Perl 6 - the future is here, just unevenly distributed

IRC log for #inline, 2014-12-27

| Channels | #inline index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:32 sivoais mohawk: the blame is on Pod::Simple which GitHub uses. I'm linking to the module name, not a URL <https://github.com/github/markup/issues/421>
00:33 leont ingy++ davido_++ # finishing the grant
00:37 sivoais mohawk: I'm going to ++ that issue. You should too :-D
00:39 mohawk well said leont
00:41 mohawk sivoais, done
00:41 mohawk thought i think the solution lies in modifying Pod::Simple itself
00:43 sivoais mohawk: it's actually easy, there's a $Pod::Simple::HTML::Perldoc_URL_Prefix variable
00:45 mohawk yes
00:45 mohawk see my spammed comment/issue ;-)
00:45 mohawk https://github.com/theory/pod-simple/issues/61
00:46 * ingy is going to yank autostubbing shortly
00:46 ingy I'll put in a deprecation warning for a month or so
00:47 sivoais mohawk: hehe, fix it upstream, nice!
00:48 mohawk sivoais, that was my thinking ;-)
00:48 mohawk do we have a source for "considered best practice" we can link to?
00:53 sivoais mohawk: stuff like <http://szabgab.com/moving-from-sco-to-metacpan.html>, <http://szabgab.com/metacpan-is-awesome.html>
00:55 mohawk any non-szabgab urls? ;-)
00:56 mohawk ingy, mind if i update Inline's "with" doc snippet?
00:56 mohawk i'm assuming not but seems polite to ask since you're here
00:57 sivoais hmm, not sure. get people in #toolchain to write blogs
00:57 sivoais :-D
00:59 mohawk ha ha
01:00 mohawk i will see what xdg thinks
01:03 GitHub91 [inline-module-pm] ingydotnet comment on issue #28: Done. Using .inline for all.
01:03 GitHub153 [inline-module-pm] ingydotnet closed issue #28: Replace blib/Inline with something more appropriate for some frameworks. http://git.io/B7fPqA
01:04 GitHub18 [inline-module-pm] ingydotnet comment on issue #17: Autostubbing is noing away, so no longer an issue.
01:04 GitHub162 [inline-module-pm] ingydotnet closed issue #17: autostubbing not working with Module::Build http://git.io/-_sjWA
01:07 mohawk sivoais, added a link
02:28 mohawk sivoais, if you have a considered view for #61, it might be worthwhile expressing it
02:41 GitHub75 [acme-math-xs-pm] ingydotnet pushed 1 new commit to cpp: http://git.io/9mOzKQ
02:41 GitHub75 acme-math-xs-pm/cpp 1025b68 Ingy döt Net: Regenerate stub without VERSION
02:43 GitHub145 [acme-math-xs-pm] ingydotnet pushed 1 new commit to eumm: http://git.io/pWKBfg
02:45 GitHub145 acme-math-xs-pm/eumm 1670706 Ingy döt Net: Regenerate stub without VERSION
02:46 GitHub52 [acme-math-xs-pm] ingydotnet pushed 1 new commit to ext: http://git.io/e4NZYQ
02:46 GitHub52 acme-math-xs-pm/ext 2cec883 Ingy döt Net: Regenerate stub without VERSION
03:01 GitHub149 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/TKcZRA
03:01 GitHub149 inline-module-pm/master 7c9595c Ingy döt Net: Remove autostub feature....
03:03 GitHub125 [inline-module-pm] ingydotnet pushed 1 new commit to master: http://git.io/8GkAFQ
03:03 GitHub125 inline-module-pm/master 7c7eeaa Ingy döt Net: CPAN Release 0.31...
03:03 GitHub96 [inline-module-pm] ingydotnet tagged 0.31 at master: http://git.io/cDuBwQ
05:03 ingy hi bulk88, mohawk
05:03 mohawk hello
05:03 ingy if we can dep on Win32::Mutex for only MSWin32, then we can apply #32?
05:04 mohawk yes
05:04 mohawk that depends (ha) on build-time deps
05:04 mohawk which is problematic in zildo
05:04 ingy ok, when I find out how dzil does it then it's easy
05:05 mohawk the keyword will probably be dynamic_config
05:05 ingy it's only problematic, if it is in dzil
05:05 mohawk as that's what we're talking about
05:05 mohawk i assume you're a packaging etc guru now?
05:05 ingy no more than before
05:05 mohawk ok
05:06 mohawk so you know about the various phases: packaging (end of author-side), (start of install-side) configure, build, install
05:06 ingy ether: how does one in dzil, require a module depending on the $^O?
