00:35 lizmat joined
dalek kudo/nom: f05c77b | lizmat++ | Configure.pl:
Fix suggested by leont++
01:00 lizmat joined 01:38 astj joined 05:07 cognominal joined 05:58 sno joined 06:48 RabidGravy joined 07:05 brrt joined
Woodi m: say "foo" if v2.1 ~~ v1.2.3+ | v2.2+; # Oops: Fixes from v2.2 backported to v1.2.3... 07:49
camelia rakudo-moar f05c77: OUTPUT«foo␤»
07:51 lizmat joined
lizmat m: say "foo" if v2.1 ~~ v1.2.3+ 07:52
camelia rakudo-moar f05c77: OUTPUT«foo␤»
lizmat Woodi: aka, 1.2.3+ *will* match 2.1 07:53
masak if I understand Woodi's point, it's that it's dangerous that the `+` matches *all* higher versions, not just the higher ones on the lowest version level
I'm not sure I agree, but that's how I understand his point
lizmat ah, perhaps that's a bug then 07:54
Woodi yes, bicouse versions are not linear thing
lizmat ah, I think I know what is going on... 07:55
masak I don't know what the spec'd behavior is
I was talking morality, not legality ;)
lizmat I'm not sure this is specced more than pretty handwavy 07:56
masak I don't have time to check right now; will try later
Woodi current implementation work at least :) some science-applied version can be sone someday... 07:57
damn... I realy need to make drafts of irc messages before sending...
masak lizmat: QAH already wound down? how was it? I really wish I could've come... 07:58
lizmat: any immediate consequences for the p6 ecosystem? 07:59
brrt what, QAH was when? 08:04
(i'm missing everything this year it seems :-o)
masak brrt: Thu-Sun, methinks 08:06
brrt .... oh well 08:07
maybe in some future time we'll have hackathons and i'll have time 08:08
weird, weird reframe-jit bug
lizmat QAH is winding down 08:15
the last people are leaving...
some people are still hacking until they're thrown out 08:16
masak: we weren't ready for the QAH, really 08:17
but we connected a lot on a personal level, which should make things easier in the coming months 08:18
masak who are "we"? the Perl 6 community?
lizmat "we" as all the people at the QAH 08:19
masak the people at the QAH weren't ready for the QAH?
lizmat ah, that "we"
masak :P
lizmat sorry, yes that was ambiguious
masak yes, sorry :)
lizmat "we" as Perl 6 hackers, weren't ready for QAH
masak well, that's a data point 08:20
lizmat "we" as all the people, connected a lot
masak nodnod
lizmat I mean, we're still trying to get CUR on the road...
on the bright side, mst will be switching on P6 distribution upload on PAUSE again soon
so that blocker should be gone 08:21
masak sounds good
lizmat meanwhile, I have a P6W to prepare, and do some 7 hours of travelling ahead
leont's PR with prove6 has been merged
masak cool 08:22
lizmat nine's patch landed, allowing BEGIN EVAL $string in pre-compiled modules
jnthn has a short list of reproducible test file failures (which are usually rare flappers) 08:23
[TuxCM] This is Rakudo version 2016.04-33-gf05c77b built on MoarVM version 2016.04
test 21.481
test-t 12.744
csv-parser 23.568
brrt well, that's progress innit :-) 08:24
09:21 pmurias joined 09:48 vendethiel- joined 09:51 M-tadzik joined 10:05 pmurias joined
dalek kudo/js: a813f25 | (Pawel Murias)++ | Configure.pl:
[js] use a js- prefix for the JavaScript backend.
kudo/js: 5ada814 | (Pawel Murias)++ | tools/build/gen-js-makefile.nqp:
[js] Fooling around with compiling parts of rakudo.
kudo/js: a309d67 | (Pawel Murias)++ | src/main.nqp:
[js] Add a MAIN signature for the js backend.
kudo/js: d5435e5 | (Pawel Murias)++ | / (2 files):
[js] Build rakudo-js assuming some stuff get copied over into node_modules in a hacky manner.
kudo/update_profile_usage: 5bff97c | timotimo++ | src/Perl6/Compiler.nqp:
elaborate on different kinds of profiles for --profile
awwaiid oops. I meant my request for review to go here 12:37
hoelzro (or anyone interested in REPLness): I updated github.com/rakudo/rakudo/pull/738 so it now works with jvm. I haven't tried on windows yet to see if that is different.
