Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2016-06-10

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

All times shown according to UTC.

Time Nick Message
01:03 idiosyncrat_ joined #marpa
01:31 idiosyncrat_ rns: The THIF just got a (probably temporary) reprieve.  The refactoring that I thought would break it (and in which case I would not have wanted to take the time to fix it) actually did not break it.
01:32 idiosyncrat_ As long as it doesn't cause trouble, I'll leave it in. :-)
01:48 ilbot3 joined #marpa
01:48 Topic for #marpa is now Start here: http://savage.net.au/Marpa.html - Pastebin: http://scsys.co.uk:8002/marpa - Jeffrey's Marpa site: http://jeffreykegler.github.io/Marpa-web-site/ - IRC log: http://irclog.perlgeek.de/marpa/today
04:34 shadowpaste "rns" at 217.168.150.38 pasted "full_test.out -- test failures under cygwin" (1660 lines) at http://fpaste.scsys.co.uk/522632
04:36 rns jdurand: re bindings -- great!
04:37 rns idiosyncrat: re THIF -- to quote ronsavage, "I just love progress".
04:38 rns I saw some test failures under cygwin, mostly
04:38 rns Can't use an undefined value as an ARRAY reference at /home/Ruslan/Marpa--R3/cpan/blib/lib/Marpa/R3/MetaAST.pm line 74.
04:38 rns Can't use string (":start") as an ARRAY ref while "strict refs" in use at /home/Ruslan/Marpa--R3/cpan/blib/lib/Marpa/R3/MetaAST.pm line 74.
04:38 rns Can't use string (":default") as an ARRAY ref while "strict refs" in use at /home/Ruslan/Marpa--R3/cpan/blib/lib/Marpa/R3/MetaAST.pm line 74.
04:39 rns in diagnostics.
04:45 rns there is also Makefile:1076: recipe for target 'test_dynamic' failed
04:50 rns which looks strange -- fell free to tell if I need to look more into it unless it's too premature.
06:05 ronsavage joined #marpa
07:07 idiosyncrat_ jdurand: Java is hotter these days
07:12 idiosyncrat_ rns: those test failures on cygwin are worrisome -- I've no immediate idea as to what the messages mean.
07:16 rns idiosyncrat: they seem to stem from the semantics of meta_recce because they refer to the value of $ast->{top_node} not being set correctly.
07:18 rns perhaps, because my perl 5.22.2 under cygwin (unlike your under linux I presume) is threaded, that may have to do with threading. Which God forbid. :)
07:18 rns Anyway, I'll (hopefully) be able to look into them more closely this weekend, will let you know.
07:19 rns s/your/yours/
10:01 maybekoo2 joined #marpa
10:37 kaare_ joined #marpa
12:43 rgrinberg joined #marpa
13:11 rns test started failing under cygwin at commit 65f7ff7, last passing commit 2c4e568
14:04 shadowpaste "rns" at 217.168.150.38 pasted "01-lua.t meta_recce({ trace_values => 2});" (132 lines) at http://fpaste.scsys.co.uk/522687
14:04 shadowpaste "rns" at 217.168.150.38 pasted "01-lua.t meta_recce({ trace_values => 99});" (3653 lines) at http://fpaste.scsys.co.uk/522688
14:07 rns idiosyncrat: traces for 1st failing script above -- for you to look when you have time
14:13 idiosyncrat_ rns: Thanks
14:14 idiosyncrat_ I've decided to halt other stuff until we know more about this.
14:14 idiosyncrat_ That way if we have to get serious and, for example, bisect, the problem space is smaller.
14:15 idiosyncrat_ I'm also perlbrew'ing a threaded Perl 5.10.1 to be my standard development Perl.
14:16 rns Good -- I take it you haven't seen/couldn't reproduce those failures so far?
14:17 idiosyncrat_ No.
14:17 idiosyncrat_ No, I've seen no hint of them.
14:17 rns Ok -- as I said tests started failing for me at 65f7ff7 -- will try windows too.