05:06 ingy mohawk: yes of course
05:06 mohawk the configure stage can write out a MYMETA.* which will then allow cpanm et al to get extra deps, if any
05:07 ingy I need this to be solved in dzil, which it probably is
05:07 mohawk so that should give enough keywords to google the right DZ plugin
05:07 mohawk because i have no idea
05:07 mohawk might be called eg "runtime deps"
05:07 ingy ether and leont are the google here
05:08 mohawk after my adventure yesterday with meta-spec 2 in EUMM, i'm now a mega-expert (ha) on this stuff
05:08 mohawk leont, YOU ARE SUMMONED
05:08 mohawk by ingy, don't blame me
05:08 mohawk ether is probably busy watching a fireplace on tv
05:09 ingy well this can happen whenever
05:09 ingy I got lots more to work on :(
05:11 ingy mohawk: is there a way for cpantesters to notify failures in IRC?
05:12 ingy maybe we can have a #inline-notify channel for all the noisy junk
05:12 mohawk ingy, sorry don't know
05:13 mohawk https://metacpan.org/pod/Dist::Zilla::Plugin::Prereqs
05:14 ingy mohawk: I don't see a solution in that url. do you?
05:18 mohawk i do not
05:19 mohawk we'll have to seek wisdom from a higher power: ether
05:24 mohawk bulk88, i also seek your wisdom on this test fail: http://www.cpantesters.org/cpan/report/aaecd1f4-6bf4-1014-936b-d7ba47a6cfb2
05:24 mohawk the first module passes, the 2nd fails
05:25 mohawk passes on my machine, naturally
05:25 mohawk so i'm after insight on windows what kind of thing might cause such a thing
05:26 ingy [OSPrereqs / MSWin32]
05:26 ingy Win32::Mutex = 0
05:27 bulk88 mohawk see my 36_fork.t patch
05:27 bulk88 *26
05:27 bulk88 that is caused by the DLL still being loaded in the process
05:27 bulk88 in I::C's case in a psuedofork
05:28 mohawk ingy, nice work
05:28 mohawk bulk88, that affects the .packlist??
05:28 mohawk sorry
05:28 mohawk i don't mean the 26 stuff, that's fine
05:28 mohawk i mean the failure on 27
05:29 bulk88 you can't delete a DLL that is loaded in the process, since if u delete the DLL,  and there is a page fault, where will the OS get the memory page from?
05:29 bulk88 SEGV!
05:29 mohawk bulk88, i'm not asking you about DLLs
05:29 bulk88 so windows doesn't allow u to delete a live DLL
05:29 mohawk please read a little further down that report
05:29 mohawk "packlist" - permission denied
05:30 bulk88 what  is test 7?
05:30 mohawk please text-search for "packlist" on that webpage
05:30 mohawk alternatively, for the word "fail"
05:30 mohawk either is good
05:32 mohawk test 7 is "make install"
05:32 mohawk the error message is all the info i have
05:32 bulk88 is .packlist Win32 file attribute Read only?
05:32 mohawk as an EUMM guy, i should probably know that
05:32 mohawk let me take a look
05:32 bulk88 cuz I know EUMM does chmod on win32 which triggers readonly flag on the file
05:33 mohawk bear in mind this passes on lots of other win32 machines
05:33 mohawk and bear in mind it just succeeded on the previous module
05:33 mohawk which would have been writing to a different location
05:33 bulk88 i dont like to say the boogeyman of virus scanner or git but maye its that
05:33 mohawk maybe it is
05:34 bulk88 it could be a hung child proc, that whas launched with open() but perl isn't reading the pipe so parent perl proc doesnt block
05:34 mohawk i feel like cargo-culting it and just reversing the list order of the modules and seeing what happens
05:34 mohawk insane that it's only on this one thing
05:35 bulk88 permission denied means something has a file lock on it or its read only, I've seen permission denied before from EUMM
05:35 bulk88 how old is the EUMM on that tester;s machine?
05:36 mohawk latest released - 7.04
05:36 mohawk latest EUI too
05:37 bulk88 i know mithaldu keeps his cpantesters sys perls git tracked (not the src, but the installed perl app),  its to clean up bad modules
05:37 mohawk https://metacpan.org/source/BINGOS/ExtUtils-Install-2.04/lib/ExtUtils/Install.pm#L842
05:37 mohawk that's the complaining line
05:38 mohawk it's ExtUtils::Packlist that's doing a croak
05:39 bulk88 im not familiar eough with stdio, but would permission denied be caused by opening a file that doesn't exist, or opening a file in a dir that doesn;t exist?
