Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-02-24

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

All times shown according to UTC.

Time Nick Message
06:00 ronsavage joined #marpa
06:01 ronsavage jeffreykegler: Thanx. But I meant the string literally. I assume it was a sample of a fake grammar whose usage was designed to trigger the warning. Is that right?
06:04 jdurand joined #marpa
06:06 ronsavage I.e. I'm taking an extremely naive interpretation here......
06:07 shadowpaste "jdurand" at 88.160.190.154 pasted "make releng" (16 lines) at http://scsys.co.uk:8002/312423
06:07 jdurand Jeffrey, in the nopaste upper, first issue might be for you. About case.t. To be confirmed. The others are for me.
06:19 jdurand Errata: this does not prevent me testing the latest Marpa, despite the build process is stopping. All architecture is ready, and I just have to add correct @INC to perl to get it right - so no harm
07:00 jdurand joined #marpa
07:04 jdurand Re: http://irclog.perlgeek.de/​marpa/2014-02-24#i_8334096 - confirmed last commit is ok about cpan/case.t - my personall issue seems to  be solved by installing package texlive-fonts-extra  - nice day||night everybody
07:13 jeffreykegler joined #marpa
07:14 jeffreykegler ronsavage: re http://irclog.perlgeek.de/​marpa/2014-02-24#i_8334088 -- the "inaccessible is ok by default" is part of the SLIF now ...
07:15 jeffreykegler as of its latest working version: https://github.com/jeffreykegler/Marp​a--R2/blob/master/cpan/pod/Scanless/D​SL.pod#inaccessible-symbol-statement
07:41 ronsavage jeffreykegler: Ah. Thanx.
15:14 jeffreykegler joined #marpa
16:02 jeffreykegler As already stated on the Google list, I've abandoned the idea of an "if" or "interface" option, however implemented.
16:02 jeffreykegler At this point, I thinking of creating a new CPAN indexed release -- there's significant value added --
16:02 jeffreykegler several new array descriptor options: "symbol", "name" and "lhs"
16:03 jeffreykegler the "name" adverb for rules
16:04 jeffreykegler the "inaccessible is X by default" statement;
16:04 jeffreykegler and documentation of SLIF's use of ambiguity.
17:12 jeffreykegler joined #marpa
18:01 jeffreykegler I've just uploaded the release candidate.  Here's the announcement to the Google group: https://groups.google.com/forum/​#!topic/marpa-parser/Ju_c1a0H3OQ
18:44 ronsavage joined #marpa
18:45 jeffreykegler joined #marpa
19:39 jdurand joined #marpa
19:42 jdurand Re http://irclog.perlgeek.de/m​arpa/2014-02-24#i_8337498: smoked ok on i86pc-solaris-64int perl-5.10, linux ia32 debian/unstabe
19:43 jeffreykegler judrand: thanks!
19:43 jdurand the last two versions are great, on my side there is all I need a priori to finish my work -; many thanks
19:47 shadowpaste "jdurand" at 88.160.190.154 pasted "2.081_001 warnings on i86pc-solaris-64int perl-5.10" (8 lines) at http://scsys.co.uk:8002/312923
19:53 jeffreykegler jdurand: On the web, I find: http://gcc.gnu.org/ml/gcc-​help/2011-05/msg00360.html -- apparently GCC has given up on unreacheable code warnings.
19:54 jdurand Ah really... Damned!
19:56 jdurand This go in the opposite way to people claming gcc -Wall is the power of eveyrthing in one of the latest comments at blogs.per.org
19:57 jdurand Personally I have always found the other compilers are really great, IRIX being the one that impressed me the most -;
19:58 jeffreykegler judrand: which compiler are you using to generate the warnings in the nopaste?
19:58 jdurand solaris
19:59 jeffreykegler Does it have a way to turn them off selectively?  "/* NOTREACHED */" or some such?
19:59 jdurand My test suite on MarpaX-Languages-C-AST fails with C::Scan do not know why yet
20:00 jdurand I don't know if they can be turned off
20:00 jdurand programmatically -; I doubt
20:00 jdurand $Noisufx08
20:00 jeffreykegler jdurand: I should mention by the way, that the unreachable code warnings are about Marpa code -- at the least the sample I looked at.
20:01 jdurand ok
20:01 jdurand These are just warnings - I bet we should leave the compiler do its job
20:02 jeffreykegler Let the compiler do its job is my attitude -- these are return statements at the end of subroutines which never reach the bottom, and I like to have them.
20:03 jeffreykegler But I also like to mark them.
20:03 jeffreykegler By "mark" I mean document, for example with a "/* NOTREACHED */" comment.
20:04 jdurand I googled quickly and did not found a way - ok IMHO that you put such comment
20:05 jdurand For the MarpaX-Languages-C-AST I think I know why - and this is again because of solaris
20:05 jdurand this platform is the only one I know that changes the spaces between tokens after preprocessing
20:06 jdurand so I was turning off the literal value and did not finish... anyway, I "think" this is ok
20:06 jeffreykegler Re: unreached -- I will think about it.  It's dangerous/misleading to add the comments if my compiler does not check them.
20:07 jeffreykegler Re: MarpaX-Languages-C-AST -- glad to hear it's probably not a Marpa bug.
20:07 jeffreykegler I need to go AFK now -- lunch
20:08 jdurand ok - I am looking still to sun compiler, and if afraid this could be a #pragma, if it is - have a nice lunch
20:55 shadowpaste "jdurand" at 88.160.190.154 pasted "probable non-destructive way to mark unreachable code" (42 lines) at http://scsys.co.uk:8002/312965
21:59 ronsavage The bot is still missing our recent input.
23:18 jeffreykegler joined #marpa
23:20 jeffreykegler I am pleased to note the Marpa community is giving back to CPANtesters, with test reports from both rns & jdurand
23:21 jeffreykegler jdurand: re http://irclog.perlgeek.de/​marpa/2014-02-24#i_8338624 -- let me think about this.
23:21 jeffreykegler I'm reluctant to do compiler-specific hacks, and when I do, I am biased in favor of, first ...
23:22 jeffreykegler GCC, because I use it and it is the most common; and second the MS compilers, because they are 2nd most common and ...
23:23 jeffreykegler among Marpa's power users, very common.

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