No platform independent way to contact a project owner (licensor) - i.e. missing EMail

[hoijui]

I Just checked, and from our 34031 projects, only 32 of the licensor fields contain an "@", meaning... there is basically no way to contact almost any of the project licensors in a hosting-platform independent way.

originated from to #20

Q: How to solve this?

We can add a dedicated field for licensor: email, and make it mandatory, but then the question remains, whether we can actually provide that information. As most of our projects are scraped form non-git based project-hosting websites, we would have to be able to scrape the email from there, which likely ... or even "hopefully" is not possible, as otherwise everyone would get spammed.

So we would have to check, if it is possible to get the project owner email on platforms:

• github
• gitlab.com
• Thingiverse
• Wikifactory
• Appropedia.org

If not, we might need to provide a hosting-platform dependent way of contacting a user, which would be very ugly, as then each software wanting to contact a user, would have to implement support for all these means of contacting.

[hoijui]

both Jaromil with dyne, and me came up with the solution independenty:
as is done with openhub.net (previosly OHLOH),
projects could initially be claimed without any need for verification,
and if a dispute over the ownership of a project would arise,
we (in this case dyne) would ask the project owners to publish an email through their OKH data
(as of now, this would be as part of the licensor field),
and would only then start to perform verification for that specific project with that data.

That requires though, that it is somehow possible to publish an email as part of the licensor field,
not just for projects on git repos, where the OKH data comes from an okh.toml file,
but also for projects scraped from other platforms, like Wikifactory or Thingiverse,
where the OKH data is (currently) solely scraped from the API/the meta-data the platform makes public about projects,
which (for obvious and good reasons),
does in most cases not contain the project authors email address.
I opened a new issue for that,
which proposes to also look for an okh.toml file on those platforms:
LOSH-krawler#10

[moedn]

not quite sure if this is an issue that should be solved via OKH metadata, see my latest comment in #7

[hoijui]

Ahh... that is where "multiple places with the same data" came from, in the other issue, I guess...
indeed, that is an issue, and a very big one. To word it out:
If we have a platform A that uses OKH data, and they want to keep 1+ emails per project for contacting, and they do it in the OpenHub way explained above, then Platform B would have to do the same thing again, separately, and each project owner would have to go through the process at each of the platforms separately.
-> data duplication, data getting outdated, data inconsistencies, ...

[hoijui]

Given that it already is possible to provide an email (as part of the string in licensor),
and the issue just explained above,
I would rather not solve this directly, with a separate email field for licensors,
but indirectly by trying to figure out a way for each hosting platform to supply a manually created TOML manifest-file to the krawler. This is trivial in the case of e.g. GitHub, but harder in the case of non-primarily-file-based hosters like Appropedia or Wikifactory & co..