Perl 6 - the future is here, just unevenly distributed

IRC log for #perl11, 2015-09-14

| Channels | #perl11 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:17 travis-ci perl11/cperl#39 (feature/CM-367-cperl-carp-builtin - 913abce : Reini Urban): The build has errored. https://travis-ci.org/perl11/cperl/builds/80148641
00:40 travis-ci perl11/cperl#40 (feature/gh6-no-miniperl - 8c7ca50 : Reini Urban): The build has errored. https://travis-ci.org/perl11/cperl/builds/80148654
00:51 travis-ci perl11/cperl#41 (feature/gh8-cowrefcnt-stats - 82f6952 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148669
01:06 travis-ci perl11/cperl#42 (feature/gh9-warnings-xs - 82d664c : Reini Urban): The build has errored. https://travis-ci.org/perl11/cperl/builds/80148684
01:15 travis-ci perl11/cperl#43 (feature/gh9-signatures3-gh25-nl - 820e303 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148701
01:35 travis-ci perl11/cperl#44 (bugfix/gh8-cowrefcnt - 7b8d24a : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148716
01:57 travis-ci perl11/cperl#46 (feature/gh23-inline-subs - 70ca9f7 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148744
01:58 travis-ci perl11/cperl#45 (feature/gh9-signatures - 48c240b : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148729
02:03 travis-ci perl11/cperl#47 (feature/gh14-native-types - 6839607 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148757
02:31 travis-ci perl11/cperl#48 (master - 653f0b6 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80148770
02:42 travis-ci perl11/cperl#49 (feature/gh9-signatures3-gh25-nl1 - 1544c0a : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148783
02:47 travis-ci perl11/cperl#50 (feature/CM-367-cperl-warnings-xs-carp - 23be95d : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148796
02:52 travis-ci perl11/cperl#51 (bugfix/gh26-qw-as-parens - 1e7d162 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148817
03:09 travis-ci perl11/cperl#52 (feature/gh9-signatures3 - 3433365 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148830
03:11 travis-ci perl11/cperl#53 (feature/gh24-new-hash-table - 6f0ef6c : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80148843
06:15 rurban joined #perl11
06:30 travis-ci perl11/cperl#54 (travis - 010b259 : Reini Urban): The build was canceled. https://travis-ci.org/perl11/cperl/builds/80184094
06:31 willthechill rurban: up late?
06:31 willthechill or perhaps early in your time zone
06:59 travis-ci perl11/cperl#55 (travis - 6ce2b3a : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80184610
07:33 travis-ci perl11/cperl#56 (travis - a2ca9c0 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80187324
07:36 dalek joined #perl11
08:02 rurban normal morning here
08:03 rurban still some travis quirks, ExtUtils-CBuilder is unstable there
08:04 willthechill I've never used EU::CB
08:04 willthechill don't know anything about it
08:04 rurban Good. It's horrible
08:04 rurban The only leftover from Module::Build left in core
08:05 willthechill haha
08:06 rurban My previous buildbot (cpanel internal) was better. travis has osx, but I had all bsd's, solaris, windows, linux, darwin
08:07 willthechill yes I wish travis was better, but it is all I have!
08:07 rurban there's also win buildbot, appconveyor or such
08:08 rurban https://github.com/perl11/p2/settings/hooks/4052765
08:08 willthechill how do you get macosx on travis?
08:08 rurban I pasted the line
08:08 rurban but double build times
08:13 rurban ci.appveyor.com
08:13 willthechill so appconveyor is for macosx?
08:13 rurban no, for windows
08:13 willthechill oh wow okay
08:13 rurban travis is for macosx and ubuntu
08:13 rurban no bsd's
08:18 rurban Your mohawk ticket is silly, but the guy has serious problems. Confirmed.
08:19 willthechill not silly, he is harassing me, in the workplace this would be grounds for termination
08:19 rurban In which modules does he have perms?
