[OPSAWG][VELOCE] Discussion (1/3): Where should VELOCE repositories live?

[Mahesh Jethanandani <[e-mail]>]

Series note: In the spirit of kicking off discussion on the mailing list before presenting it in a WG meeting, this is the first of three emails soliciting WG feedback on specific open issues in the VELOCE draft, which was recently adopted (but the adopted WG document will be posted when the gates open up). Two more will follow — on how an RFC references a YANG module in a repository, and on proposed issue sequencing. All three topics will also be presented and discussed at the upcoming IETF meeting in Vienna.

Dear OPSAWG participants,

The VELOCE experiment requires each participating working group to maintain a git repository for its YANG modules. Issue #8 [1] asks a question that has remained open since the beginning of the experiment: which platform should host these repositories?

THE OPTIONS

Three options have emerged in the discussion:

GitHub — the current de facto platform for IETF work, including this draft. Widely familiar, strong tooling, and easy for contributors. However, it is a commercial third-party platform, and some WG members have raised concerns about long-term reliance on it.

IETF-operated GitLab — the IETF holds a GitLab licence, which would give the community more control over the infrastructure. Less familiar to many contributors, and it is not yet clear what operational support IETF would provide.

IANA registry — IANA already maintains a registry of YANG modules. It could serve as the canonical reference point for published module versions, even if the day-to-day development repository lives elsewhere (e.g., on GitHub or GitLab). PR#49 explores this direction.
A FOURTH POSSIBILITY

These options are not mutually exclusive. The VELOCE process could be platform-agnostic — specifying what a repository must provide (public access, stable tagged versions, archival guarantees) without mandating a specific host. This would give each WG flexibility, but could make tooling and process consistency harder to achieve.

QUESTIONS FOR THE WG

Does the WG have a preference among these options?

Should VELOCE specify a single platform, or define a set of requirements that any platform must meet?

What archival guarantees should be required — specifically, what happens to a repo if the hosting platform shuts down or changes its policies?

Please share your views on the list. This topic will also be on the agenda at the upcoming IETF meeting in Vienna, and your mailing-list input will directly inform that session.

[1] 
http://github.com/mjethanandani/veloce
mjethanandani/veloce: IETF draft to talk about the VELOCE effort
github.com

Cheers

Mahesh Jethanandani
[e-mail]