05:40 bulk88 if I hd cmd line access I'd say use cmd line handle tool or process explorer to figure out what process has a lock
05:40 mohawk https://metacpan.org/source/BINGOS/ExtUtils-Install-2.04/lib/ExtUtils/Packlist.pm#L161
05:40 bulk88 lock or open handle
05:41 mohawk it's really quite ordinary
05:41 mohawk it tries to open for writing, gets perm denied
05:41 mohawk if only there were a way to get more info
05:43 bulk88 since u wanna debug through automated releases, improve that error message by adding the "-e" or whatever for file exists and chang ethe error message to say "my $xists = -e $packfile; open($fh, ">$packfile") || Carp::croak("Can't open"($exists? "existing ":"")" file $packfile: $!");"
05:44 bulk88 also is the path over 255 chars?
05:44 mohawk ah, that might be it
05:45 mohawk i don't control the software
05:45 mohawk the bit i'm doing is "make install"
05:45 bulk88 looks like 162 chars to me
05:45 bulk88 according to me IDE but you check
05:45 mohawk the pathlength - is that a problem?
05:45 GitHub191 [inline-c-pm] ingydotnet pushed 1 new commit to master: http://git.io/H3P2UA
05:45 GitHub191 inline-c-pm/master 425eaca Ingy döt Net: CPAN Release 0.70...
05:45 GitHub48 [inline-c-pm] ingydotnet tagged 0.70 at master: http://git.io/CRbFfw
05:46 GitHub179 [inline-c-pm] ingydotnet tagged 0.63 at rm: http://git.io/HWuPIA
05:46 GitHub179 inline-c-pm/0.63 37d3474 ingy: importing cpan stuff...
05:46 GitHub179 inline-c-pm/0.63 54805ed ingy: more files...
05:46 GitHub179 inline-c-pm/0.63 f6152e9 ingy: Release ToDo files ...
05:46 bulk88 Win32 goes nuts over 255 chars
05:46 bulk88 brb
05:46 mohawk yeah, 161
05:46 ingy bulk88: can you reapply #32 over the new inline-c-pm (which now deps on Win32::Mutex for windows only)
05:47 ingy and test and such, and I'll push it out
05:48 ingy mohawk: trivial zd patch. zd is usually only as difficult as dz.
05:50 bulk88 ntfs supports 65k paths, native api supports 65k, win32 is hit or miss on over 255, explorer cant delete a file in a path over 255 u get weird error boxes ,b google the issue 4 details
05:50 bulk88 i'd improve that error message to say if the file paases exists test
05:51 ingy I think zd has 178 releases now. it will pass dz here soon enough: http://stats.cpantesters.org/leadercpan.html#top20distros
05:51 bulk88 thats a good guess if its  locked by another proc
05:51 ingy muhahahaha
05:52 bulk88 im on a phone on gprs ill look in detail when im at a real pc for the perm denied thing
05:53 ingy mohawk: I can try to integrate #32, and then you can release after bulk88 tests it
05:53 ingy if you want
05:53 ingy just make sure to install zd-1.80+
05:57 bulk88 i am not installing zd, ive never gotten dz to pass  testes on any of my dev  machinbes, ive never gotten it to build a single dz  module, also im not installing 37000 files
05:57 bulk88 *3700
06:00 GitHub134 [inline-c-pm] ingydotnet created PR/32-B (+1 new commit): http://git.io/RLxlsQ
06:00 GitHub134 inline-c-pm/PR/32-B f22516c bulk88: fix hang in test/26fork.t on Win32...
06:00 ingy bulk88: nobody asked you to install zd or dz
06:00 ingy I just asked you to test branch PR/32-B
06:01 ingy I asked mohawk to install the latest zd
06:01 bulk88 ok didnt realize tghat
06:01 ingy since he will release it (if he wants)
06:02 mohawk bulk88, you're a hero
06:02 mohawk thanks!
06:03 mohawk ingy, i'll upgrade zd now but we'll have to wait for bulk88 to be back at a real comp for his testing
06:03 ingy mohawk: git checkout master; git pull --ff-only; git rebase PR/32-B; zild release
06:03 ingy yeah I don't care. whenever
06:04 mohawk ingy, i feel like there's no rush; we'll validate it as good, then one of us can put it on master and do it the normal way
06:04 ingy yeah
06:04 mohawk i'm THIS close to finishing this EUMM metafiles thing
06:05 mohawk a process Mithaldu and bingos started in march
06:06 ingy who is the guy all up in arms because all my stuff fails on his homemade win32 perl?
