[Coke] RT: 1268; GLR: 6; JVM: 58; LHF: 1; LTA: 53; OSX: 3; PATCH: 3; PERF: 15; POD: 3; PRECOMP: 3; SEGV: 22; STAR: 1; UNI: 5; WEIRD: 2 01:10
there, that was completely automated. 01:11
timotimo nice 01:19
03:14 tomboy64 joined 03:23 tomboy64 joined
stmuk_ pl6anet.org/drop/rakudo-star-2016-04.dmg 03:46
can a mac user have a quick look at that please?
06:22 sno joined 06:52 tomboy64 joined 07:05 MadcapJake joined 07:21 JimmyZ_ joined 08:50 RabidGravy joined
dalek kudo/nom: c8bd913 | Altai-man++ | src/perl6-debug.nqp:
Ability to use a custom debugger module.
kudo/nom: 43c57f0 | Altai-man++ | src/perl6-debug.nqp:
Pass debugger at any position, not only first
kudo/nom: 23229c7 | Altai-man++ | src/perl6-debug.nqp:
Exit the arguments loop if we found debugger
rakudo/nom: f7613f7 | jnthn++ | src/perl6-debug.nqp:
rakudo/nom: Merge pull request #752 from Altai-man/nom
11:05 brrt joined 11:16 TimToady joined
[Coke] stmuk_: I was able to run hello world and p6doc. 11:31
I wonder if, in the future, if it makes sense to pre-compile star DMGs, since we know what the install dir looks like. 11:32
I can't use panda install on my own module, but I'm used to network failures when I'm at work. 11:34
Looks good. Thanks for putting that together! 11:35
timotimo awesome 11:37
11:42 TimToady joined 11:49 dalek joined
[Coke] We have a bunch of tags like RFC/REQUEST/TODO/NYI that we need to go through at some point and decide if we actually still want 'em. 11:52
(in RT)
timotimo mhm
[Coke] (should probably add STD and SPEC to those as well) 11:53
12:16 astj joined 13:14 tomboy64 joined 13:15 skids joined 13:46 tomboy64 joined
stmuk_ [Coke]: thanks! precomp sounds a good idea too 14:12
timotimo oh, yeah, precomp bites us often enough 14:13
[Tux] This is Rakudo version 2016.04-40-gf7613f7 built on MoarVM version 2016.04 14:41
test 22.292
test-t 13.351
csv-parser 23.235
timotimo uh oh 14:42
times's going up again?
masak .oO( time's up! ) 15:01
hoelzro has anyone on Windows noticed an issue where the moar runner builder doesn't work because $*DISTRO.Str eq 'mswin32 (6.3)' or something similar? 15:23
I was building a fresh rakudo on Windows last night and this happened to me; wondering if anyone else has seen it
tomboy64 i somehow broke rakudo compilation; it worked last night. bpaste.net/show/639b63cc1986 15:39
are .nqp files supposed to be installed with nqp? 15:40
hoelzro also, I set up a VM for Windows building so I could test there without rebooting, but for some reason, running perl6-m fails because it's a batch file, and /bin/sh invoked by gmake doesn't seem to like batch files. Does that sound familiar to anyone on Windows? 15:51
stmuk_ hoelzro: I built star ok with MSVC but failed with the strawberry perl mingw but I didn't see a moar runner issue on wither 16:12
hoelzro hmm 16:13
strawberry is what I'm using atm; I didn't see this before, but that was months ago
stmuk_ I think mingw failed to build moar even
I was using 32 bit
hoelzro I got the build to work once I fixed the moar runner issue 16:14
(using 64 bit)
I should probably make my VM use MSVC
stmuk_ SB P 5.22.1003 32 bit and MSVC 2013 community 16:15
17:53 hankache joined 18:08 sno joined 18:16 tomboy64 joined
[Coke] tomboy64: do you mean to be working on the jvm? 18:54
tomboy64 [Coke]: yes 19:03
[Coke]: i see the QRegex.nqp being built with nqp, but it's not getting installed along with the .jar and the .moarvm files 19:04
[Coke] it's probably getting compiled to moarvm, or included in a different file. 19:08
but I'd have to check to be sure. 19:09
someone broke make -j 19:30
bisecting... 19:32
dalek p/helpful_utf8_read_failure: 403bf33 | timotimo++ | src/HLL/Compiler.nqp:
explain b0rked utf-8 files

for example:
   Error while reading from file: Invalid UTF-8 encountered on line 3, character 15
   # oh no, a comm�xE9t
   ^ here maybe?
[Coke] so, the make problem appears to be the conversion of create-moar-runner to perl6 - it was changed to use perl6, but it's complaining that: 19:59
timotimo wait
[Coke] While looking for 'CORE.setting.moarvm': no such file or directory
timotimo i forgot a line in the initial commit
dalek p/helpful_utf8_read_failure: f245b06 | timotimo++ | src/HLL/Compiler.nqp:
actually use handle-read-failure
p/helpful_utf8_read_failure: 1c339d9 | timotimo++ | src/HLL/Compiler.nqp:
when we find nullbytes, point user towards 16bit encodings
p/helpful_utf8_read_failure: 10717b3 | timotimo++ | src/HLL/Compiler.nqp:
move default error back, return "success", handle missing utf8-c8
[Coke] Anyone mind if I revert 531a044? 20:05
(fixing it is better, but I'm happy to leave that to leont.
bartolin [Coke]: +1 (I have just seen that failure, too) 20:06
[Coke] bartolin: ok, pushed. 20:12
dalek kudo/nom: 66883a4 | coke++ | / (3 files):
Revert "Convert create-moar-runner to perl 6"

This reverts commit 531a044db162e2d0189c22df91ff2af8daac81a4.
This broke `make -j` - the perl6 command that is used needs more dependencies specified.
p/helpful_utf8_read_failure: 5dabf45 | timotimo++ | src/HLL/Compiler.nqp:
include filename (and line number) in output
20:57 sortiz joined
[Coke] btw, I mentioned this on #toolchain, but I can't rakudobrew at work. I suspect something is trying to make a direct connection somewhere (either non HTTP or not respecting my HTTP proxy settings) 21:18
RabidGravy [Coke], it appears to be working for me right now 21:29
oh, wait proxy 21:30
dunno then
I thought it only used git to grab the stuff so it should just DTRT with the proxy 21:31
22:55 skids joined
sortiz qq 22:58
23:06 geekosaur joined 23:56 synopsebot6 joined