timotimo github.com/rakudo/rakudo/pull/756 - this is the pull-request that goes with update_profile_usage; i'd like someone who knows more about how a usage output should be formatted to look over this 12:38
12:40 perlpilot joined 12:42 [TuxCM] joined
[Coke] (pr with prove 6) landed where? nom? 12:43
timotimo yes, nom 12:44
but you have to set an env var to actually use it
[Coke] acking the repo for prove6 doesn't show anything. 12:45
timotimo gimme a sec :)
i ... can't find it in the irc backlog :o 12:47
[Coke] ah, t/harness6 12:48
timotimo ah! just search for files with a 6 in their name, clearly
[Coke] there's a 200 in t/harness6 also, btw. 12:49
13:03 RabidGravy joined 13:17 [TuxCM] joined
[Coke] we have a few unanswered perl6-user emails if someone wants to be the face of perl6 this week. :) 13:19
timotimo butterflyface? :)
stmuk_ NativeHelpers-Blob 13:50
that fails on 32 bit Windows .. does anyone know if it works on 64 bit? 13:51
hmmm I suspect Linenoise doesn't work on any arch of Windows anyway 14:15
at least MSVC/nmake
hoelzro stmuk_: Linenoise *should* work on Windows; I can't confirm with MSVC, though 14:16
awwaiid: I saw your updates yesterday; nice work! I was a bit swamped, so I didn't get a chance to really look through 14:17
stmuk_ hoelzro: you use cygwin I guess? 14:26
hoelzro stmuk_: it's been a while since I set things up, but iirc I just use the tools that came with Strawberry Perl
stmuk_ ah I should try that .. mingw I guess 14:27
14:34 cognominal joined
timotimo stmuk_: fails in what way? "failed to find offset"? 15:12
that's something i thought was fixed a few weeks ago
stmuk_ timotimo: well I see the dependency NativeHelpers-Blob fail to install due to test failure on 32 bit MSVC and was also reading github.com/hoelzro/p6-linenoise/issues/13 15:14
windows linenoise didnt work in the Jan R* MSI (based on MSVC) either 15:15
hoelzro stmuk_: that issue has to do with building the MSI with MSVC|GCC and then installing Linenoise on a system that has the other compiler 15:16
timotimo oof, that issue was open a long time since the last activity ;(
stmuk_ I'm wondering if building the MSI is possible with the Strawberry tools since I'm not a MSVC fan 15:21
ugexe won't that just result in Linenoise failing when building with msvc? if so, whats deciding to use the wrong compiler? 15:24
MadcapJake So i just did `git pull upstream nom` applied my patch, config'd and make'd, now `./perl6 -v` says "2015.11-1620-gf05c77b" why would that be? 15:31
ugexe did you make/make install it?
if you just make (and dont `make install`) it won't work 15:32
MadcapJake yeah I just make'd 15:33
ugexe well, not that it wont work. but usually i have enough artifacts and stuff laying around that i have to
MadcapJake yeah i keep getting this dependency error too,
"Unhandled exception: Missing or wrong version of dependency 'gen/moar/stage2/QRegex.nqp' (from 'src/Perl6/Pod.nqp')" 15:34
ugexe that seems like what i would get until i `make install`
MadcapJake really wishes he could have a separate rakudo for hacking
ugexe i just put my rakudobrew perl6 at the start of PATH, and my dev perl6s somewhere else 15:36
so `make install` doesnt affect my regular perl6 incantation
MadcapJake ahh that's a good idea
make install gives me "no writeable path" error, is it because I didn't provide a prefix? 15:37
ugexe eh, i've never even used prefix so i hope not
maybe check git log to make sure you are actually on a modern commit and not that 2015 one it showed earlier 15:38
MadcapJake yep i am
ugexe hmm. well the no writable path error is from (i believe) CUR::*s 15:39
MadcapJake install-core-dist.pl it says
psch well, the version from -v is "$(cat VERSION)-{$num-of-commits-sinse-last-version}-{$git-ref}" 15:40
and if that starts with 2015.11 something is up with your VERSION file, i'd say
MadcapJake weird because `cat VERSION` returns 2016.04
geekosaur it's not just using git describe? 15:41
(which would suggest maybe a tag is checked out)
psch geekosaur: oh, yeah, no. you're right i think
geekosaur (although in that case git should be whining about detached HEAD)
psch git describe does what i said there i assume
ugexe rakudo/install/share/perl6 (this is where install-core-dists.pl saves to) 15:42
geekosaur hm, or just for some reason not on nom itself and the closest annotated tag in the history is 2015.11
MadcapJake should I remove that directory? maybe it won't overwrite?
geekosaur git status
and, maybe, git branch
MadcapJake geekosaur: on branch nom, one modified file
ugexe MadcapJake: if its a modern rakudo you should be able to delete it
psch git clean -xdf 15:43
orr maybe cd .. && rm -rf rakudo # :P
MadcapJake lol
psch well, maybe copy the patch first :)
MadcapJake ok i did git clean now I'll recompile
psch: it's a 3 letter patch :P 15:44
geekosaur hm... maybe: git fetch --tags
MadcapJake well git clean appears to have solved it! thanks all! 15:51
geekosaur actualy I think we should have thought of that first; iirc there is a known (but undiagnosed) glitch in updating git checkouts from 2015.11 to something more recent 15:54
we just haven't seen it much since 2016.01 because most people had gotten past it by then 15:55
MadcapJake ahh lol I was late to the party
how do i install modules for use with rakudo? I tried installing panda against rakudo and it still just sees all the rakudobrew modules. 16:11
RabidGravy I'm not quite sure what the distinction is 16:26
psch panda just uses the perl6 in your path, no? 16:37
16:37 [TuxCM] joined
MadcapJake zef allows me to do it just by calling the script with which rakudo you want to install against 16:39
16:39 sno joined 17:12 pmurias joined
[Coke] tries to get the RT CLI working again. 17:25
huh. "finger" no longer comes standard? (missing on hack) 18:24
geekosaur often not 18:26
vendethiel- nine_++ # no more "use Test" issues, everything works nicely :D 18:34
18:51 hankache joined 19:09 skids joined
[Coke] is glad we don't seem to have someone who is trying to contribute who feels that we're doing everything wrong. 20:44
21:11 [TuxCM] joined 21:27 sortiz joined 22:23 lizmat joined 23:27 cognominal joined
stmuk_ rakudo.org/2016/04/25/announce-raku...e-2016-04/ 23:51
23:55 cognominal joined