Martynas Jusevičius
Martynas Jusevičius
When submitting a `multipart/form-data` form, Jersey returns `org.jvnet.mimepull.MIMEParsingException: Missing start boundary`. This seems to be [specific to [Chrome/IE](https://liviutudor.com/2016/03/21/issue-with-jersey-multipart-libraries-missing-start-boundary/).
Because containers in the Docker network can by default access each other, a user of the end-user endpoint (`fuseki-end-user`) can potentially run such a federated query which will expose data...
Saxon-JS in the browser struggles to render large SPARQL results (e.g. in container model or in the SPARQL editor). But it should be trivial for `client.xsl`/`container.xsl` to request `text/html` instead...
OpenLayers [support WKT](https://openlayers.org/en/latest/examples/wkt.html) literals. See also a PoC which uses TypeScript: https://github.com/AtomGraph/WKTMap
It should be rather trivial to replace most of the legacy jQuery-based code in [`jquery.js`](https://github.com/AtomGraph/LinkedDataHub/blob/master/src/main/webapp/static/com/atomgraph/linkeddatahub/js/jquery.js) with an IXSL (Saxon-JS) implementation. We need to keep the `fetchDispatchXML` function though.
We should have a facet type for string values which would filter results by `regex()`.
There were some reports that UI doesn't change after authentication with Google. We should investigate whether the `ETag` header value changes in this case as it should.
Currently the CLI scripts only support authentication with WebID-TLS client certificate. Looks like it should be possible to enable OpenID authentication as well: https://www.janua.fr/bash-oauth2-openid-connect-wrapper/
Saxon-JS 2 attempts to parse XML body from a non-XML response, e.g. Tomcat's HTML error page: https://sourceforge.net/p/saxon/mailman/message/37142695/ Waiting for reply from Saxonica.
