Groups keyboard shortcuts have been updated
Dismiss
See shortcuts

Moving forward on the "blocker" issues

10 views
Skip to first unread message

Roberto Di Cosmo

unread,
Apr 4, 2023, 11:03:12 AM4/4/23
to swhid-...@googlegroups.com
Dear all,
     the current open issues for the SWHID specification are here: https://github.com/swhid/specification/issues

Among them, the following ones are marked "blocker" and have an associated pull request (PR):
Specify precisely the way a core identifier is computed for a revision node blockerSpecification incomplete or incoherent
#11 opened on Feb 21 by rdicosmo
Specify precisely the way a core identifier is computed for a release node blockerSpecification incomplete or incoherent
#10 opened on Feb 21 by rdicosmo
Specify precisely the way a core identifier is computed for a snapshot node blockerSpecification incomplete or incoherent
#9 opened on Feb 21 by rdicosmo
Specify precisely the way the core identifier is computed for a directory node blockerSpecification incomplete or incoherent
#8 opened on Feb 21 by rdicosmo

The associated PRs, listed below for your convenience, are meant to provide the required precise specification for the computation of the SWHID of directories, revision, release and snapshot nodes:

Please look at the proposed language, and comment directly on the PRs: if no significant unsolved issue arises by April 27th, they will be merged.

The last issue marked 'blocker' that needs to be resolved before we produce a candidate v1.0 is the following one, and it needs discussion

Hashing algorithm blockerSpecification incomplete or incoherent
#16 opened on Mar 2 by zvr

Cheers

--
Roberto (on behalf of the SWHID specification core team)

P.S.: the poll to change the SWHID acronym expansion from SoftWare Heritage IDentifier to SoftWare Hash IDentifier (see issue #19 on GitHub) is still open at https://evento.renater.fr/survey/community-vote-on-swhid-acronym-jfu5toc9
Reply all
Reply to author
Forward
0 new messages