08:20 willthechill https://metacpan.org/author/ETJ
08:22 rurban Uh, a lot. Inline::C
08:24 willthechill apparently he has forked PDL
08:24 rurban In the perl community harassment by people with perms is accepted behavior. ether just took over a module I was fixing, and managed to break the API. complaints went backwards. hostile takeovers are accepted if it's from the module@ group.
08:24 willthechill https://metacpan.org/pod/PDLA
08:24 willthechill https://metacpan.org/pod/PDL
08:24 willthechill the Perl community needs help, and I'm not willing to ignore the situation
08:25 rurban mohawk is in the toolchains group, so no chances. they will back him.
08:25 willthechill whatever.  they can't stop me from standing up to this bully.
08:25 rurban complaints will always go backwards. they will throw the new civilty and faith arguments against you.
08:25 rurban perl5 is now officially a religion
08:26 rurban you'll see. they did with me.
08:26 willthechill yes I know it is a religion and I don't mind that part, it is the CRAPPY part of the religion that must be excised
08:27 rurban Not with RJBS. he came up with those ideas
08:27 rurban You need to wait until his term is over. Sigh
08:27 willthechill I think my github code streak has proven me to be at least somewhat patient and long-suffering
08:28 rurban yes. that's why I'm concerned about burnout. take a weekend off
08:29 rurban That's why I came up with different, perl6-like development policies: http://perl11.org/cperl/#Development-policies
08:30 bulk88 rurban, is cperl supposed to be able to build without USE_CPERL?
08:30 rurban I think so
08:30 rurban but that's undertested
08:31 willthechill rurban: your development policies issue is the same with alioth, I am dealing with him now, he is very mean
08:31 rurban I added that to make patches to upstream easier
08:31 bulk88 cl *CUT THINGS*   B.c
08:31 bulk88 const-xs.inc(329) : error C2065: 'SVf_NATIVE' : undeclared identifier
08:31 bulk88 const-xs.inc(329) : error C2099: initializer is not a constant
08:31 bulk88 NMAKE : fatal error U1077: 'cl' : return code '0x2'
08:32 rurban alioth? the python guy? he hates perl.
08:32 rurban he took over the perl script, and converted it to python
08:32 rurban bulk88: thanks
08:33 willthechill yes his initials are IG
08:33 willthechill I hereby officially predict the imminent downfall of Alioth
08:35 rurban That's going on for years. No chance. All the new fast little languages were thrown out
08:36 willthechill just wait and see
08:44 travis-ci perl11/cperl#60 (bugfix/gh32-instable-travis - 91d3cfa : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80195585
09:00 bulk88 rurban, you downgraded 5.23.0 to 5.22.1, how is cperl's version scheme supposed to work? you missed a couple places but it isn't that important ATM
09:07 rurban I took the non-breaking optimizations from 5.23 (return). As I added much more optimizations it is kosher for my release scheme
09:07 rurban but I added no API breaking changes to 5.22
09:08 rurban I tried more goodies from 5.23.x but these were not kosher
09:08 rurban 5.24.0 will have the promised features. This is just faster variant to get the discussion started
09:08 basiliscos joined #perl11
09:13 bulk88 so patches to change missed places from 5.23.0 to 5.22.1 are fine?
09:13 rurban missed places? everything which is useful to backport and does not break the API is fine
09:14 rurban even internal optimizations
09:14 rurban this is different to perl5
09:14 bulk88 https://github.com/perl11/cperl/blob/master/win32/Makefile#L8  https://github.com/perl11/cperl/blob/master/win32/Makefile#L610 and so forth
09:14 rurban oopsi
09:15 rurban I haven't tested windows at all
09:15 bulk88 i have, your master branch is a disaster, I've gotten a cperl interp built in my working dir tho
09:16 rurban those should be called cperl522
09:16 rurban a disaster?
