One thing is that we'll make testing doc changes to repos like https://github.com/apache/incubator-pekko harder.

Today, if you run sbt paradoxBrowse, it will build the docs into target/paradox/main/site and you can load up the index.html page from there and the CSS, etc. will be found. In theory, users could copy the missing files into target/paradox/main/site themselves or we could script something.

In the end, it might be easier if we don't rely on relative paths and hardcode that we can get the static files from explicit URLs. We could make the hostname of these URLs configurable. Default to https://pekko.apache.org but users could set https://pekko.staged.apache.org.

One thing is that we'll make testing doc changes to repos like https://github.com/apache/incubator-pekko harder.

Today, if you run sbt paradoxBrowse, it will build the docs into target/paradox/main/site and you can load up the index.html page from there and the CSS, etc. will be found. In theory, users could copy the missing files into target/paradox/main/site themselves or we could script something.

In the end, it might be easier if we don't rely on relative paths and hardcode that we can get the static files from explicit URLs. We could make the hostname of these URLs configurable. Default to https://pekko.apache.org but users could set https://pekko.staged.apache.org.
@pjfanning
I have added the assets.hostname variable for handling the hostname. Default is configured to https://pekko.apache.org