Those permission entries might have been this way ever since the Crixa[1] mishap, then... as I believe that the Moose distribution always excluded them until 2.2014.
[1] sorry Chris, I know you'll never live this down... :D but for Neil: several years ago there was a separate distribution released by Chris that accidentally bundled all of Moose inside it, causing all Moose modules to be reindexed to the Crixa distribution.. it was easily fixed, but it shows what happens when errant releases happen by people who have permissions to the world :D On Tue, Apr 6, 2021 at 10:06 AM Neil Bowers <neil.bow...@cogendo.com> wrote: > I recall that there was a metadata change between releases 2.2013 and > 2.2014 (last year) to allow some internal modules to be indexed. I did > those releases.. how did PERIGRIN end up with first-come? that seems like a > pause bug. > > > Ah, Checking META.json for 2.2013, these four packages were all excluded > from indexing by the no_index. > > So I don’t think this was down to a PAUSE bug. My guess is that the > packages already had indexing permissions (either from earlier releases, > not necessarily of Moose), but were being ignored by the PAUSE indexer > because of the no_index. > > When no_index was changed, PAUSE would have considered them for indexing. > And because they already had indexing permissions, PAUSE wouldn’t try and > enforce the recent rules on consistent permissions — that applies to > packages that don’t have any indexing permissions. > > Neil > >
