-sprintf", which I deleted locally to make sure no strange things 09:42
I build rakudo from master with fill Configure.pl
nine Does a git submodule update before the rebase help? 09:44
lizmat feels I just repeated the steps I did before, but now it works :-( 09:47
argh, no, making changes to the new Sprintf file doesn't trigger a re-compile 09:55
fatal: cannot rebase with locally recorded submodule modifications 10:01
argh
lizmat gives up for now
|Tux| Rakudo version 2019.07.1-260-g9dd67cb37 - MoarVM version 2019.07.1-188-g289e9dfce
csv-ip5xs0.672 - 0.676
csv-ip5xs-204.966 - 4.981
csv-parser20.735 - 21.451
csv-test-xs-200.425 - 0.426
test7.071 - 7.099
test-t1.721 - 1.729
test-t --race0.763 - 0.763
test-t-2029.386 - 29.518
test-t-20 --race8.991 - 9.066
 10:13
Geth rakudo/new-sprintf: bdf04746f4 | (Elizabeth Mattijsen)++ | 2 files
Second attempt at working on new sprintf
 10:28
lizmat and now any changes are not picked up in "make install" 10:44
and now it does... I guess the sha somehow wasn't changed with what I had compiled already 10:47
Question: if I only make changes in a 6.e file, why does it insist on recompiling core.c and core.d as well ?? 10:48
changes in 6.e shouldn't affect 6.c and 6.d precomps ?
Geth rakudo/new-sprintf: 3caced7ac8 | (Elizabeth Mattijsen)++ | src/core.e/Rakudo/Internals/Sprintf.pm6
zprintf -> sprintf to facilitate testing

