For example, the TEI consortium publishes HTTP URIs for standard TEI schemas, but when you resolve those URIs you are redirected to an HTTPS URI which then responds with the actual schema. The response to the HTTP URI is a "301" redirect with a message body containing HTML (saying "301 Moved Permanently").
What should happen is that the Linter should follow the 301 redirect and retrieve the actual schema.
What currently happens is that the Linter ignores the 301 response code, and attempts to interpret the HTML response body as a RelaxNG schema, which naturally fails.
Example XML document (which should validate according to the linked schemas):
For example, the TEI consortium publishes HTTP URIs for standard TEI schemas, but when you resolve those URIs you are redirected to an HTTPS URI which then responds with the actual schema. The response to the HTTP URI is a "301" redirect with a message body containing HTML (saying "301 Moved Permanently").
What should happen is that the Linter should follow the 301 redirect and retrieve the actual schema.
What currently happens is that the Linter ignores the 301 response code, and attempts to interpret the HTML response body as a RelaxNG schema, which naturally fails.
Example XML document (which should validate according to the linked schemas):
NB if you change the
http
scheme withhttps
in those schema URIs, it does work.