Agree!
In my case, yes 💯
Indeed!
Thanks for sharing that! I agree that using an id is more modular than hard-coded enums
Thanks for mentioning that!
I do. Kinda happy paths. I agree that a better coverage would catch it
Haha, I don’t. Just moved a bit fast here, but I used to have a 360° view of my products
The DX of the client looks good to me. Just sharing an experience here. If I wanted an error to be raised, I’d use this throwOnError: [https://github.com/supabase/supabase-js/blob/b835e3c870f3767980776cc25fd16df72fa416be/packages/core/storage-js/src/lib/common/BaseApiClient.ts#L87-L103](https://github.com/supabase/supabase-js/blob/b835e3c870f3767980776cc25fd16df72fa416be/packages/core/storage-js/src/lib/common/BaseApiClient.ts#L87-L103)
Not reallY. The DX looks good to me. Just sharing the experience here
In my DB, I had a constraint on a column called "subscription\_type" that only allowed specific values, 'starter' and 'pro'. I end up by switching to 'premium' instead of 'pro'. The issue was that the database constraint was still expecting 'pro', so inserts with 'premium' was silently failing I fixed it by writing a migration to update the constraint and then ran the migration on the database. Then I generated the Supabase types and initialized my client with createClient<Database>, so the linter can catch this kind of mismatch
Yeah, that was indeed totally my fault. The way Supabase handles this is probably not very intuitive, but it provides a good developer experience and is fine with me. Else, i'd just wrap with an high order function
Yes, it is totally reliable. When an issue is found, the user is notified so they can fix it. After applying the fix, they can rescan. I do not store any data. Plus, since I only need the site URL, if my intentions were bad, I could simply scrape places like Reddit instead.
Thanks a lot for testing! I try to reduce false positives by detecting the nature of the key (e.g. anon vs sensitive keys). In cases where it’s unclear, I end up by leaving the judgment to the user