06:06 ingy suni or something?
06:06 mohawk sinan unur
06:07 ingy that's it
06:07 mohawk he also refuses to specify what's going on specifically, or to make bug reports
06:07 mohawk i know your views on language, so i will instead say:
06:07 mohawk forget that guy
06:07 ingy all he has to do is give a build script of how he makes perl and someone (including me) could look into it
06:07 mohawk nah
06:07 mohawk then he'd have no sense of righteousness
06:08 ingy true
06:08 mohawk a solved problem is no fun for that type of personality
06:08 ingy ok, well if nobody else cares, then I won't either
06:09 ingy I wish there was a good article on when File::Spec->catfile matters and when it doesn't
06:09 ingy Mithaldu: ^^
06:11 mohawk i can have a go
06:11 mohawk it always matters
06:11 mohawk oh, you mean catdir vs catfile?
06:11 mohawk or what?
06:13 travis-ci ingydotnet/inline-c-pm#128 (PR/32-B - f22516c : bulk88): The build passed.
06:13 travis-ci Change view : https://github.com/ingydotnet/inline-c-pm/commit/f22516ce0fa6
06:13 travis-ci Build details : http://travis-ci.org/ingydotnet/inline-c-pm/builds/45197736
06:13 ingy just when you need File::Spec and when you can "$dir/$file"
06:14 ingy and all that jazz
06:14 mohawk so you nearly always want File::Spec
06:14 mohawk only exception i know is File::Find, which makes $File::Find::name have / regardless, which is wrong
06:14 ingy well that's what I'm saying. sometimes you don't and I'd like to know when
06:15 mohawk i did just answer that
06:15 mohawk however i will expand
06:15 ingy %INC keys don't want it
06:15 mohawk that's another exception, you're right
06:15 mohawk there are scattered other ones
06:15 ingy yes
06:15 mohawk probably - can't think of any though
06:16 mohawk safe way is use File::Spec everywhere and wait for win32 test fails ;-)
06:16 ingy not only that, but I don't want it if I don't need it
06:16 ingy yeah, forget that
06:16 mohawk you need it unless you don't care about your users
06:16 mohawk nearly everything else you need it
06:16 mohawk if you want to be all funky you could use Path::Class
06:17 ingy that's kind of a crap argument. and also I didn't like that you said we don't care about users in some issue comment
06:17 mohawk remind me which one
06:17 mohawk saying "a crap argument" is kind of inviting a hostile response, don't you think?
06:17 ingy I wish I knew because I want to delete it
06:18 ingy yeah sure, but so is saying you don't care about your users
06:18 mohawk so when you said "forget that", what did you mean?
06:19 ingy I meant I'm not going to start using File::Spec everywhere regardless of whether it's needed
06:19 mst please put the "take offense instead of considering the bug report" emo crap down and back away slowly.
06:19 ingy o hi mst
06:20 * ingy is completely befuddled as to mst's sleep hours
06:20 mst amusingly, ingy got stroppy earlier because we referred to a bug in his code as 'his fault' even though he hadn't expected that code at being installed in that context
06:21 mst mohawk: so, nah, people just want to point out problems; sometimes the recipients invent hostility rather than fixing the bug, sometimes they respond like a programmer
06:21 ingy I just wanted to understand. Once I realized it was mst's fault, I was fine. :)
06:21 mst :D
06:23 mohawk it does seem to me that code that wants to be widely used needs to deal with odd filesystems
06:24 mohawk i dare say that wants to be relied on by others, to be part of a "tool chain", to invent a phrase
06:25 ingy sure, but it isn't the #1 concern
06:25 ingy unless it happens to be at some point in time
06:25 ingy I started this off by offereing to fix the sinar issue
06:26 mohawk i've told you what the fix is
06:29 * ingy gets back to work
06:34 ingy davido_: good news, no test failures in any of the current I:M based module releases
06:59 GitHub16 [inline-module-pm] ingydotnet closed issue #24: Look at cpantesters reports for all modules http://git.io/h3jp6w
06:59 GitHub166 [inline-module-pm] ingydotnet comment on issue #24: Done. All passing!
07:00 GitHub110 [inline-module-pm] ingydotnet opened issue #29: Add module VERSION to Dynloader stub http://git.io/5WOq2w
07:01 ether bulk88: if you can't get Dist-Zilla to install on your box, please send cpantesters reports
07:02 ether or nopaste a build.log or something?
