Camelia, the Perl 6 bug

IRC log for #moe, 2013-04-09

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

All times shown according to UTC.

Time Nick Message
00:38 * rjbs is not pleased with the next piece of homework, whose description is very unclear.
01:09 awwaiid those parenthesis are like placing the wrong emphAhasis on the wrong sylAhble
01:18 jasonmay oh right, new week, new assignments
01:20 stevan_ awwaiid: the messed up thing is that if he didn't have the type there, it would have just worked
01:21 stevan_ actually, oddly enough, it wouldnt
01:25 stevan_ rjbs: why do it the easy way when you could do it the exhastive brute force way
01:25 stevan_ def is_odd (n: Int): Boolean = if (n == 0) false else is_even(n - 1); def is_even (n: Int): Boolean = if (n == 0) true else is_odd(n - 1)
01:25 stevan_ mutually recusive predicates FTW
01:25 stevan_ THIS
01:25 stevan_ IS
01:25 stevan_ FUNCTIONAL PROGRAMMING!
01:26 * stevan_ kicks nperez into the pit
01:28 rjbs is_odd = 1
01:28 rjbs right pretty often
01:29 stevan_ well at least 50% of the time anyway
01:29 rjbs possibly more, hard to say
01:29 stevan_ wont get better odds then that in Vegas baby
01:38 jasonmay were return types required for that?
01:38 * jasonmay test
01:38 jasonmay s
01:40 sahadev_ joined #moe
01:41 jasonmay yup, no matter how much indirection
01:41 jasonmay makes sense
01:44 stevan_ jasonmay: yeah, they were
01:44 stevan_ I suspect it was because of the mutual recursion
02:14 rjbs Your overall score for this assignment is 10.00 out of 10.00
02:15 rjbs This is like the worst diagnostic ever, guys: [Observed Error] org.scalatest.exceptions.TestFailedException: "{[2,4,5,6,8]}" did not equal "{[0,2,3,4,6,999]}"
05:32 hiratara joined #moe
06:37 tobyink joined #moe
09:59 hiratara joined #moe
11:09 tobyink joined #moe
12:08 jnap joined #moe
12:50 gizmomathboy joined #moe
12:59 stevan joined #moe
13:00 jasonmay am I in the minority when I say I would like method metadata done in the vein of how java implements modifiers?
13:00 jasonmay @or whatever those things are called
13:00 jasonmay @foo
13:01 jnap annotations?
13:01 jasonmay that's it
13:01 jasonmay but not necessarily the exact syntax
13:01 jnap I wouldn't mind a sort of superpowered method attributes
13:01 jasonmay just the idea that all the data is not cramming itself between the declaration and signature
13:01 jasonmay or between the signature and block
13:01 jnap I'd love that for CataMoe
13:02 stevan I don't like how it is above the method
13:02 jnap yeah that always felt like a hack
13:03 jnap I don't think Perl's approach is bad personally, just not complete and it has too many wonky behaviors
13:03 jnap too many unexpected behaviors if you are a perl programmer I mean
13:05 stevan I like the Perl 6 attributes
13:05 stevan sub foo is public { ... }
13:05 stevan sub bar is fucking(awesome) { ... }
13:06 jnap could we find a way to split the difference such that backcompat is not impossible
13:07 stevan back compat is pretty impossible
13:07 stevan perl 5 attributes are complete shit, both in implementation and API
13:07 jnap or… ug support both
13:07 stevan why?
13:07 stevan why support a shitty API and a bad implemenation
13:08 stevan we could support syntax maybe
13:08 stevan sub bar :fucking(awesome) { ... }
13:08 stevan sub foo :public { ... }
13:08 jnap I guess I don't hate the current so much, but then again I don't need to hack on the core of it,  just find what we do with Catalyst seems to work ok
13:09 jnap yeah, partial support would be fine, most people don't use them much, outside of Catalyst AFAIK
13:11 stevan I think we could have something very close to it, with the same root capabilities ... then have so much more on top of it
13:14 jnap I guess I need to reread the Perl6 spec, not sure what the more would be, but as long as it doesn't end up superpowering methods so much that it confused people regarding how to compose logic is cool
13:16 stevan yeah, I will do my best ot keep it sane :)
13:17 jnap stevan:  doing hacking at Baltimore workshop?
