After using prisma on a few projects I would now rather abstain from using it. When there are type problems the error output is nigh unreadable. And as thos article indicates they seem to be more concerned about expanding their product line than fixing old issues (no multi insert statement). A bare bones way to generate types from SQL (readable types mind you) is a more fruitful path to explore imo
Your criticism re: old issues is spot on. However, fixing backlogged issues is easier said than done. Any OSS project with more than 350k monthly active devs won’t be able to fix everything unless they throw an army of engineers at it.
Another thing we're not very good at is sharing with our community the reasoning behind how we prioritize bug fixes. This opacity leads to confusion and we're going to fix it.
Re: the aforementioned fruitful path, I am assuming you're referring to our recently announced feature "TypedSQL," which incidentally was one of those most-requested features.
While the commercial products are a focus, we're still headstrong on delivering ongoing value to our OSS users. As evidence, we've religiously released a new version every 3 weeks for the past (almost) 4 years. A quick glance at our changelog (https://prisma.io/changelog) should reassure you that the OSS ORM product is, and will always be, top of mind.
I really appreciate the feedback though. Keep it coming!
