Hello,

On Thu 20 Feb 2025 at 11:18am +01, Simon Josefsson wrote:

> Do consumers of this field have to handle both SHA1 and SHA256 git
> object ids?

I don't believe that this has to be specified in Policy.  There isn't
any ambiguity -- SHA1 and SHA256 checksums cannot be mistaken for each
other.  Whether they have to handle both just depends on what is in the
archive; at some point that will be all SHA1, then mixed, then all
SHA256, presumably.

> OpenPGP v5 or v6 fingerprint?

Here my knowledge is lacking.  Can they be unambiguously distinguished?

> How are they supposed to behave if they cannot understand parts of the
> information in the header?

Surely that would depend on what the tool was trying to achieve and
couldn't be specified in general?

-- 
Sean Whitton

Attachment: signature.asc
Description: PGP signature

Reply via email to