09:17 bulk88 i assume that due to constant rebasing nature of cperl, commit sha1s are meaningless and I should only refer to commit titles?
09:17 rurban yes, please. master is pretty stable nowadays though
09:17 rurban from tomorrow on, I will not rebase master anymore
09:18 rurban I had to, to make accepting patching upstream easier.
09:19 bulk88 I dont care, I've only quoted commit titles in my branch so far
09:21 rurban where is your branch? I don't see it
09:22 rurban I haven't taken your https://github.com/bulk88/perl/tree/removecvfile yet, will do later when I work on the compiler
09:25 bulk88 removing cvfile is pointless if I ever get around to inter-perl-thread HEKs for pathnames/curcops
09:25 bulk88 you only save 1 ptr per CV then, since the GV and GV share teh sanme inter-thread HEK with the file name
09:25 rurban yes, but still, much better than before.
09:26 rurban just getting rid of this stupid _< prefix
09:26 bulk88 not necessery at all
09:26 rurban then we can share all filename HEKs
09:26 bulk88 all malloc blocks are 4 byte aligned on all platforms, just always mask off low 2 bits before
09:26 rurban bytes
09:27 bulk88 $/%/@/_< all can be stripped
09:27 rurban hmm, not sure
09:27 rurban the comppad_names?
09:27 rurban or in the GV?
09:27 bulk88 only the full string can be used for hases tho, not the offset_ok 3 bytes removed game
09:28 bpmedley joined #perl11
09:28 bulk88 IDK what comppad_names is off the top  of my head
09:31 rurban yes, no OOK hack for HEKs. they need to be properly aligned to guarantee fast hashing
09:31 bulk88 IDK if a PADNAMELIST struct was supposed to be ultimately shared between threads or not from Father C's implementation, maybe it was, since it was an AV previously
09:33 rurban they are shared, I hope so.
09:33 rurban everybody who changes it in a thread should deal with the consequences by himself
09:33 rurban type promotion is safe, name changes unimportant
09:34 bulk88 they aren't shared, new malloc block instead
09:34 rurban oh my
09:34 rurban those people…
09:35 bulk88 http://perl5.git.perl.org/perl.git/blob/HEAD:/pad.c#l2765
09:35 rurban but threads are doomed anyway
09:35 rurban they clone all the data, but cannot clone ops yet, which is needed for sub optimizations. oh my
09:36 bulk88 I dont think so, green threads/Coro are a failure, GIL is a failure
09:36 rurban green threads are pretty fast. just the Coro API is crazy
09:36 bulk88 perl's "interp context" is the only way to max out more than 1 core at a time
09:37 rurban yes, but the data needs to be COW
09:37 rurban and needs to have an owner field
09:38 bulk88 a green thread can never max out more than 1 CPU, it just hooks IO/mutex blocks and switches "context" in the same OS thread
09:38 rurban perl is fast for it's IO, not CPU
09:39 rurban perl5 ops are super slow. not much need to duplicate that slowness
09:40 willthechill not much slower than perl5 ops
09:40 willthechill interpreted C is slower
09:40 willthechill some other toy languages which are written in Perl
09:41 bulk88 the lack of locking a SV struct makes data passing comicly slow (threads::shared), they should have just created a SVt_PVSH type above SVt_PVLV that includes a mutex/owner field in the SV body and the whole SV* is always magical
09:42 willthechill OR not allowed anything but P5 into parrot so that it would not have failed
09:42 willthechill then PONIE would not have failed
09:42 willthechill then I would not have needed to create RPerl possibly
09:42 rurban lock, not copy? did anyone try that?
09:43 bulk88 5.005 threads I think? there are deprecated empty macros that do SV locks and unlocking in the headers
09:43 rurban but threaded data has it's magic fields in the mg
09:43 rurban yes, probably.