13:17 stevan jnap: I plan on some, but we are doing training on the second day, so it wont be a huge amount
13:18 jnap I can't stay late on Sunday anyway, can't leave my wife alone on the farm for too long
13:19 stevan yeah I am staying over until monday, cause the train choices are pretty bad for sunday night
13:19 jnap Not sure if driving yet or not
13:21 jnap my best friend from China might be coming, is a bit of a wild card, since his dad is really sick and lives in Rochester.  If he comes I'll meet him in Rochester and drive down, otherwise train it one way or another, since I get wifi on the train and can work the day
13:46 jnap joined #moe
14:21 isBEKaml joined #moe
14:25 isBEKaml stevan: I have a niggling feeling that that guy was trolling us on the latest issue - First, he started off with the "Misleading" thing, later jumped to syntactic quibbles. Should we direct him to our wiki instead so he can focus on getting his issues out onto some page we can see?
14:26 isBEKaml stevan: (I'm willing to give him the benefit of doubt since I can see English is not his native language)
14:26 stevan isBEKaml: no, I know him
14:26 stevan he is a japanese Perl guy
14:26 stevan he had similar issues with Moose
14:26 stevan he is not trolling
14:27 isBEKaml stevan: Ah, okay - that settles it, then. :)
14:32 lestrrat stevan/isBEKaml: um, just kind of, um, don't take that guy, um, too seriously...
14:33 isBEKaml lestrrat: was that you? If so, I kind of feel weird saying "Welcome to Moe!" :)
14:33 lestrrat no way
14:33 lestrrat I speak English ;P
14:33 perigrin isBEKaml: no lestrrat has better English than I do.
14:34 perigrin also he isn't crazy :)
14:34 isBEKaml :)
14:34 isBEKaml We all are crazy. :)
14:34 prammer joined #moe
14:34 perigrin (most of the time, I'm still wondering about taking on 0mq support)
14:34 isBEKaml that guy had an escape latch to say "not crazy" :P
14:35 lestrrat I just happen to be Japanese and his, um, odd twitter posts all come up in my twitter search for "perl lang:ja" all the time
14:35 jasonmay this early in the game I don't think anything short of insults could be considered trolling
14:36 isBEKaml jasonmay: I think I was caught off by how he was jumping all over the field...
14:37 perigrin isBEKaml: that's okay I read a page today that insinuated Moose was just a fad.
14:38 perigrin so jumping all over the place regarding Moe doesn't even break the top ten of "weird world views"
14:38 isBEKaml "ha ha, ummm.. I got stuff to do, have fun!" (paraphrasing steve yegge)
14:38 stevan perigrin: Moose is a fad, all the cool kids have moved onto Moo
14:39 lestrrat Just my $.02: I think he means well, but I don't think he can really be convinced against his vision. so just don't put too much effort.
14:39 perigrin stevan: Moo? Screw that ... M is where it's at.
14:39 stevan perigrin: M for minimalist
14:39 * perigrin gives it a Mondarin for a logo.
14:40 moe [moe] ribasushi opened pull request #77: Avoid confusing curious bystanders (master...patch-1)  http://git.io/fMgxTA
14:40 jnap I wish we had a sane core so we didn't have all these problems :(
14:41 jasonmay that was a fast merge
14:41 lestrrat anyway, I was just intrigued which one of my crazy Japanese friends filed an issue that could possibly be interpreted as trolling.
14:41 moe [moe] stevan pushed 2 new commits to master: http://git.io/kB7Mww
14:41 moe moe/master 57b0b3b Peter Rabbitson: Avoid confusing curious bystanders
14:41 moe moe/master ff028a9 Stevan Little: Merge pull request #77 from ribasushi/patch-1...
14:41 * lestrrat goes back to lurking.
14:41 stevan he is right, why not
14:42 perigrin jnap: you gotta change 34 years of development then.
14:42 stevan lestrrat: I know he wasn't trolling, which is why I took the time to answer him :)
14:43 lestrrat yeah, I know. I was interested. turns out it wasn't even a friend, but anyway
14:43 * lestrrat REALLY goes back to lurking.
