Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2015-05-11

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

All times shown according to UTC.

Time Nick Message
02:15 idiosyncrat A Lua indenter, in Perl: https://github.com/jeffreykegler/kollos/blob/master/components/test/lua_bf.pl
02:16 idiosyncrat I needed something to indent Lua, and found that -- it's a good example of the relationship I expect between Lua and higher level languages.
02:17 idiosyncrat The Lua indenter was originally written in Lua, but a Lua hacker wanted to change it and found it just as easy to rewrite the whole thing in Perl.
02:17 idiosyncrat I expect, in contexts where minimalism is the most important thing, that Lua will rule.
02:18 idiosyncrat But minimalism comes at a price, and for things like an indenter, where minimalism isn't important, the convenience of more extensive languages weighs in.
02:19 idiosyncrat Btw, this indenter would be a great thing to rewrite to use Marpa.
04:12 ronsavage jk: Is that a hint, a challenge or a request :-)
04:31 rns Marpa-R2-3.0 built, installed and runs my module tests ok under Windows (msvc 2010, perl 5.20.1) and cygwin (gcc 4.9.2, perl 5.14.4).
06:06 ronsavage joined #marpa
16:27 jeffreykegler joined #marpa
16:27 jeffreykegler rns: I'm working up an example of the KIR for Kollos
16:28 jeffreykegler and we should decide on the form of internal/external symbol names.
16:28 jeffreykegler I propose the following:
16:29 jeffreykegler External symbols names are allowed Lua variable names, plus those with a non-initial hyphen.
16:31 jeffreykegler KHIL internal names include an '!'.  Where an internal name is formed from an internal one, it should have the form "internal!stuff",
16:32 jeffreykegler Oops -- scracth that last line --
16:32 jeffreykegler KHIL internal names include an '!'.  Where an internal name is formed from an external one, it should have the form
16:33 jeffreykegler "external!stuff", where "external" is the external symbol name and "stuff" is the stuff added for internal purposes.
16:34 jeffreykegler KLOL internal names include a '?' and when formed from external or KHIL internal names, they are of the form "khil?klol-stuff"
16:35 jeffreykegler where "khil" is the name at the KHIL level and "klol-stuff" is stuff added at the KLOL level.
16:36 jeffreykegler KLOL internal names might be made from KHIL internal names, and would look like "external!khil-stuff?klol-stuff"
16:37 jeffreykegler Finally, I am thinking of also allowing spaces in all symbol names: external, KHIL internal and KLOL internal.
16:37 jeffreykegler This so I can have my beloved angle-bracketed names:
16:38 jeffreykegler lhs ::= short-rhs-symbol <rhs symbol whose name goes on and on and on> sym3 sym4 sym5
16:43 koo7 joined #marpa
16:53 jeffreykegler joined #marpa
17:00 idiosyncrat joined #marpa
17:00 * idiosyncrat switches to a topic of wider interest, which affects Marpa::R2 going forward.
17:01 idiosyncrat http://www.dagolden.com/index.php/2466/the-annotated-berlin-consensus/
17:01 idiosyncrat This "Berlin consensus" divides CPAN modules into up-, down- and mid-river.  Marpa::R2 is mid-river.
17:02 idiosyncrat I'm pleased to see we are already following almost of the suggested practices.
17:02 idiosyncrat Discussion and comments on Marpa::R2's compliance with these is welcome.
17:11 rns jeffreykegler: re http://irclog.perlgeek.de/marpa/2015-05-11#i_10583827 -- sounds good, I've tried to write it up below:
17:11 rns An external (LUIF) name is Lua Name or a Lua String (to allow spaces and hyphens), e.g. external, '<long name>', luif.S'long-name'
17:11 rns KHIL name is LUIF name + !khil-stuff, e.g. 'external!khil-stuff'
17:12 rns KLOL name is KHIL name + ?klol-stuff, e.g. 'external!khil-stuff?klol-stuff'
17:18 rns re http://irclog.perlgeek.de/marpa/2015-05-11#i_10584102 -- of "in the middle" best practices, minimum perl version required seems to be the most applicable.
17:25 rns continuous integration with Travis looks hardly worth the effort at the time as Marpa::R2 is semi-frozen, don't really know about documented succession plan.
17:28 koo6 joined #marpa
17:28 rns And yes, for the most part Marpa::R2 looks like a compliant midriver distribution.
19:12 lwa joined #marpa
19:33 jeffreykegler joined #marpa
19:34 jeffreykegler about the "documented succession plan" as suggested for mid-river modules --
19:34 jeffreykegler One reason I prefer "idiosyncrat" to "Benevolent Dictator for Life", is that the former does not remind me of my mortality :-)
19:36 jeffreykegler More seriously, Marpa::R2 would be a very hard transition, and I cannot think of any plan which would make things easier ...
19:36 jeffreykegler that aside from the fact that such a plan would almost certainly be totally irrelevant in 6 months time.
22:43 ronsavage joined #marpa

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