Add Zenodo to the registries check?

[sdruskat]

If Zenodo were added to the list of registries to check against, this would cover an awful lot of "generic" software releases, e.g., those that don't go into any of the language/packaging registries, e.g., if it's distributed as binary/zip.

As far as I can see, there are two ways to achieve this:

1. Check for a Zenodo badge in the README. Cheap but may be imprecise (could additionally check if the record exists and has an isSupplementTo record for this repo).
2. Do an inverse search for records on Zenodo that contains the repo in the JSON. Expensive but precise:

"metadata": {
                "related_identifiers": [
                        {
                            "identifier": "https://github.com/this/repo/tree/v1.1",
                            "relation": "isSupplementTo",
                            "scheme": "url"
                        }

[jspaaks]

Just to clarify, there is a test for presence of a Zenodo badge, but it is part of the citation checks, not the registry checks

https://github.com/fair-software/howfairis/blob/d82eb0f92620f066f28bb5083eb3097717187606/howfairis/mixins/CitationMixin.py#L62-L65

[mattkretz]

Is there any reason why Zenodo has not been added to the list of registries other than "nobody submitted a PR"? It's listed on https://github.com/NLeSC/awesome-research-software-registries, so why shouldn't it count?

[jspaaks]

The reasoning at the time was that Zenodo as a community registry does not serve the purpose of discovery very well, i.e. items in Zenodo are not ranked well by search engines. FWIW I still agree with this.

[dennisklein]

What is the decision making process here? Is it possible to re-open it to give zenodo inclusion another shot?

[ChristianTackeGSI]

The OSSR (also now member of SciCodes) currently uses zenodo's communities as its authoritative source of truth.

So using zenodo or "being part of some list of communities" would be highly welcomed by projects using the OSSR.