Results from days 1 to 10, from overall 492 days
Result pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
| 09:06 | sorear | moritz: isn't there a way to get at the default through the MOP? | |
| 09:07 | masak | ah, moritz++ had the same idea but with an array. | |
| 09:07 | moritz | sorear: through introspection, yes | |
| 09:07 | moritz | but well, if the caller goes to *those* lengths to fool us, I guess he deserves to get the result of us being fooled :-) | |
| 09:08 | * masak | .oO( you can always hack the system by bombarding the CPU with alpha particles ) | |
| 13:38 | ajr_ | ) | |
| 13:38 | PerlJam | ajr_: in a way, that's okay because the context around the problems shifts. | |
| 13:38 | arnsholt | Also, when you look at the kind of introspection and tweaking that was possible in Smalltalk and Lisp-machine systems you realize that modern OSes aren't all that awesome | |
| 13:39 | masak | arnsholt: if you're wiped out by Worse is Better, you'll either spend the rest of your life being bitter about it, or (like Eirich Gamma et al.) you rebuild the tools in Java instead of Smalltalk. now, Eclipse is a weak ghost of what Smalltalk was, but at least it has the original ideals and it's being used a lot. | |
| 13:39 | PacoAir joined #perl6 | ||
| 20:50 | bruges joined #perl6 | ||
| 20:50 | TimToady | but I think the original idea was that you'd get these types by instantiating a role, and that role just instantiates to Str,Any by default | |
| 20:51 | TimToady | doing it by attaching info to the constructor means you're always forced to do introspection to find out if two types are equal, and introspection is kinda smelly | |
| 20:52 | TimToady | with classes you can use named equivalence sometimes | |
| 20:54 | lizmat | so effectively you're saying they're *all* typed hashes |
| 12:44 | jnthn | Under the hood we don't really look up most lexicals by name. | |
| 12:44 | lizmat | FROGGS: well, that's exactly what I'm trying to figure out | |
| 12:45 | lizmat | jnthn: it's more for introspection, I guess | |
| 12:47 | huf_ joined #perl6 | ||
| 12:47 | * FROGGS | .oO( my $dog = Dog:auth({/cpan/}).new( ... ) ) |
| 20:44 | pmichaud | use 1 if @array and 1 if +@array | |
| 20:44 | pmichaud | { @array } just tests the cost of a lexical lookup | |
| 20:44 | gtodd | lizmat: so re . perl5 non-equivalent of perl6's .^methods ... in the perl5 debugger if one does "use Mojo;" and then x \%{Mojo} you get a hash of a bunch of things and m \%{Mojo) just gives you object methods like ... UNIVERSAL: isa ... etc. which is nowhere near perl6 introspection | |
| 20:45 | gtodd | and never can be ... | |
| 20:45 | lizmat | pmichaud: ok |
| 16:05 | TimToady | or the iterator extracts unique .values | |
| 16:06 | TimToady | but probably forward is hash Wed => 3, and backwards just an array 3 => Wed | |
| 16:06 | PerlJam | masak: I guess because I don't see another way for introspection yet. | |
| 16:06 | TimToady | and the iterator runs over the array | |
| 16:07 | PerlJam | (and if I put in 5 things, I expect to get 5 things our somehow) |
| 16:02 | TimToady | the copy should be empty by the end of BUILDALL | |
| 16:02 | jnthn | TimToady: The tricky bit I don't quite see yet is how calling a user-defined BUILD will mark things as used. | |
| 16:03 | jnthn | Short of introspection, which is usual a bad sign. :) | |
| 16:08 | TimToady | well, the most obvious place to do it is in the value fetcher; since named args are weird anyway, they could use a separate type from normal hashes | |
| 16:08 | TimToady | the hard problem is knowing where to store the resulting "I've been used" bit |
| 17:40 | TimToady | assuming it can't be known at compile time | |
| 17:40 | TimToady | which some methods might be knowable depending on which classes can close/finalize at CHECK time | |
