problem

In a dockerized environment all services must be usable via network, same for the reflow via headless browser.

solution

• Change the usage of the headless browser used for reflow, so that it can be reached via HTTP instead of plain file access.
• Suggestion: the existing path config is evaluated, if it is a URL, the HTTP based approach is used, if not then the local one.
• As discussed the data must be accessible via local URL access. Doable as symlink from data to public task dir.
• Here keep in mind, that the whole data dir may be located somewhere else.
• Also the URL for that access from the browser must be separatly configurable, similar as the "runtimeOptions.worker.server" config
• So that new URL config defaults to runtimeOptions.worker.server (with the same "runtimeOptions.server.protocol" and "runtimeOptions.server.name" fallback applied)
• format should be as for worker config: SCHEME://HOST:PORT