as a repository operator, I'd like search engine discoverability to not break #5406
Description
Spin-off from #5400 ; having multiple <head> elements in generated HTML may prevent search engines from correctly parsing meta-tags, schema.org json-ld and other machine-readable parts of pages that used to assist discoverability.
It would seem reasonable that any pages a dataverse installation would like to have show up in search results to be checked.
Pages:
- dataset page (should be ok once 5400 duplicate html tags #5405 is merged)
- dataverse page
- (custom?) home page
- other pages?
How to check:
- no fatal errors with vnu (https://validator.github.io/validator/), or online variant. Getting pages to 0 errors is explicitly out of scope here; goal is to have the generated html parsable to the point of being able to report how many errors there are.
- visual inspection?
- other / better ways?
Metadata
Metadata
Assignees
Labels
No labels