Verify bioschemas in production environment about metabolicatlas

[mihai-sysbio]

If I understand correctly, this is issue is only about verifying in the production environment. My follow-up question would be why can't this be verified locally though, knowing it would be the same in the production environment?

[e0]

The markup (HTML document) in local development mode is not the same as in production. This issue is a reminder to double check that things are done correctly after deploy.

[mihai-sysbio]

The markup (HTML document) in local development mode is not the same as in production

@e0 besides the host url changing, what other changes in the HTML should we be aware of?

[e0]

There is a lot of Vue specific stuff that's helpful during development but is stripped away for production. There is also other type of verifications (which I should have mentioned in the original issue description such as link previews when sending over Slack for example and verifying search engine titles and summaries) that makes more sense to test only on production.

[nanjiangshu]

Hi @mihai-sysbio and @e0, I've deployed the branch feat/bioschemas_MA-675 on the dev server and verified the Genes page using Schema Markup Validator. The profile looks fine as shown by
https://validator.schema.org/#url=https%3A%2F%2Fdev-metatlas.csbi.chalmers.se%2Fexplore%2FHuman-GEM%2Fgem-browser%2Fgene%2FENSG00000136010

However, title and description are not interpreted correctly by a search engine or Slack.
I've created a simple test page https://pcons1.scilifelab.se/test1.html and it seems Slack can catch the metaInfo.

[e0]

@nanjiangshu Thanks for deploying and trying this out. I was wondering if it was maybe due to caching but I realized it is probably because these fields for title and description are populated after the webpage already loads and it makes another request to get the component data. To get around this issue, I think we would need to do some SSR. That would be a much larger issue though with changes to our infrastructure so we should probably have a talk before deciding if we want to make this change.

For now, I think the best things to do is to only focus on the bioschemas for genes and metabolites in this issue.