07:06 ingy ether: wishful thinking :)
07:06 ether Sinan's not ignorable. He's testing a particular combination of MSWin32 that we don't have coverage on in the cpantesters network
07:06 ether so his bug reports add value
07:06 mohawk does he make bug reports?
07:06 ingy There are a number of people who are not interested in dzil
07:07 ingy I think sinan has made some
07:09 ingy ether: as I understand it though the MSWin32 build is only used by sinan
07:09 * ingy looks for reports
07:09 ingy sec
07:10 ether he's basically testing MSVC
07:10 ether most MSWin32 reports are apparently coming from strawberry
07:11 ether I don't see how we could verify how many users are using the configuration he is testing
07:11 ingy activeperl uses MSVC afaik
07:12 ingy I don't really know of many/any people using Perl on win32 that is not Strawberry or Active
07:12 mohawk activeperl 5.18, if i remember right, told me to get mingw
07:13 mohawk so, no msvc
07:13 ingy I'm willing to help here. just saying.
07:13 mohawk using perl with msvc is valid
07:13 mohawk if we make software that breaks on that, that's on us
07:13 ingy yeah I recall Jan saying they switched
07:14 ingy but if only one person uses MSVC then?
07:14 mohawk ironically enough, i now have msvc 12 installed, but haven't made the effort to get a config working
07:14 mohawk bulk88 brilliantly made a diff of Configs, so i have the pieces to do it
07:15 mohawk msvc is clearly not a mass-market thing
07:16 mohawk but people use it
07:16 mohawk i meant perl on msvc, obviously
07:16 mohawk i'm sure loads of people use msvc
07:16 ingy so now you want to support sinan?
07:19 mohawk of course
07:19 mohawk the individual has "problems"
07:19 mohawk but he is simply the bellwether of other users
07:20 mohawk think of him as the dickish canary in a gold-mine
07:22 mohawk i should have said coal-mine there
07:22 mohawk phonetic fail
07:22 ingy http://blog.nu42.com/2014/12/int-main-in-c-and-hard-coded-directory.html
07:22 ingy https://github.com/ingydotnet/pegex-pm/issues/26
07:25 ingy https://github.com/ingydotnet/testml-pm/issues/10
07:25 mohawk that blog post wasn't as unpleasant as the HN/slashdot one
07:25 ingy I'm done with this topic for now
07:27 GitHub100 joined #inline
07:27 GitHub100 [module-build-inlinemodule-pm] ingydotnet closed issue #1: Fix autostubbing http://git.io/T01XDg
07:27 GitHub100 left #inline
07:27 GitHub40 joined #inline
07:27 GitHub40 [module-build-inlinemodule-pm] ingydotnet comment on issue #1: n/a. no more autostubbing.
07:27 GitHub40 left #inline
07:28 mohawk fair enough
07:28 GitHub0 joined #inline
07:28 GitHub0 [module-build-inlinemodule-pm] ingydotnet opened issue #2: Add `.inline/` to `./Build clean` http://git.io/-Tp_6w
07:28 GitHub0 left #inline
07:28 mohawk i think the URLs you just pasted here, he's being constructive
07:28 mohawk however, enjoy the other stuff you're doing ;-)
07:29 GitHub110 joined #inline
07:29 GitHub110 [module-install-inlinemodule-pm] ingydotnet opened issue #1: Add .inline to clean targets http://git.io/4Ngheg
07:29 GitHub110 left #inline
07:29 GitHub126 joined #inline
07:29 GitHub126 [module-build-inlinemodule-pm] ingydotnet opened issue #3: Add makestub support http://git.io/_OC-kw
07:29 GitHub126 left #inline
07:30 GitHub159 joined #inline
07:30 GitHub159 [module-install-inlinemodule-pm] ingydotnet opened issue #2: Add makestub support http://git.io/oaz4dQ
07:30 GitHub159 left #inline
07:47 bulk88 so I need to test branch PR/32-B ?
07:48 mohawk that's what i understood
07:50 ingy bulk88: please
07:52 bulk88 "prove test<enter>" is how you do it?
07:52 ingy prove -lv test
07:53 ingy you may want to install the latest Inline as well.
