Open aartaka opened 2 years ago
Perhaps (optionally?) include bookmarks for a source, too?
History, bookmarks, and all current opened websites/buffers all seem to be good source.
I think this is a good idea. Perhaps we can also store a cache of all DNS records? It is not too large is it?
Apparently there are about 360 million domains. If each domain is about 10 bytes, then that is about 3.5 gigabytes. I don't think it is /too/ unreasonable.
To check whether user-inputted URL is an actual website, we check it doing DNS resolving. This is slow and fast typing can invalidate it. We need a local, fast-to-access, and user activity-based source of information on which URLs are reliable. History fits it perfectly.
See https://github.com/atlas-engineer/nyxt/issues/1886#issuecomment-1015806875 fir initial discussion of it.