search

ebi-ait / hca-ebi-dev-team

Repository for hca ebi dev team agile management. See zenhub board
0 stars 0 forks source link

Metadata documentation website onto schema.humancellatlas.org #135

Open clairerye opened 4 years ago

clairerye commented 4 years ago

As a contributor, I want to find information about the HCA metadata schema at the schema.humancellatlas.org address.

We need an avenue to provide information to the community about our metadata standards. They need to be able to pre-plan their metadata collection, understand how what fields are required and what fields are option and understand the design and evolution principles of our schema.

@mshadbolt has produced a prototype site which is hosted here: https://ebi-ait.github.io/hca-metadata-community/

The contents of this site should be hosted alongside the JSON schema definitions themselves on https://schema.humancellatlas.org/

Requirements

Acceptance criteria

clairerye commented 4 years ago

@lauraclarke can you add a better description of what you want here, in terms of being able to host the metadata community documentation from the github website that Marion created and also keeping the schema stuff that is already on schema.humancellatlas.org

lauraclarke commented 4 years ago

@mshadbolt @zperova Do you think I missed anything from high-level requirements/acceptance criteria?

mshadbolt commented 4 years ago

I think making the metadata dictionary automatically updated from schema would need a different ticket.

lauraclarke commented 4 years ago

Should we turn this into an epic and have one ticket for static content and one for automated update of the metadata dictionary?

zperova commented 4 years ago

If this work is going to be done by the devs, shouldn't they be asked this question?

mshadbolt commented 4 years ago

Should we turn this into an epic and have one ticket for static content and one for automated update of the metadata dictionary?

I see them as independent jobs personally.

lauraclarke commented 4 years ago

If this work is going to be done by the devs, shouldn't they be asked this question?

This work needs collaboration across the devs and the wrangers and depending on the task an individual with the right skills should complete it.

lauraclarke commented 4 years ago

Should we turn this into an epic and have one ticket for static content and one for automated update of the metadata dictionary?

I see them as independent jobs personally.

What is the current process for updating the dictionary?

mshadbolt commented 4 years ago

generate a spreadsheet from the schema_template_generator with all fields, run a script to do some filtering and cleaning up and transposing, save in a folder that the app recognises, deploy to shiny.io.

I think that the website as it stands could be deployed to https://schema.humancellatlas.org/.

The metadata dictionary is a separate entity and decisions need to be made about how it should be implemented in the future. Either editing the current version to query the schemas themselves within R or transitioning to a different implementation.

gh-pages is a static hosting service so if we continue on this then we couldn't have a service that was fully automated and would update when schemas are updated unless it is hosted somewhere else.

Anyway there are many considerations that require a conversation about it going forward but I don't think it should block the documentation moving somewhere that is accessible by the community.

Also doesn't seem that important to have it automatically update since we can't update schema at the moment anyway