|
07:41
lizmat left
08:34
lizmat joined
08:54
lizmat_ joined
08:57
lizmat left
09:05
lizmat_ left
09:06
lizmat joined
|
|||
| lizmat | hmmm.. I've been removing obsolete module references from the p6c module list, but I don't see the number of p6c modules on raku.land going down | 11:53 | |
| are the statistic calculated at a different time ? | 11:54 | ||
| jjatria JRaspass ? | |||
| jjatria | lizmat: you're looking at raku.land/stats ? | 11:56 | |
| lizmat | yes | ||
| jjatria | IIRC, that's calculated at ingestion time with the data from the database. The distributions you removed from p6c, are they also gone from the RL index? | 11:58 | |
| lizmat | RL index? | 12:02 | |
| isn't ecosystem/META.list authoritative ? | 12:03 | ||
| jjatria | RL aggregates data from those sources into its own index, so yeah, it should still be authoritative. RL tries to not be a primary source of any data, if it can avoid it | 12:04 | |
| lizmat | looks like the modules from p6c are indeed gone from the RL index | 12:06 | |
| aahh.. and stats is updated now | |||
| looking at the log now, not sure why it complains about e.g. not finding github/raku-community-modules/ClassX-StrictConstructor ? whereas github.com/raku-community-modules/...onstructor is defnitely a thing ? | 12:08 | ||
| anyways, looks like the index didn't complete (as it didn't show the final result table) | 12:10 | ||
| ok, I'll continue cleaning up p6c's meta list then :) | |||
| thanks for looking into it! | |||
| ok, this one got me puzzled: | 13:37 | ||
| the error log mentions: github: Could not resolve to a Repository with the name 'lizmat/Object' | |||
| but I have *no* idea where that could be coming from: jjatria JRaspass any ideas? | 13:38 | ||
| jjatria | Let's look | 13:58 | |
| That's a GraphQL error. We construct the query (owner / repo) from the URL we ingested: gitlab.com/raku-land/raku-land/-/b...heads#L251 | 14:08 | ||
| Not sure what distribution in particular is the one that points to `lizmat/Object`, though | |||
| lizmat | so it could be a module that lists a dependency incorrectly ? | 14:09 | |
| jjatria | It appears it comes from version 0.0.10 of Object::Trampoline, which lists the URL as "github.com/lizmat/Object/Trampoline" | 15:07 | |
| In any case, it's more of a warning than an error. That version was indexed correctly: raku.land/zef:lizmat/Object::Tramp...e?v=0.0.10 | 15:09 | ||
| lizmat | ah, ok, so in that sense, it's unfixable (as it has apparently been fixed in more recent versions) | 15:54 | |
| JRaspass | Regarding immutable errata it might make sense to have a quirks file for overwriting legacy releases | 16:14 | |
| lizmat | looks like I made changes to the p6c ecosystem in the wrong branch today: master instead of "main" | 16:59 | |
| the fact that I *did* see changes in raku.land, makes me wonder whether raku.land is maybe still looking at the master branch as well ? | |||
| yup: gitlab.com/raku-land/raku-land/-/b...=heads#L17 jjatria JRaspass | 17:05 | ||
| gitlab.com/raku-land/raku-land/-/m...equests/11 | 17:09 | ||
| JRaspass: thanks, p6c now down to 371, 15.9% :-) | 21:43 | ||