also remove debugging code		 10:53
rakudo/new-sprintf: ae0657bf1e | (Elizabeth Mattijsen)++ | src/core.e/Rakudo/Internals/Sprintf.pm6
Another case of args-have needing an Int
 10:59
10:59 robertle_ left
Geth rakudo/new-sprintf: 5a5cc6f9a9 | (Elizabeth Mattijsen)++ | src/core.e/Rakudo/Internals/Sprintf.pm6
And yet another one
 11:00
11:01 robertle_ joined
Geth roast/new-sprintf: d4f6021df0 | (Elizabeth Mattijsen)++ | 9 files
Mark new sprintf tests as 6.e.PREVIEW
 11:07
Guest15407 lizmat++ new cool sprintf implementation 11:19
11:34 titsuki left
Guest15407 Missing or wrong version of dependency 'gen/moar/Metamodel.nqp' (from 'gen/moar/CORE.c.setting') 11:36
I just wiped my Rakudo repo and started fresh ... 11:37
no, it was my rakudbrew build that failed, sry 11:38
nine lizmat: a wrong dependency in the Makefile: $(SETTING_C_MOAR): $(PERL6_MOAR) $(PERL6_BOOTSTRAP_C_MOAR) $(M_CORE_ALL_SOURCES) 12:33
There's no reason for SETTING_C_MOAR to depend on M_CORE_ALL_SOURCES instead of M_CORE_C_SOURCES
12:35 anatofuz left 12:55 lucasb joined 13:08 anatofuz joined
vrurg Guest15407: did you resolve it? 13:11
nine: thanks for spotting, I'll get it fixed later today.
Guest15407 vrurg: yeah, I removed the old installation and rebuilt everything 13:15
vrurg Guest15407: ok. 13:16
lizmat: regarding your submodule problem, what does git status reports? 13:26
lizmat nothing anymore since ditching the branch and restarting it *after* the submodule changes 13:27
it mumbled something about "cannot rebase with locally recorded submodule modifications"
git status then said there were changes that I hadn't (consciously) made 13:28
Guest15407 vrurg, now that many changes have been made to the build does that affect how we should update to a new version of a submodule, e.g. libuv? 13:29
previously we have checked out the new version in the lib uv dir, after that we've moved to the MoarVM dir and 'git add' the entire libuv dir 13:30
does that still work? 13:31
13:31 anatofuz left
vrurg Guest15407: the changes are in rakudo build only for now. I also plan doing similar changes to NQP. libuv is a part of MoarVM where I done nothing. 13:32
Guest15407 vrurg, thx for the information 13:34
vrurg BTW, what people think about using cmake? Does it worth spending time on learning it or we'd stick to perl6 Configure? 13:36
jdv79 any idea how many more times building will be broken? i think this is the third time lately for me. 13:39
nine vrurg: what issue would a move fix? 13:41
jdv79 nopaste.xyz/?e1fbce71f249ce2e#5p8J...f9Suo4Qb0= 13:43
vrurg nine: Getting rid of perl5. Nothing else. This was discussed back in May/June, but so far no other solution is good enough for this task.
jdv79: you nqp-configure submodule didn't get updated. Though I wonder how did you get around implicit recursive submodules update which Configure.pl sets for you? 13:44
jdv79 i didn't do anything outside of what is in that paste afaik 13:45
vrurg jdv79: aha, looks like you have changes in the submodule and it rejects to update it. 13:46
M3rdparty/nqp-configure
jdv79 i made no changes
vrurg jdv79: could you cd 3rdparty/nqp-configure; git status?
jdv79 nopaste.xyz/?8021e52ffdae654b#udG+...p2p6xvVDc= 13:48
idk much about submodules but it seems an issue wrt to that
lizmat afk&
13:48 lizmat left
vrurg jdv79: try just manually run git submodule update. Should help. The problem is that git pull --rebase reject updating submodules if encounters a problem. 13:51
jdv79 ok. idk what the problem would be. i don't edit that checkout - that update routine is the extent of what i do with that checkout aside from run `perl6` so it wasn't something i did in there. 13:54
Guest15407 jdv79, I had that problem yesterday. I nuked my Rakudo install and started from scratch. 13:57
14:00 Kaiepi joined
jdv79 the submodule update seems to have worked. i would prefer to at least try to learn than just blow stuff up. 14:00
Guest15407 indeed, I just got impatient :) 14:02
vrurg jdv79: unfortunately, I'm of little help here because I don't use rebase (though I should perhaps). Just know that it's quite subtle and easily fails on submodules. 14:04
BTW, can anyone with macos try to replicate R#3100 locally?
synopsebot R#3100 [open]: github.com/rakudo/rakudo/issues/3100 [macOS] IO::Notifications for changes to file contents (2019.03 vs 2019.07.1)
Kaiepi can someone review github.com/MoarVM/MoarVM/pull/1166 ? 14:32
15:14 ufobat_ joined 15:18 ufobat left 15:19 robertle_ left 15:28 anatofuz joined
Geth problem-solving/path-to-raku: 4 commits pushed by (Jonathan Worthington)++ 15:40
16:03 anatofuz left
Geth ¦ problem-solving: ajs assigned to AlexDaniel Issue Perl 6, if renamed, needs new filneame extensions github.com/perl6/problem-solving/issues/101 16:03
¦ problem-solving: AlexDaniel assigned to jnthn Issue Perl 6, if renamed, needs new filename extensions github.com/perl6/problem-solving/issues/101 16:04
¦ problem-solving: AlexDaniel unassigned from jnthn Issue Perl 6, if renamed, needs new filename extensions github.com/perl6/problem-solving/issues/101
¦ problem-solving: AlexDaniel self-unassigned Perl 6, if renamed, needs new filename extensions github.com/perl6/problem-solving/issues/101
¦ problem-solving: AlexDaniel assigned to jnthn Issue Perl 6, if renamed, needs new filename extensions github.com/perl6/problem-solving/issues/101 16:05
16:12 lizmat joined
AlexDaniel treegrep: t6 16:19
greppable6 AlexDaniel, gist.github.com/e3cf9d22a131fb03ca...a7783eef31
Geth problem-solving/path-to-raku: 9310222f55 | (Jonathan Worthington)++ | solutions/language/PATH-TO-RAKU.md
Add a draft module author guide to the changes
 16:27
problem-solving/path-to-raku: 9285582c0a | (Jonathan Worthington)++ | solutions/language/PATH-TO-RAKU.md
Grammar tweak
17:20 MasterDuke joined 17:58 anatofuz joined
Geth ¦ problem-solving: vrurg assigned to AlexDaniel Issue Use of cmake for build subsystem github.com/perl6/problem-solving/issues/102 18:13
18:32 anatofuz left
Geth ¦ problem-solving: vrurg assigned to jnthn Issue Roles, submethods, constructors, destructors. github.com/perl6/problem-solving/issues/103 18:54
[Coke] github.blog/2019-05-23-announcing-...en-source/ 18:58
Followed by 18:59
github.blog/2019-09-10-acceleratin...sors-beta/
19:00 Kaypie joined, Kaiepi left 19:02 anatofuz joined 19:09 Kaypie is now known as Kaiepi 19:10 AlexDaniel left, leont joined 19:11 AlexDaniel joined, AlexDaniel left, AlexDaniel joined 19:16 anatofuz left 19:30 Ven`` joined 19:41 AlexDaniel left, AlexDaniel joined, AlexDaniel left, AlexDaniel joined 19:57 lizmat left 19:59 lizmat joined, lizmat left 21:12 anatofuz joined 22:06 lizmat joined 22:08 Ven`` left 22:24 leont left 22:33 anatofuz left 23:05 lucasb left 23:08 anatofuz joined
Geth roast: vrurg++ created pull request #576:
Add test for file change events.		 23:37

Powered by Cro and the Raku® Programming Language.

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

Thank you!