14:43 stevan :)
14:52 hobbs at the risk of trolling, I think the answer to that question is "0%" unless you use a really generous definition of the word "compatible"
14:53 stevan hobbs: I do, cause syntax is only skin deep
14:54 masak syntax is at the same time irrelevant and compellingly important :)
14:54 isBEKaml beauty is only skin deep, so is syntax. :)
14:55 masak on behalf of all pretty skeletons, I protest! :P
14:55 isBEKaml you can all protest in your cupboards
15:09 jjl stevan: sorry if i'm annoying you on the github trail
15:14 moe [moe] prakashk created prakashk/list-op (+13 new commits): http://git.io/Z1eDUA
15:14 moe moe/prakashk/list-op 6b9803c Prakash Kailasa: barehashKey should be non-empty
15:14 moe moe/prakashk/list-op 9eb2a60 Prakash Kailasa: sync comments with code changes
15:14 moe moe/prakashk/list-op 30901a1 Prakash Kailasa: rule for list operators (leftward); now we can 'say "hello"', in addition to "hello".say.
15:18 moe [moe] prakashk opened pull request #78: ArrayClass methods, array/hash elements as lvalues etc (master...prakashk/list-op)  http://git.io/3d17TA
15:31 jnap joined #moe
15:35 moe [moe] prakashk pushed 1 new commit to prakashk/list-op: http://git.io/mtI_1A
15:35 moe moe/prakashk/list-op 8445b11 Prakash Kailasa: added tests for array/hash elements as lvalues; changed tests to use ShouldMatchers
15:37 moe [moe] stevan pushed 1 new commit to master: http://git.io/Nnnpmw
15:37 moe moe/master 041c1a3 Stevan Little: Merge pull request #78 from MoeOrganization/prakashk/list-op...
15:39 moe [moe] prakashk deleted prakashk/list-util at a504fc7: http://git.io/BCu7XA
15:41 moe [moe] prakashk deleted prakashk/context-attempt at 89f2f74: http://git.io/zzrHEA
15:41 moe [moe] prakashk deleted prakashk/explicit-coercion at 81e4a15: http://git.io/hib0MA
15:45 isBEKaml joined #moe
15:55 stevan jjl: not at all, sorry if it is getting lost in the high level of noise in that ticket :)
16:07 jhannah hmm... what would I do with all these sigil tattoos?
16:07 jhannah (luckily they're only skin deep)
16:28 jasonmay jhannah++
16:38 sahadev ArrayClass needs a flatten method, I think. Otherwise, I see no way to implement flattening of nested lists.
17:01 stevan sahadev: yes, that is true
17:01 stevan sahadev: also got a regression in the parser with your branch
17:02 prammer_ joined #moe
17:04 stevan sahadev: trying to isolate a test case right now
17:08 sahadev oops. is it about the array/hash lvalue or the listOp rule?
17:08 sahadev all tests did pass in my repo, though.
17:10 stevan looks like I just needed to swap listOpLeftward and anonCodeRule
17:10 stevan what was happening is that this
17:10 stevan $foo = $foo + 1;
17:10 stevan was getting parsed as
17:11 stevan the second $foo being a subroutine call
17:11 stevan this is because of the anonCodeRule
17:12 stevan actually, no that didn't solve it
17:12 stevan hmm
17:14 sahadev hmm, it is working in my local repo.
17:14 sahadev moe> my $foo = 10; $foo = $foo + 1; {"CompilationUnitNode" : {"ScopeNode" : {"StatementsNode" : [{"VariableDeclarationNode" : {"name" : "$foo", "expression" : {"IntLiteralNode" : 10}}}, {"VariableAssignmentNode" : {"name" : "$foo", "expression" : {"BinaryOpNode" : {"lhs" : {"VariableAccessNode" : "$foo"}, "operator" : "+", "rhs" : {"IntLiteralNode" : 1}}}}}]}
17:14 sahadev 11
17:14 stevan sahadev: run bin/moe
17:15 stevan it seems to only be breaking there
17:15 jhannah joined #moe
17:15 stevan I will keep looking
17:16 stevan doing a full clean first
17:16 sahadev bin/moe works too.
17:16 sahadev I am up-to-date with current master branch
17:17 stevan hmm
17:17 stevan ok, let me pull again
17:17 stevan sahadev: no, I am up to date too
17:17 stevan odd
17:17 stevan I will figure it out
17:20 stevan oh, hmm
17:20 stevan I think I see it
17:21 jnap joined #moe
17:22 sahadev created a brand new clone off master, and tried it. works there too.
17:23 stevan yeah I think it is something to do with what is getting bundled in the assembly on my end
17:23 stevan I actually think it is a problem with eh ident from JavaTokenParsers
17:23 stevan cause $ is a legal character in that
17:24 stevan yup, that was it
17:25 stevan sahadev: I am going to not use ident anymore and just use [a-zA-Z_][a-zA-Z0-9_]* instead
17:25 stevan which is more correct anywya
17:25 jhannah joined #moe
17:26 sahadev how come I am not seeing the problem on my end? is it possible that I have a different version of JavaTokenParsers than yours?
