Discord Name: Lionirdeadman#7572
GitHub Repository (Must be V3): https://github.com/Jintaku/Jintaku-Cogs-V3
Description: V3 Cogs made by the Jintaku community
Hey!, A couple issues here to be resolved. (and I don’t know that this is an exhaustive list, though it appears to be based on standards set unless something was missed)
-
Repo level info.json’s author key should be a list of strings (see https://red-discordbot.readthedocs.io/en/v3-develop/framework_downloader.html#info-json)
-
xkcd should be an all caps class name. I understand this isn’t ideal, but it breaks the help formatter’s notion of help for cogs to not be. (I literally can’t invoke the command help, as it will always show the cog’s)
-
print statement needs to go in xkcd
-
probably shouldn’t hardcode a latest xckcd, and instead check what the latest is periodically (given xkcd’s updates on a mostly regular schedule, you could even store in config and see if you should check if you’d like to minimize requests made) (see https://xkcd.com/json.html for details if needed)
-
booru does mention that it has nsfw content, but does not respect discord channel settings for this. While your settings for this may be fine in supplement to these, we really can’t accept cogs placing known nsfw content into channels not marked as such.
-
booru’s print statements need to go. use logging.debug to a custom logger if you need them for testing.
-
missing doc strings in booru’s commands also breaks help formatter, and same cog name issue. (
Booruwould be the preferred class name) -
User facing tracebacks in booru: can happen on filter remove (attempting to remove a not set filter), may have missed elsewhere.
I’d be willing to approve should these be fixed, and the xkcd checking latest one is merely a suggestion.
On the strength of the AniList search cog alone though, your repo could easily stand up, so if you aren’t ready to handle everything in booru, you could move that into another branch than the one being submitted here until you are.
Alright, I’ll take note of the things for XKCD, it seems simple enough.
As for Booru though, I think I need to rework the whole filtering engine and I don’t want to deal with having a different branch (in part because I’m lazy and in another because I’m bad at git).
So just wait until I fix it, I’ll probably contact you on Discord or add another reply here.
After some talk behind the scenes, 1 more additional option available is to mark booru as hidden in the info.json for it and to include a clear statement that it is not ready for public use in the install message.
For clarity, this was to keep with the following ideals we had set out to achieve:
- Good cogs should be available to the community
- The community should be protected from issues where possible
- The barrier to entry for a repo being approved should ideally just be contributing something worthwhile to the community.
As I already stated your anime search cog could easily stand alone, the only point needing addressing is the protecting the community one, which the info.json data can be used to accomplish.
I’ll also be ensuring that this level of clarity about flexibility is made available where otherwise stated.
sorry to tack on one more thing, noticed this after the fact, but your keynames in info.json are expected as all lower case (see previously linked documentation if needed)
I believe I’ve done all the changes required then. Thank you. 
(Except Booru which I’ve simply hidden as proposed)
One last thing. You now have a trailing comma which doesn’t belong in the xkcd cog’s info.json file (At the end of the tags section), Marking as approved pending that small change.
My bad, done now.
