|
Fire is step THREE! | github.com/perl6/toolchain-bikeshed | Channel logs: irclog.perlgeek.de/perl6-toolchain/today | useful prior art: metacpan.org/pod/CPAN::Meta::Spec Set by moderator on 10 March 2016. |
|||
|
02:01
FROGGS_ joined
02:48
ilbot3 joined
|
|||
| moderator | Fire is step THREE! | github.com/perl6/toolchain-bikeshed | Channel logs: irclog.perlgeek.de/perl6-toolchain/today | useful prior art: metacpan.org/pod/CPAN::Meta::Spec | ||
|
04:18
MadcapJake joined
06:45
JimmyZ joined
07:16
sivoais joined,
sjn joined
07:18
domidumont joined
07:45
FROGGS joined
07:57
domidumont joined
08:23
FROGGS joined
|
|||
| nine | Ok, my plan for serializing DependencySpecification objects didn't work out. Using EVAL is an absolute no-go during precompilation as it leads to broken precomp files. | 08:34 | |
| So we will have to define the known serializable subset of matchers beforehand, i.e. strings, substring matching, version ranges, maybe boolean operations, ... | 08:36 | ||
|
14:11
MadcapJake` joined
15:58
perlpilot joined
|
|||
| ugexe | what was your original plan? | 16:17 | |
|
16:23
Kassandry joined
|
|||
| ugexe | `method install($dist,` needs to be `method install($dist is copy,`. For instance: `@CURs.map: {.install($dist)}` will fail on all iterations past the first due to `.install` directly modifying `$dist.files` and `$dist.provides`. This changes the structure to one `.install` cannot understand (for reading) | 16:32 | |
| i also wonder if `.install` should return the Distribution object it ends up saving instead of just True (yes the Distribution object would be almost identital to the one passed in to .install($dist), but at the very least the META6 is different) | 16:35 | ||
| i think ideally the META6 stuff should be in its own encapsulation that can be arbitrarily attached to a distribution (like META6::S22). so in my last statement it would just `does META6::Installable` to $dist before returning it (no need to create a copy or new $dist). Then all the `provides` etc interfaces use the META6 encapsulation instead of direct member access | 16:43 | ||
| nine | ugexe: the original plan was to just write $spec.perl to the .deps file. This plan may actually work out after all if I'm able to fix the BEGIN time EVAL precomp issue in rakudo | 16:51 | |
| ugexe: I sometimes wonder why you just go forward and implement those suggestions. I mean you obviously think a lot about this stuff and know what to do. | 17:34 | ||
|
18:21
domidumont joined
18:42
hankache joined
18:51
cognominal joined
19:43
lizmat joined
20:40
kmel joined
|
|||