14:19 rns BTW, 65f7ff7 is huge or I'm looking at it all wrong.
14:24 idiosyncrat_ I suspect 65f7ff7 was a move of a big block of code from one place to another -- conceptually simple, at least in a sense ...
14:24 idiosyncrat_ but it looks complex on the diff and, alas, may be complex in practice for debugging.
14:25 idiosyncrat_ We'll hope there's a simpler way. :-)
14:25 idiosyncrat_ Btw, I'm also creating a new CPAN developer's release.
14:27 idiosyncrat_ Thanks for the 2 traces -- they suggest a specific place to look.
14:27 rns just tried windows -- all tests ok, will try to update cygwin
14:27 rns re dev release -- looks like a good idea
14:28 shadowpaste "rns" at 217.168.150.38 pasted "warnings under msvc 2010" (20 lines) at http://fpaste.scsys.co.uk/522690
14:30 rns updating cygwin now -- takes some time -- will retest and let you know.
14:34 idiosyncrat_ rns: re the warnings at http://irclog.perlgeek.de/marpa/2016-06-10#i_12642386
14:34 idiosyncrat_ Once we get this other thing checked out, we should do something about those.
14:34 rns Ok, will take it.
14:35 idiosyncrat_ Don't do anything now -- until we find out about the "undef top_node" issue, I'm reluctant to take new pull requests.
14:36 idiosyncrat_ Also, the real solution might be for me to enable warnings in my development compiles -- I think I had them in Marpa::R2, but regressed when I changed the build.
14:38 rns Good -- will let you know if cygwin updating helped.
15:08 idiosyncrat_ I just run releng under a threaded 5.10.1 and everything is 100% OK.
15:10 rns Great -- hope this is cygwin-only glitch.
15:30 idiosyncrat_ Just uploaded Marpa-R3-4.001_011
15:44 idiosyncrat_ [2016-06-10T15:36:39Z] [rns] [pass] [JKEGL/Marpa-R3-4.001_011.tar.gz] [MSWin32-x86-multi-thread] [perl-v5.20.1] [00d80904-1c21-1023-a187-aa4a7abada3d] [2016-06-10T15:36:39Z]
15:44 idiosyncrat_ [2016-06-10T15:37:55Z] [rns] [fail] [JKEGL/Marpa-R3-4.001_011.tar.gz] [cygwin-thread-multi-64int] [perl-v5.22.2] [f4baa1ad-1c57-1023-80fb-f54607c966b2] [2016-06-10T15:37:55Z]
15:44 rns yes, ok under windows xp & 7, failed on cygwin under windows xp, trying cygwin under windows 7
15:44 idiosyncrat_ but you knew that :-)
15:45 rns oh, yes. :)
15:45 idiosyncrat_ btw, would you do me a favor and look into a bot which echoes out entries from the CPANtesters log onto this channel?
15:46 idiosyncrat_ when I first set up the channel moritz offered to help with bots & I told him we didn't need any, which we haven't up until now.
15:47 idiosyncrat_ [2016-06-10T15:41:54Z] [ruslan shvedov] [pass] [JKEGL/Marpa-R3-4.001_011.tar.gz] [MSWin32-x86-multi-thread] [perl-v5.22.1] [ef51c29c-6d80-1014-88c7-00044ad7832b] [2016-06-10T15:41:54Z]
15:55 maybekoo2 joined #marpa
15:58 idiosyncrat_ [2016-06-10T15:52:43Z] [Slaven Rezić (SREZIC)] [na] [JKEGL/Marpa-R3-4.001_011.tar.gz] [amd64-freebsd] [perl-v5.8.9] [5e29d8ec-2f23-11e6-b32c-ccb0c38313ad] [2016-06-10T15:52:43Z]
16:01 rns re http://irclog.perlgeek.de/marpa/2016-06-10#i_12643078 -- haven't heard about such bot, sorry
16:16 idiosyncrat_ That is, somebody would have to write it, as a shell script or some such
16:30 rns havde to go -- AFK, will backlog.
16:30 rns left #marpa
17:11 idiosyncrat_ Big series of fails for Marpa-R3-4.001_011.tar.gz on SREZIC's testbeds -- but I notice that everybody's else's modules are also failing, so perhaps it's an issue with SREZIC's testbeds.
19:29 JPGainsborough joined #marpa
19:55 rgrinberg joined #marpa
20:09 maybekoo2 joined #marpa
22:08 rgrinberg joined #marpa

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