07:53 ingy just to be pedantic
07:54 ingy mohawk: I'll release it (if passes for bulk88)
07:57 ingy bulk88: also I'm requiring Win32::Mutex 1.09
07:58 ether ingy: re file separators, you should read the section on files in perlport.pod
07:58 ingy which seems to be the latest
07:58 ingy ok
07:58 bulk88 failed http://paste.scsys.co.uk/453100
07:58 bulk88 I havent seen file 28 fail before
07:59 ingy bulk88: the 26 output makes me think you are the wrong branch
07:59 ingy or I messed up
07:59 ingy sec
08:00 mohawk ingy, roger
08:00 bulk88 mohawk pushed the skip patch due to having a Win32:: module dep for a portable module (I::C) not being alowed
08:00 mohawk it was a temporary fix because failing test = can't easily install
08:01 ether ingy: I do agree he's quite unpleasant though :)
08:01 bulk88 https://github.com/ingydotnet/inline-c-pm/commit/256e4706885d7663d22193549bbbb647fca1fe4b
08:01 * ether was pretty annoyed by his stream of blog posts and nary an RT ticket in sight
08:02 ingy bulk88: can you get things right based on that branch?
08:03 bulk88 someone used C/++ function "erf", google says "erf" doesn't exist on VC
08:03 ingy it has your code and I guess mohawks no longer needed skips in it
08:03 ingy and we have the conditional install of Win32::Mutex now
08:06 bulk88 i've reverted the skip on file 26 locally
08:07 bulk88 push it to my repo publically or you/mohawk will write your own revert?
08:07 GitHub76 [inline-c-pm] ingydotnet pushed 1 new commit to master: http://git.io/Y41eRg
08:07 GitHub76 inline-c-pm/master ab7a556 Ingy döt Net: Update prereq versions
08:07 ingy bulk88: please push your fix
08:07 ingy you can PR it or I can just pull it
08:08 ingy I mean push a patched version of the PR/32-B branch
08:08 bulk88 retesting now
08:08 ingy ok
08:09 ingy if you can get it all passing, push that and I'll pull it in
08:10 ingy ether: I asked him for repro instructions (testml-pm)
08:10 ingy repro on his win32 build process
08:10 ingy if he plays along, so will I
08:10 ether cool
08:11 ether & # zzz
08:11 ingy o/
08:13 bulk88 http://paste.scsys.co.uk/453101  revert patch on https://github.com/bulk88/inline-c-pm/commits/master
08:13 bulk88 notice test\28autowrap.t fails
08:15 ingy bulk88: I'll try it here
08:18 bulk88 http://www.johndcook.com/blog/math_h/ erf is C99 and POSIX only, no MS compilers support it, git blamed to  https://github.com/ingydotnet/inline-c-pm/commit/f200a23aa742b062729a25cd11866d84688c2d2b
08:18 bulk88 I dont know/remember why I didn't see 28 fail before, unless I never ran the full test suite before (maybe I was just running 26_fork.t only)
08:19 ingy 28 blames out as mohawk
08:19 GitHub158 [inline-c-pm] ingydotnet pushed 1 new commit to PR/32-B: http://git.io/FM561w
08:19 GitHub158 inline-c-pm/PR/32-B cbbc690 bulk88: Revert "skip_all if WinXP as test hangs there"...
08:19 ingy bulk88: all passes here. just repushed PR/32-B
08:20 ingy can you skip 28 on win32?
08:20 ingy or we should ask mohawk what that test is for
08:20 mohawk what's the question?
08:20 ingy Test, add cookbook recipe for autowrap use to replace h2xs.
08:21 ingy mohawk: your commit f200a23aa742b062729a25cd11866d84688c2d2b fails for bulk88
08:21 ingy test/28…
08:21 mohawk hold on
08:22 ingy bulk88: 26 and 27 are ok for you?
08:22 ingy all tests pass here btw
08:22 mohawk hmm
08:23 ingy mohawk: can you skip 28 on win32 for now? on PR/32-B
08:23 bulk88 my git says ingy PR/32-B and bulk88 master are the same SHA1 so there is no point of me retesting it
08:23 ingy bulk88: yeah I don't need you to retest until we patch 28
08:23 mohawk if you're on a branch or whatever, just delete the file
08:23 bulk88 26 and 27 are fine for me, not skipped, and they passed
08:24 ingy mohawk: you want to me to remove 28?
08:24 mohawk no, i don't
08:24 mohawk it's testing the coderef facility in Inline::C
08:25 ingy ok, well what should we do?
08:25 mohawk ah, hang on
08:25 mohawk it's a unix-y function
08:25 mohawk do "man erf"
08:26 mohawk The erf() function returns the error function of x, defined as
08:26 mohawk erf(x) = 2/sqrt(pi)* integral from 0 to x of exp(-t*t) dt
08:26 ingy can you skip on win32?
