Improve automatic issue raising?

[tfardet]

The automatic issue on feed fails seems to be generating a lot of issues which, as far as I can tell, are just temporary issues that do not require any action (please correct me if I'm wrong).

Would it be possible to improve this part to reduce spaminess?
Potential improvements I have in mind could be:

• caching the fails and raising the issue only after X fails
• mentioning only the maintainer of the failing feed

[jbruechert]

I'd recommend splitting up the json file in case you are mentioned to often for feeds you didn't add.

per-feed maintainer info seems a bit too verbose to me.

The first point definitely makes sense, but is more work than it might seem. We currently don't preserve anything about the failures between runs apart from the GitHub issues.

[tfardet]

OK, thanks for the information.
Is there a standard for splitting the json in terms of naming conventions?
I.e. if it's {country}-{something}, is there a rule for what "something" should be ?

[jbruechert]

If your feed applies for some specific district, you can append the appriviation of its name.
We aren' t currently following any standard there.

[Pi-Cla]

No one has to follow any sort of standard when naming a file but so far everyone has followed ISO 3166-2 codes which include the subdivisions when you click on the name of the code.

For example, for France these would be the subdivision codes: https://en.wikipedia.org/wiki/ISO_3166-2:FR