09:44 rurban 5.005 is a worthwhile tradeoff I think. locking should be faster than copying
09:45 rurban parrot failed for people reasons, not p5
09:45 rurban same as perl5 is failing now
09:46 bulk88 other VMs won't have any better solution than locking data structures between interp contexts, there are design problems when you lock a hash and that takes 100s or 1000s of lock aquisitions but that is really at the user code side, but most users wont have any idea that triggers a mass locking of a data tree :-/
09:47 rurban to lock hashes you need to go into the pairs and version the changes
09:47 rurban not easy. only clojure got that right
09:47 rurban and some bigger lisps
09:47 rurban clisp did not yet
09:47 bulk88 there are algorithms that use atomic compare excahnge, but the problem with that is, atomic compare exchange has no concept of blocking until the data is free, it doesn't exist anymore after it was taken ownership of
09:47 willthechill one of the reasons parrot failed is because they tried to make it a do-all VM like CLR
09:48 rurban parrot got the locking problem right. same as rust. moarvm not
09:50 bulk88 IDK rust or parrot, but what happens in those 2, if you dont do explicit lock aquisition on a var before modifying it? what happens to a var that is used from another thread that wasn't marked as "shared"?
09:50 bulk88 perl5 has a no crash policy, IDK about the other languages
09:51 rurban they disallow concurrent changes by not-owners
09:51 rurban such changes are deferred, jvm also
09:52 bulk88 so forgetting to lock just creates a race condition of whose's trasnactional write will complete last/win the race, but never a partial or incomplete write (a crash)?
09:53 rurban locks are implicit, a user cannot lock
09:53 rurban otherwise you are plagued with errors and deadlocks
09:54 rurban each setter needs to check ownership and locks
09:54 rurban in parrot this was compile-time
09:55 bulk88 what about aggregate data? each sub-var has its own lock or it relies on the parent container's lock? how do you make a ref to a hash slice lock safe
09:55 bulk88 ?
09:55 rurban but copying is a fine strategy also, esp. with massive parallel cores. "message passing"
09:55 rurban aggregate data was not as fine-grained as with moar. the writer had to wait
09:56 rurban lock-free, but lot of waiting
09:56 rurban not sure about the jvm
09:56 rurban they use lock bits in the data
09:56 rurban I think they call it "optimistic locking"
09:58 bulk88 message passing is best with atomics, atomic swap remove free message, fill at your liesure, atomic swap put on passing queue, signal other thread to wake up or other thread was already awake and polled/picked up the message with another atomic swap, read at your leisure, add to free queue with atomic swap
09:58 rurban Cliff Click has some good videos on youtube explaining it
09:59 rurban (author of the jvm compilers and azul optims)
09:59 willthechill JW did a good talk on P6 parallel/concurrent/async at YAPC::Asia a few weeks back
09:59 rurban atomics are pretty costly btw. ~100 cycles
09:59 willthechill there are some lessons to be learned from it
09:59 willthechill from P6
09:59 willthechill especially the async stuff y'all are talking about now
10:00 bulk88 memcpy is more expensive
10:00 rurban async/await is an API improvement, yes. but underlying it must also be consistent and fast
10:01 rurban I rather look at experienced VM authors like Hotspot or v8
10:03 bulk88 https://developer.mozilla.org/en-US/docs/Mozilla/Projects/SpiderMonkey/Internals/Thread_Safety I've written C code for this before
10:03 willthechill yes v8 is very fast but it is still not FAST
10:04 willthechill and moarVM is TRYING to be as fast as P5!  ACK!
10:05 willthechill and honestly, Perl 6 is quite an insane meta-language
10:05 willthechill I think Perl 6 can someday be useful for prototyping other languages
10:06 willthechill maybe even someday fast enough to write a not-ultra-fast language like Javascript using P6
10:06 willthechill and certainly it can be fast enough to run P5
10:07 willthechill because it is hard to get much slower than P5 already is now!
