Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2017-12-17

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

All times shown according to UTC.

Time Nick Message
01:20 idiosyncrat OK, now I have "We've successfully synchronized your details from GitHub.
01:20 idiosyncrat We will redirect you to your profile in a few seconds."
01:21 idiosyncrat I'm going to assume that "a few seconds" means "a hour or two,", like it did on the previous screen.
01:34 idiosyncrat Success!  A few set of Travis builds finished with all tests passing.
01:44 idiosyncrat s/few set/full set/
02:02 idiosyncrat joined #marpa
02:57 ilbot3 joined #marpa
02:57 Topic for #marpa is now Start here: http://savage.net.au/Marpa.html - Code paste/run: https://f.perlbot.pl/#marpa - Jeffrey's Marpa site: http://jeffreykegler.github.io/Marpa-web-site/ - IRC log: http://irclog.perlgeek.de/marpa/today - Youtube channel: https://www.youtube.com/channel/UCYKVfGBtfTqbs1JdYq-dc5g
05:12 ronsavage joined #marpa
05:13 ronsavage JK: Gotta love 'automation', even when it's a PITA.
11:34 ceridwen joined #marpa
12:01 amon_ joined #marpa
12:02 amon_ Re: Travis and [ci skip] https://irclog.perlgeek.de/marpa/2017-12-16#i_15589119
12:04 amon_ I absolutely don't know, but it's probably going to be some sane behaviour. I'd expect the most recent non-skip commit to be built upon push
12:05 amon_ And regarding ”wait a few seconds“: Yes, the GitHub API can sometimes be noticeably sluggish.
12:06 amon_ GH used to have response time graphs on their status page, but they got rid of those a few months back. Suspicious…
12:08 amon_ The syncing was slow for me as well.
12:08 amon_ But Travis has a nice status page where you can see how long the queue of waiting builds is: https://www.traviscistatus.com/ (during peak times, build jobs have to wait some time)
16:40 idiosyncrat joined #marpa
16:41 idiosyncrat Well,Travis, even when delayed, is still as fast as CPANtesters.
16:42 idiosyncrat Of course, when I started using CPANtesters, there was no alternative.
16:42 idiosyncrat AFAIK Travis is the first alternative to CPANtesters.
16:43 idiosyncrat Lots of archives claimed to compete with CPAN, but none of them had testing the way CPAN did, which is why none of them were anywhere near as good.
16:44 idiosyncrat amon_: More changes with G1 locations and related calls coming up.
16:45 idiosyncrat I'm grateful for your g1_to_block_last() issue, because it's forced me to rethink G1 positions, which is an important concept internally.
16:46 idiosyncrat One of the many reasons to prioritize user bugs over new development is that the bugs often alert me to conceptual issues, and save me from amassing a lot of new code on the basis of inadequately-thought-out ideas.
19:45 amon_ Moving to R3 revealed a problem in my grammar, related to rule ranks and sequence rules.
19:45 amon_ I have a PEG-ish grammar "List ::= (VAR '=' VAR / VAR '=' / VAR)*". Because of ordered choice, that grammar is unambiguous.
19:46 amon_ I'm now trying to do the same with Marpa::R2, using rule ranks and ranking_method => high_rule_only
19:46 amon_ List ::= Item*; Item ::= VAR '=' VAR rank => 2 | VAR '=' rank => 1 | VAR rank => 0
19:47 amon_ On an input "a = b" this produces the expected parse List(Item(a = b)) but also List(Item(a =) Item(b))???
19:47 amon_ Didn't notice this with R2, but R3 dies by default on ambiguous parses. Both versions always produce the exact same parses.
19:48 amon_ I see that both parses are valid, but shouldn't all but the first be discarded by the ranking method?
19:48 amon_ On an input "a = b c = d" we get the expected "List(Item(a = b) Item(c = d))" but also "List(Item(a =) Item(b) Item(c = d))" but curiously never "List(Item(a = b) Item(c =) Item(d))".
19:49 amon_ I think the cause is that Marpa doesn't evaluate rankings from left to right?
19:49 amon_ Is there a way to implement the _exact same_ language as the above PEG grammar, but with Marpa?
19:49 amon_ Here's my code: https://gist.github.com/latk/1aa3265ed938ee4e309470d648827f39
23:00 ronsavage joined #marpa

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