17:26 stevan yes, that is what I am thinking it is
17:26 isBEKaml JavaTokenParsers is in scala-library
17:26 stevan isBEKaml: but it seems to use the Java Regexp stuff
17:27 stevan """\p{javaJavaIdentifierStart}\​p{javaJavaIdentifierPart}*""".r
17:27 stevan so might be a java version issue between me and sahadev
17:27 stevan either way, the new regexp is more correct and gives us more control to later add unicode stuff if we want
17:28 isBEKaml stevan: I don't see how that could introduce differences, since this has been the same from jvm1.x days.
17:28 moe [moe] stevan pushed 1 new commit to master: http://git.io/4OCO5Q
17:28 moe moe/master 62f1c22 Stevan Little: fixing the ident that was causing me issues but no one else :)
17:29 isBEKaml I don't recall any recent changes to java literal syntax either at java source or bytecode level.
17:29 stevan isBEKaml: I have no idea, either
17:31 moe [moe] stevan pushed 2 new commits to master: http://git.io/ZpgZ4A
17:31 moe moe/master 202725d Stevan Little: might as well keep the test, it passes
17:31 moe moe/master 012ae35 Stevan Little: take advantage of our new fancy listop syntax
17:31 stevan sahadev++ # say 1, 2, 3
17:31 sahadev got tired of saying @foo.say etc :)
17:32 stevan :)
17:32 stevan well I added it as a function in CorePackage
17:32 stevan so you could have said say($foo)
17:32 sahadev i like @obj.method for some things, but with say, the functional style feels more natural
17:32 stevan yes, I agree
17:33 sahadev yeah, i saw that. in your Test::More package.
17:33 stevan yup
17:33 sahadev by then, I was already looking into this change, so I continued.
17:34 stevan no, it is a good thing
17:34 stevan less parans the better, this isn't LISP :)
17:34 sahadev :)
17:41 perigrin it's a scheme.
17:41 perigrin sorry twitter has me being trollish
17:48 tobyink joined #moe
18:03 stevan perigrin: Scheme 6 though, not Scheme 5
18:05 isBEKaml stevan: R7RS is in public draft. :)
18:12 gizmomathboy joined #moe
18:23 perigrin stevan: Scheme 6 light
18:32 prammer joined #moe
18:54 tobyink joined #moe
19:02 umask001 joined #moe
19:10 jnap joined #moe
19:29 gizmomathboy joined #moe
19:55 jhannah but maintained
19:57 moe [moe] prakashk created prakashk/string-literal-fix (+3 new commits): http://git.io/IeF2_w
19:57 moe moe/prakashk/string-literal-fix 4c9cefb Prakash Kailasa: fix string-literal parsing rule to preserve leading whitespace
19:57 moe moe/prakashk/string-literal-fix a22327d Prakash Kailasa: .pad() workaround no longer needed.
19:57 moe moe/prakashk/string-literal-fix 24ffac6 Prakash Kailasa: chr(35) issue seems to have been fixed as well.
19:59 moe [moe] prakashk opened pull request #79: String Literal parsing fix (master...prakashk/string-literal-fix)  http://git.io/JYr5NQ
20:03 moe [moe] prakashk deleted prakashk/string-literal-fix at 24ffac6: http://git.io/DCt6Fg
20:05 moe [moe] prakashk opened pull request #80: String Literal parsing fix (master...prakashk/string-literal-fix)  http://git.io/lDkfuQ
20:10 moe [moe] stevan pushed 1 new commit to master: http://git.io/PWNuHw
20:10 moe moe/master 425446f Stevan Little: Merge pull request #80 from MoeOrganization/prakashk/string-literal-fix...
20:10 stevan sahadev++
20:20 moe [moe] prakashk deleted prakashk/string-literal-fix at dfc5f2d: http://git.io/Bc9ssA
20:20 moe [moe] prakashk deleted prakashk/list-op at 8445b11: http://git.io/mP7DHg
20:26 jnap joined #moe
20:44 jnap1 joined #moe
21:52 umask001 joined #moe
22:01 umask001 joined #moe
22:14 jnap joined #moe
22:31 umask001 joined #moe

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