10:08 bulk88 mentioning other languages doesn't help Perl 5, if you want to use PHP or Perl 6, go use them, both are inspired by P5, neither claim source code compatibility
10:09 bulk88 same with javascript, JS is limited enough in syntax that is can cook down to JIT easily, P5 can't
10:09 willthechill actually P6 does claim eventual source code compatibility using Saint Larry's STD5.pm6
10:10 willthechill supposedly it can parse everything but some obscure v-strings and maybe a few other bits
10:10 willthechill but I don't know if there are any semantic actions yet, P6 is such a mystery
10:11 travis-ci perl11/cperl#65 (bugfix/win32 - 0c1d1a4 : Reini Urban): The build was canceled. https://travis-ci.org/perl11/cperl/builds/80210182
10:12 rurban P6 lost that battle against rpython long time ago. Now they claim their specs are no specs anymore…
10:13 rurban P5 can cook down to JIT even more easily than javascript. javascript is much more insane than p5
10:13 rurban dart and the various type efforts tried to get the insanity level lower
10:14 rurban firefox gave up on thread safety? cool
10:15 willthechill RPerl will help fix it
10:24 bulk88 firefox is single threaded, there are 100 ways 1 bad tab can freeze the browser
10:25 bulk88 there are no memory limits on a tab, so just do a JS memory leak, and firefox will go into  continous GCing, even though your tab isnt using any ops
10:28 bulk88 spidermonkey and gecko used to have stable APIs for embedding, mozilla org then decided that embedders are worthless (FF4-6ish) and removed API stability, the only client of mozilla's code is FF and TB, embedders and their tickets by policy are ignored
10:31 bulk88 firefox's perl tab CPU limits can be defeated with events or 0 ms timers
10:31 bulk88 *perl tab
10:31 travis-ci perl11/cperl#66 (bugfix/win32 - 4c97a03 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80210449
10:46 rurban chrome got it right with seperate processes per tab. no need to mess with threads
10:49 stephen joined #perl11
10:49 willthechill chrome did good
10:55 travis-ci perl11/cperl#68 (bugfix/gh32-instable-travis - 36f008d : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80214073
11:06 bulk88 rurban https://github.com/bulk88/cperl/commits/master
11:11 rurban name[sizeof("FETCH_svtype_ATTRIBUTES")] really?
11:13 willthechill MEGA_camel_CASE
11:13 willthechill hahaha
11:13 rurban a cperl522.dll is better
11:18 willthechill it has now been shown to be possible for RPerl to output .so files which can be used elsewhere
11:20 bulk88 rurban, var len is not a const expression, sizeof("FETCH_svtype_ATTRIBUTES") is
11:21 bulk88 it was in the line above
11:22 bulk88 in C99, char my_array [some_auto]; is an alloca() call in assembly
11:23 bulk88 so whether len really is const or not, for the purpose of creating my_array, whether len is const or not is an implementation detail
11:26 rurban oh my
11:27 rurban thanks a lot
11:28 rurban Now only VMS, os2, Cross + NetWare is broken
11:28 rurban I hope they will come forward somewhen
11:29 rurban for debian and redhat only Cross is relevant for the arm ports
11:30 rurban system ('copy', $Config_PM, $Config_mini); is better written as oneliner. 'copy lib\Config.pm lib\Config_mini.pl'
11:30 rurban I think
11:30 travis-ci perl11/cperl#69 (bugfix/gh32-instable-travis - d61b022 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80218039
11:31 bulk88 os2, last known build was 5.16 with some OS2 FOSS communities patches and never merged to p5p, Netware had some paid dev working on it but he never published patches, VMS, only person with a C compiler is Craig and I dont think he will respond
11:32 bulk88 I'd rather use $Config_PM $Config_mini for consistency, list vs string cat/""s IDC about
11:32 bulk88 the above code used list so I used list
11:32 rurban yes, I'll keep it asis
11:33 rurban and now I will disable the failing porting/authors.tests, aargh
11:38 bulk88 rurban, how is Config.o from ext\Config supposed to build without a Makefile.PL?