08:26 ingy or I can
08:26 mohawk yes, i'll push it on master
08:26 ingy no
08:26 mohawk ok
08:26 ingy push on PR/32-B
08:27 bulk88 there are 3 choices on what to do, find a different c std lib function to use, find erf()  for Visual C src code,  or skip on Win32
08:27 mohawk what's a different stdlib func?
08:27 ingy bulk88: we can do skip for now, then fix later
08:27 mohawk oops
08:27 mohawk i know what it is
08:27 mohawk i mean, which one to use?
08:28 bulk88 why was erf picked? can't sin() or other double taking function work?
08:28 bulk88 +use Inline C => sub { q{ double erf(double); } } => enable => "autowrap";
08:28 bulk88 +like erf(1), qr/^0\.8/, "erf(1) returned 0.8-ish";
08:29 mohawk done
08:29 GitHub15 [inline-c-pm] mohawk2 pushed 1 new commit to PR/32-B: http://git.io/gt1gRQ
08:29 GitHub15 inline-c-pm/PR/32-B 9e9c8df Ed J: Skip test 28 on Win32 as no "erf" in stdlib there
08:30 mohawk that is a damn good question, i don't know
08:30 mohawk let me try a different one
08:30 bulk88 http://msdn.microsoft.com/en-us/library/634ca0c2%28v=vs.71%29.aspx pick something using VC 2003
08:32 GitHub84 [inline-c-pm] ingydotnet pushed 2 new commits to master: http://git.io/7xy-kw
08:32 GitHub84 inline-c-pm/master 6e66d45 bulk88: fix hang in test/26fork.t on Win32...
08:32 GitHub84 inline-c-pm/master 3bf4a9b bulk88: Revert "skip_all if WinXP as test hangs there"...
08:33 GitHub123 [inline-c-pm] ingydotnet deleted PR/32-B at 9e9c8df: http://git.io/I92A8w
08:33 GitHub100 [inline-c-pm] mohawk2 created PR/32-B (+4 new commits): http://git.io/rLXV9Q
08:33 GitHub100 inline-c-pm/PR/32-B f22516c bulk88: fix hang in test/26fork.t on Win32...
08:33 GitHub100 inline-c-pm/PR/32-B cbbc690 bulk88: Revert "skip_all if WinXP as test hangs there"...
08:33 GitHub100 inline-c-pm/PR/32-B 9e9c8df Ed J: Skip test 28 on Win32 as no "erf" in stdlib there
08:33 GitHub71 [inline-c-pm] ingydotnet deleted PR/31 at 6bd908a: http://git.io/xY7Rlg
08:33 GitHub192 [inline-c-pm] ingydotnet deleted PR/32 at 92df031: http://git.io/tX2Hmg
08:33 mohawk ingy, sorry i just recreated
08:34 mohawk what do you want me to do with this change?
08:34 mohawk it's worth putting on master
08:34 ingy ff master and push your change to master
08:34 bulk88 I will refresh my local repo and retest now with new PR/32-B, ok?
08:35 ingy I rebased PR/32-B and deleted that branch
08:35 ingy bulk88: wait for mohawk to push, then test master
08:36 bulk88 that branch should be rebased/merged, it skips then switch to sine1 and de-skips
08:36 mohawk done
08:36 bulk88 *squashed
08:36 GitHub69 [inline-c-pm] mohawk2 pushed 1 new commit to master: http://git.io/v55J3A
08:36 GitHub69 inline-c-pm/master b9ce6d7 Ed J: Use sin because erf not on Win32
08:36 mohawk don't release yet
08:37 mohawk part of my "skip on win32" also tidy up the TODO section so i want to just reinstate that
08:37 mohawk hold on
08:37 ingy ok
08:38 bulk88 28 passes for me now
08:41 ingy mohawk: are you set?
08:41 mohawk on 26, are we happy to TODO it?
08:41 mohawk nearly
08:41 mohawk on win32, i mean
08:41 mohawk it still runs
08:42 mohawk you know what, go ahead
08:42 mohawk the tidyup i was thinking of was a different thing
08:42 ingy mohawk: we don't need to TODO anything since all passes for bulk88
08:42 ingy ok
08:42 mohawk great
08:42 mohawk green light from me
08:42 bulk88 im still runing full test
08:43 ingy ok I killed the release
08:43 ingy waiting for bulk88
08:44 bulk88 http://paste.scsys.co.uk/453102
08:44 bulk88 everything passes for me
08:44 ingy restarting zild release
08:44 mohawk great
08:46 GitHub109 [inline-c-pm] ingydotnet pushed 1 new commit to master: http://git.io/Eq2pjQ
08:46 GitHub109 inline-c-pm/master b09d5cf Ingy döt Net: CPAN Release 0.71...