| 17:42 | TimToady | and yes, it's an introspection, but hopefully it's the least smelly way to get something that dwims most of the time | |
| 17:43 | TimToady | in any case, it's an introspection that doesn't need to be made on every node, just on the initial hyper dispatch | |
| 17:43 | pmichaud | well, it's introspection at the level of the operator, which means we can likely hotpath some of them | |
| 17:43 | TimToady | two minds with but a single thought... :) | |
| 17:48 | arnsholt | masak: http://translate.google.no/#no/sv/postoppkrav | |
| 17:48 | moritz | wait, doesn't it need to introspect for each element again? | |
| 17:48 | arnsholt | (And anyone else Scandinavian-phone, I guess. moritz perhaps?) | |
| 17:55 | sorear | do these two words refer to specific species of fish? | |
| 17:55 | TimToady | (the .duckmap vs .deepmap distinction is really only useful when the operators are not know to be disjunct) | |
| 17:56 | TimToady | so the initial dispatch itself can in many cases serve as the "introspection" | |
| 17:56 | sorear | Aha | |
| 17:56 | TimToady | *the initial dispatch on the item | |
| 21:24 | jnthn | If one extra allocation per method call (until I elimiante it) does that, we're in trouble... | |
| 21:24 | pmichaud | exactly. | |
| 21:24 | pmichaud | anyway, I'm fine with always generating the %_ parameter. We probably want it for introspection anyway. | |
| 21:24 | jnthn | And maybe TimToady++ will backlog this and decide that *% is enough :) | |
| 21:24 | pmichaud | *% can work too. |
| 16:49 | rurban joined #perl6 | ||
| 16:49 | TimToady | I suspect the semantics we're actually looking for there is "Can this value be coerced to the enum?", which probably puts it into the bailisick of a Day() coercive declaration, not a smartmatch | |
| 16:51 | TimToady | on the / $x / and / @x / questions, is there anyone currently depending on the ambiguity to pass both strings and regex to the same matcher? I'm inclined to rip out the introspection and say / $x / is always / "$x" /, and you must say / <$x> / to get the other thing | |
| 16:52 | TimToady | likewise / @x / would always be literals, and / <@x> / would always be regex | |
| 16:52 | TimToady | and if you want to intermix, then you'll have to preprocess the strings in your arrays into regexen |
| 16:09 | lizmat | std: foo->bar | |
| 16:09 | camelia | std 316b40b: OUTPUT«===[0mSORRY!===[0m�Unsupported use of -> as postfix; in Perl 6 please use either . to call a method, or whitespace to delimit a pointy block at /tmp/ICAjMSc0PH line 1:�------> foo->�bar�Undeclared routine:� 'foo' used at line 1�Parse fa… | |
| 16:10 | gtodd | PerlJam: yes, and it seems to be getting better at it ... is it the introspection ? | |
| 16:10 | gtodd | lizmat: nice ... | |
| 16:10 | lizmat | experience frozen in grammar |
| 13:41 | tgt joined #perl6 | ||
| 13:45 | gtodd | /usr/local/bin/perl Configure.pl --gen-parrot=RELEASE_5_3_0 --gen-nqp --prefix=$HOME/perl6 seems to work | |
| 13:46 | arnsholt | Awesome module idea of the moment: (De)serialize arbitrary objects to/from a graph database using MOP introspection | |
| 13:51 | gtodd | sort of like that ZODB thingie in python? | |
| 13:53 | arnsholt | Yeah, that looks pretty close to what I really want instead of ORM systems |
| 18:15 | FROGGS | I dont know these | |
| 18:15 | lizmat | I agree there is some database needed in each @*INC directory (similar to decided at QA Hackathon for Perl 5) | |
| 18:16 | lizmat | but I'm not sure we will want to consult that for loading a module, it should be for introspection only, I would think | |
| 18:16 | FROGGS | lizmat: but you need that for locating a module | |
| 18:17 | FROGGS | since you can install a module from different auths and version at a time, and load a specific one |
Result pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50