|
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 25 February 2017. |
|||
|
00:14
llfourn joined
02:38
llfourn 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 | ||
|
06:34
llfourn joined
07:33
llfourn joined
09:08
domidumont joined
11:55
domidumont joined
12:53
domidumont joined
13:55
SmokeMachine joined
15:02
lizmat joined
|
|||
| nine | ugexe: resolve-unknown-repos does not work yet as smoothly as it should. I hit issues in that area when trying to create rpm packages for Inline::Perl5. It's important that no none of the precompiled code and objects refer to the custom repo in any way. | 15:24 | |
|
16:27
lizmat_ joined
16:33
lizmat joined
|
|||
| ugexe | nine: so if I have custlib#/foo and have Foo::A and Foo::B where Foo::B has `use Foo::A`? I guess that explains that | 16:45 | |
| from what I got into it seemed like next-repo doesn't get set to the right thing somewhere | 16:50 | ||
| but that might just be a symptom | 16:51 | ||
| I assume it should technically get past that specific problem if I integrate it into rakudo (so it can short-id2cur etc) and re-compile? | 16:54 | ||
| I wonder if we can't expose that interface like RAKUDO_EXCEPTIONS_HANDLER | 16:56 | ||
| "If present, the print_exception routine will use a class of that name to process the exception for output. Rakudo currently ships with Exceptions::JSON (invoked by setting this variable to "JSON"), to override the default output." | 16:57 | ||
| Also CUR::Lib is to meant to proof of concept a Recommendation Manager as `method candidates` (mentioned in s22 as well, but not as a rec manager) github.com/ugexe/Perl6-CompUnit--R...ib.pm6#L74 | 17:29 | ||
| in CURI such a method would be like .resolve except that it considers the current repo only, and it returns all matches | 17:30 | ||
| (and thus .resolve and others would use this underneath) | 17:31 | ||
| this could allow the rec manager to be changed by the user for CURs, but thats not the real point. Rather it allows zef to use the same rec manager roles as the CURs to resolve | 17:34 | ||
|
17:57
lizmat joined
|
|||
| ugexe | which I think might become more important if superceded-by, emulates, etc are implemented | 18:50 | |
|
19:21
lizmat joined
20:36
llfourn joined
|
|||
| nine | ugexe: Yes. Ironically your struggles show me that mine with keeping CompUnit::Repository::Staging out of core were not in vain. I could have easily put it into the setting and be done with it, but it was a good test bed for custom repo implementations. | 21:20 | |
| Unfortunately while I got it working well enough for packaging rakudo, it's still too fragile. But I'll have another look once I got require working well again. | 21:21 | ||
|
21:37
llfourn joined
22:36
llfourn joined
23:37
llfourn joined
|
|||