When syncing initially fails after a new release, we'd like to be able to re-issue a sync (maybe also restricting a specific sync for just certain pipelines).
Need to think about how / where to trigger this and who should be allowed to do it.
Command line? Website? By calling the @nf-core-bot in an issue? (like conda-forge does).
My suspicion is that the easiest way to do it would be to build it in to nf-core tools somehow. Probably just by a script in the bin folder that you supply with the login credentials for the bot account. It's not super user friendly that way, but it would be secure and easy, and I guess the idea is that we shouldn't have to do this manually very often (or ever..)
Phil
Technically it already works :) You can now run the sync script via command-line and pass a single pipeline name. If you run the script without any arguments, all pipelines will be synced. #307
Two environment vars need to be set:
TRAVIS_TAG, representing the Github release build you want to syncNF_CORE_BOT, with the bot credentialsAn adaption of the script COULD however allow users to pass their own credentials, either via ENV vars or command-line input. For that, the code needs to be refactored a little, as it expects the bot to perform all actions.
But I am not sure, if we should provide this functionality to be honest. I think the use cases are very rare, and the nf-core admins can run the script manually, if necessary.
What do you think?
Sven
Thinking the same - just maybe document that in the developer docs for the nf-core team?
Yes, docs are crucial!
nf-core/nf-co.re#105
Done :)