Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-02-06

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

All times shown according to UTC.

Time Nick Message
00:24 ronsavage lucs: I did wonder about the lifetime of pastes. Thinks ... OK I've put the script on my web site and fixed the link in http://savage.net.au/Perl-modules​/html/marpa.papers/chapter3.html. It's now called null.lexeme.demo.pl
00:38 lucs ronsavage: Nice.
00:40 lucs ronsavage: In that FAQ, you talk about the distinction between G0 and G1; maybe that's pre-Marpa::R2 nomenclature? (L0 now, right?)
00:46 ronsavage lucs: No. See https://metacpan.org/pod/release/JKEGL​/Marpa-R2-2.078000/pod/Scanless/G.pod
00:51 lucs Hmm... Apart from some "discouraged" g0_,,, methods, I see only L0/G1 there.
01:10 jeffreykegler joined #marpa
01:11 jeffreykegler lucs: ronsavage: G0 is allowed for backward compatibility.  L0 is preferred, especially in material directed at newcomers.
01:12 jeffreykegler The switch, from "G0" to "L0" was motivated by the fact that the (potential) expansion of the grammar hierarchy did not go the way I originally thought.  I thought that there might be a G2 someday (and who knows?)
01:14 jeffreykegler But multiple lexers seems much more likely, and the additional ones will be named "L1", "L2", etc., so that it makes sense to call the main one, "L0".
01:15 lucs jeffreykegler: Aha, thanks.
01:15 jeffreykegler Summary: L0 == G0.  L0 is preferred.  G0 is allowed for backward compatibility, and will have some presence in my documentation probably until there is an R3.
01:16 jeffreykegler By the way, I'm hoping that by the time I'm ready for Marpa::R3, others will have jumped in the interface business, so that there never really is a Marpa::R3 SLIF.
01:33 ronsavage jeffreykegler: So it is OK to leave G1 and just change G0?
01:34 jeffreykegler ronsavage: Yes, the only name for G1 at present is G1.
01:34 jeffreykegler The idea is that G1 is structural, and grammars in the Ln series are lexers.  ("L" for lexer.)
01:36 ronsavage All: Docs updated.
01:36 jeffreykegler I was reluctant to introduce new nomenclature, especially since the G0/G1 terminology was one of my successes ...
01:36 jeffreykegler for once I used new terms and everyone intuitively seemed to pick up what I meant by them ...
01:37 jeffreykegler so (wouldn't you know it) that turns out to be the case where I'm forced to update the terminology.  Alas.
02:09 jeffreykegler left #marpa
02:44 ronsavage joined #marpa
03:05 jeffreykegler joined #marpa
16:31 lucs In the 'Semantics' POD, 'Blessings' section, it says "Implicitly, an action is not blessed.", but then a bit later, "For both RHS alternatives and lexemes, the implicit blessing is ::undef."
16:31 lucs These sound contradictory. Am I missing something?
17:25 jeffreykegler joined #marpa
17:26 shadowpaste "jeffreykegler" at 108.70.210.132 pasted "Change language re implicit blessing" (38 lines) at http://scsys.co.uk:8002/300718
17:28 jeffreykegler lucs: re http://irclog.perlgeek.de/​marpa/2014-02-06#i_8241819 -- Saying the blessing is "::undef" and saying that something is not blessed says the same thing in two different ways.
17:29 jeffreykegler I've touched up the doc to be slightly clearer.
17:29 jeffreykegler Incidentally, the talk about "implicit" vs. "explicit" options was something I came up with to deal with a common problem in docs.
17:31 jeffreykegler In docs, folks often say things have "defaults".  But very often you are allowed to specify the default behavior explicitly.  So in this case what do you call the "default default".  Most documents I read just throw around the word "default" in two different senses, expecting the reader to figure out what is meant.
17:32 jeffreykegler "Implicit" vs. "default" was my way of trying to state the distinction precisely.
17:32 jeffreykegler "Default's" are something you can set.
17:33 jeffreykegler "Implicit" behavior can never be changed by the application, only overriden.
17:50 lucs Aha, I see (re. :undef vs 'not blessed')
17:50 lucs Interesting distinction between implicit and default -- clearer indeed.
18:54 jdurand joined #marpa
19:58 * jdurand is installing freebsd on virtualbox to understand the MarpaX::Languages::ECMAScript::AST failure on this OS...
20:31 * jdurand is angry, freebsd installer fails
20:34 * jdurand is now trying to install openbsd
20:58 jdurand success with openbsd - now checking MarpaX::Languages::ECMAScript::AST on it!
21:02 jeffreykegler judrand: best of luck!
21:02 jeffreykegler left #marpa
21:28 ronsavage jdurand: So, OpenBSD installs under VirtualBox, does it? The last time I tried to install it (no VM), it asked for absolute sector numbers, grrr. Does it still do that?
21:29 jdurand not in my case, installed fine under virtualbox - I took the install54.iso and installed by http - was rather quick
22:43 jdurand Jeffrey - perl5.18 introduced set operation in character classes - Is Marpa affected (in good: supports it automatically) or bad (set operation is not visible as such when Marpa transfers the character class to perl) ?
22:47 jdurand Jeffrey, you are cited in this article: http://arstechnica.com/information-​technology/2014/01/why-does-the-gov​ernment-disallow-dynamic-languages/
23:55 jeffreykegler joined #marpa

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