It is scary, but it already seems more stable and backward-compatible than Deno. With some testing and further stabilization, I have a feeling bun might be a much more feasible and beneficial move.
It's not backwards compatible, it's just Node compatible. Demo is not, and it's stated as much clearly. Thus the suggestion that bun is more backward-compatible than Deno is incorrect and speaks to a fundamental misunderstanding of the two projects.
Sure, a poor choice of words. Node compatibility is what matters to me (and likely many others) practically as a node developer. Deno's benefits don't outweigh the costs of migrating to it and adopting it more broadly, in my opinion.
tried deno briefly the other day, was a real pain. just wanted to hash some files. they had an std hash lib but dropped it in favor of web crypto which bazzarely doesn't have a hash.update method for streaming?? not very good. prefer node's impl