08:46 GitHub69 [inline-c-pm] ingydotnet tagged 0.71 at master: http://git.io/w8jwVg
08:46 ingy __________________________________
08:46 ingy < Inline-C-0.71.tar.gz Released!!! >
08:46 ingy ----------------------------------
08:46 ingy \   ^__^
08:46 ingy \  (oo)\_______
08:46 ingy (__)\       )\/\
08:46 ingy ||----w |
08:46 ingy ||     ||
08:46 ingy thanks guys
08:46 ingy &
08:52 mohawk nice work guys
08:52 mohawk a fantastic bit of international internet-facilitated teamwork!
08:52 mohawk bulk88, you're a hero
08:52 bulk88 the timezones involved say the developers are undead
08:53 mohawk your point being?
08:54 bulk88 international development isn't supposed to happen in real time due to timezones between eur/us/asia/aus, in this case it is working
10:02 mohawk ha ha
10:02 mohawk ingy, can we close the open PRs and issues related to t/26 on the repo?
11:51 rurban1 joined #inline
13:55 mohawk davido_, would it be possible for me to have commit bit on EUCG?
14:42 mohawk davido_, also, co-maint?
14:42 mohawk i see it hasn't been released since 2011?
19:00 willthechill joined #inline
20:30 GitHub121 [inline-c-pm] ingydotnet comment on issue #32: Fixed
20:30 GitHub2 [inline-c-pm] ingydotnet closed pull request #32: fix hang in test/26fork.t on Win32 (master...master) http://git.io/r9RIPQ
20:31 GitHub61 [inline-c-pm] ingydotnet closed pull request #31: skip tests that hangs on Windows XP (master...master) http://git.io/5hZ6Ng
20:31 GitHub103 [inline-c-pm] ingydotnet comment on issue #31: Fixed
20:32 GitHub53 [inline-c-pm] ingydotnet comment on issue #21: Fixed
20:32 GitHub79 [inline-c-pm] ingydotnet closed issue #21: 26fork.t frequently hangs on WIndows http://git.io/TX1IxA
20:39 GitHub183 [inline-c-pm] ingydotnet comment on issue #24: We keep the Pegex grammar in the share/ dir.
20:41 GitHub125 [inline-c-pm] ingydotnet comment on issue #22: I think this is handled now. Closing.
20:41 GitHub54 [inline-c-pm] ingydotnet closed issue #22: USING is not back compatible. http://git.io/1XhpwQ
22:08 GitHub73 [inline-c-pm] ingydotnet closed issue #33: Move grammar in share to a repo http://git.io/30I4-A
22:08 GitHub120 [inline-c-pm] ingydotnet pushed 2 new commits to master: http://git.io/ioNluw
22:08 GitHub120 inline-c-pm/master 4f51bf2 Ingy döt Net: git subrepo clone git@github.com:ingydotnet/inline-c-pgx ext/inline-c-pgx...
22:08 GitHub120 inline-c-pm/master 0434c1c Ingy döt Net: Fix #33
22:26 travis-ci ingydotnet/inline-c-pm#136 (master - 0434c1c : Ingy döt Net): The build was broken.
22:26 travis-ci Change view : https://github.com/ingydotnet/inline-c-pm/compare/b09d5cfcd8a0...0434c1c84a1f
22:26 travis-ci Build details : http://travis-ci.org/ingydotnet/inline-c-pm/builds/45243499
22:49 ingy bulk88: ping. look at that travis error before I rerun it. it's only on one of several runs. 26fork.t got confused on linux
22:51 ingy bulk88: well I copied it here: https://gist.github.com/36a2935e60a87c50c8ff
22:51 ingy going to try to rerun it
22:53 ingy mohawk: ^^
22:54 ingy mohawk: I think I fixed testml for sinan
22:54 ingy we'll wait and see
22:55 ingy it basically effects Inline, since Inline deps on Pegex and Pegex on TestML
22:57 ingy travis rerun PASS!
22:59 ingy travis should notify here, but guess not
23:09 willthechill joined #inline

| Channels | #inline index | Today | | Search | Google Search | Plain-Text | summary