11:39 rurban It's in the main Makefile when being built dynamic
11:39 rurban as static a Makefile.PL will get created.
11:39 rurban quite fragile, yes
11:39 rurban I took the rules from EUMM
11:40 rurban solaris and bsd's were ok with it, but haven't tested windows
11:43 rurban just pushed my rebased win32 branch. I usually do g lb, which is git pull —rebase
11:43 rurban and I have a ./cp-lb which pulls all updated branches
11:45 rurban oh my. win32/config_h.PL disallows shell logic in config_h.SH
11:45 rurban the test can go
11:45 bulk88 I figured out the Config.o problem, will have a patch eventually.
11:53 rurban updated bugfix/win32
12:02 rurban and travis is finally green
12:02 travis-ci perl11/cperl#70 (bugfix/gh32-instable-travis - 1b19372 : Reini Urban): The build was fixed. https://travis-ci.org/perl11/cperl/builds/80221386
12:02 rurban I'll wait for the Config.dll hell :)
12:02 rurban biggest problems are parallel builds, but looks good now
12:07 bulk88 rurban https://github.com/perl11/cperl/commit/08868f206236ff427d7b57d41d6fbc920d32cc36#diff-e071e70f3844cf489cc1622eed9329ecL5318 no " interpolation either
12:07 bulk88 Fo..\av.obj ..\av.c
12:07 bulk88 av.c
12:07 bulk88 ..\lib\CORE\config.h(5296) : error C2008: '{' : unexpected in macro definition
12:07 bulk88 NMAKE : fatal error U1077: 'cl' : return code '0x2'
12:08 rurban I have no line 5296. Is that the PERL_HASH_FUNC_ orgy?
12:10 rurban ok, so I just go without the #if and do it in Configure
12:12 bulk88 https://github.com/perl11/cperl/commit/08868f206236ff427d7b57d41d6fbc920d32cc36#diff-e071e70f3844cf489cc1622eed9329ecL5318
12:13 rurban I removed it now
12:14 rurban Ah, I see, not fixed. {} comes from the hash_func
12:14 rurban Is WIN32 always defined?
12:15 bulk88 yes
12:15 rurban #ifndef WIN32
12:15 rurban #definePERL_HASH_FUNC_${hash_func}/**/
12:15 rurban #endif
12:15 rurban this looks better
12:15 rurban and hv_func.h does the real thing then on windows
12:15 bulk88 IDK if its a CPP or C lang syntax failure
12:21 rurban got it now
12:21 rurban added a regex to replace it
12:23 travis-ci perl11/cperl#71 (bugfix/win32 - 443412e : bulk88): The build passed. https://travis-ci.org/perl11/cperl/builds/80221412
12:24 rurban I don't like the $() shell syntax. ${} is more stable
12:25 rurban So maybe we'll get Config also done on win32 today, so it will be useful on win32 also
12:26 rurban I can do the gnu part in makefile.mk, but not the nmake part
12:35 bulk88 nmake's syntax is nearly non existant, it can put things in a temp file, no other concept from $(cmdhere $(MACROS)) in dmake exists
12:41 bulk88 https://github.com/perl11/cperl/commit/066d84e2dabfa0f6563a9cf96fdaf7d283f33a62 this didn't work
12:41 bulk88 const-xs.inc(329) : error C2065: 'SVf_NATIVE' : undeclared identifier
12:41 bulk88 const-xs.inc(329) : error C2099: initializer is not a constant
12:42 rurban I declared it for !USE_CPERL also
12:42 rurban Ah, I see
12:43 rurban have to leave for 30min
12:43 rurban add my temp work to bugfix/win32-wip
12:45 travis-ci perl11/cperl#72 (bugfix/win32 - 08868f2 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80223600
13:00 travis-ci perl11/cperl#74 (bugfix/win32 - 2c7798b : Reini Urban): The build has errored. https://travis-ci.org/perl11/cperl/builds/80226524
13:00 bulk88 SHA-1: 69b4870f8b80c9b18b196822ffa6f1a39aa84ca2
13:00 bulk88 * win32: remove the #if also in config_h.SH
