Should we expose a sync version of `fetch_ohttp_keys`?

[benalleng]

Following some review in the liana implementation it was brought to my attention it might be useful to make a sync version of fetch_ohttp_keys() publicly available so we don't have to manually re-implement it where async would be a burden.

[DanGould]

wait didn't @spacebear21 just say that this is a very unlikely scenario? Is your link to Liana a counterpoint to his claim which I can't find right now

[spacebear21]

My claim was regarding implementations that use FFI language bindings fwiw

[DanGould]

I feel like Liana is the type of implementation where they'd want to roll their own fetch_ohttp_keys unless they coincidentally also used the same HTTP client. Definitely a place bitreq makes more sense than reqwest, but if we're supporting bitreq then it makes sense for liana to use io and perhaps for us to expose that functionality for them.

[xstoicunicornx]

FWIW I was the one who made the Liana review and I had not realized at the time that fetch_ohttp_keys was async. So I think probably appropriate to close this issue until there is at the very least one other implementation that would use this.