[Unison-hackers] OCaml version mismatch breaks Unison 2.48.4

Benjamin Pierce bcpierce at cis.upenn.edu
Sun Sep 13 12:47:05 EDT 2020


> Anyway, whatever the hash function, I think it should be more
> standardized than just "OCaml's Digest module". If we insist on no new
> dependencies (do we?), that hash function would have to be in Unison
> sources.

Depends what you mean by “insist”. :-)

IMO, restricting Unison to depend only on the standard OCaml installation has served us well over the years by making it pretty easy for people who needed to compile from scratch to do so.  But if the OCaml world has reached the point where there are really easy ways of satisfying dependencies on other packages (across all the platforms Unison runs on!), then I’d be happy to know about them.  I’m a bit out of date on best practices.

In any case, though, if we’re just talking about adding one file to the Unison distribution (and it has an appropriate license so that we can do so), then why wouldn’t we take that route?

> And my marshalling implementation changes everything in an
> incompatible way, so it's in my opinion a good time to change the hash
> function. And as a person doing cryptography in my day job, I cannot
> imagine using MD-5 or SHA-1 today. CRC-32 could fit some purpose, but I
> don't think it's what we want here.

If that’s your day job, I am very happy to follow your lead here!



More information about the Unison-hackers mailing list