13:00 bulk88 finishes "nmake all" sucessfully
13:01 bulk88 with -DUSE_CPERL
13:05 travis-ci perl11/cperl#73 (bugfix/win32 - 05c1d44 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80224607
13:27 travis-ci perl11/cperl#75 (bugfix/win32 - 69b4870 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80228379
13:31 bpmedley joined #perl11
13:32 rurban joined #perl11
13:41 rurban joined #perl11
14:06 travis-ci perl11/cperl#77 (bugfix/win32 - f98290e : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80241710
14:30 travis-ci perl11/cperl#78 (bugfix/win32-wip - fdf614d : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80241742
15:24 travis-ci perl11/cperl#79 (bugfix/gh34-no-usecperl - 5900156 : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80257659
15:45 travis-ci perl11/cperl#80 (bugfix/gh34-no-usecperl - 9da8c98 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80264849
15:59 travis-ci perl11/cperl#81 (bugfix/win32 - fd28174 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80265495
16:26 travis-ci perl11/cperl#83 (bugfix/win32-wip - 04cd416 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80270473
16:29 travis-ci perl11/cperl#82 (bugfix/gh34-no-usecperl - 909c995 : Reini Urban): The build was fixed. https://travis-ci.org/perl11/cperl/builds/80268804
16:45 rurban bulk88: should I wait for your Config.dll fixes?
16:46 basiliscos joined #perl11
16:48 travis-ci perl11/cperl#84 (bugfix/win32-wip - fbc92f2 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80275683
17:00 rurban I'll close master now. I pushed some mingw fixes, untested. have no idea about msvc. my msvc installation is broken. -Zi is invalid
17:51 travis-ci perl11/cperl#85 (bugfix/win32 - 61e04ae : Reini Urban): The build passed. https://travis-ci.org/perl11/cperl/builds/80282313
18:43 bulk88 rurbsn, no, Config.dll is getting very complicated
18:43 bulk88 rurban ^
20:19 rurban joined #perl11
20:26 travis-ci perl11/cperl#86 (bugfix/win32 - 8df5d71 : Reini Urban): The build was broken. https://travis-ci.org/perl11/cperl/builds/80314009
20:45 travis-ci perl11/cperl#87 (feature/CM-367-cperl-carp-builtin - f9980ea : Reini Urban): The build failed. https://travis-ci.org/perl11/cperl/builds/80314053
20:48 rurban https://github.com/perl11/cperl/releases released
21:08 travis-ci perl11/cperl#90 (feature/CM-367-cperl-warnings-xs-carp - de31d03 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80314143
21:11 travis-ci perl11/cperl#88 (feature/gh14-native-types - ed78012 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80314068
21:16 j0e joined #perl11
21:33 travis-ci perl11/cperl#97 (master - 9d657af : Reini Urban): The build was fixed. https://travis-ci.org/perl11/cperl/builds/80314258
21:37 travis-ci perl11/cperl#98 (feature/CM-626-cperl-use-dots - 1a0ee72 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80314289
21:48 travis-ci perl11/cperl#104 (feature/gh24-new-hash-table - 4d5b107 : Reini Urban): The build is still failing. https://travis-ci.org/perl11/cperl/builds/80314380
22:00 rurban https://www.reddit.com/r/perl/comments/3kymcg/cperl5221_released/
22:15 willthechill rurban++
22:19 travis-ci perl11/cperl#106 (cperl-5.22.1 - 9d657af : Reini Urban): The build passed. https://travis-ci.org/perl11/cperl/builds/80319026

| Channels | #perl11 index | Today | | Search | Google Search | Plain-Text | summary