A Flow board is currently a single resource identified by a single URL, and content is exposed as the user scrolls down. That is an antipattern: scroll position is lost whenever a browser tab is reloaded, closed and reopened, or the user navigates away and back; search engines might not discover or might rank down part of the board, or links from the search result page might lead to a page which does not contain the searched expression.
A better way to handle that is paginated infinite scroll where the URL is updated to track which "virtual pagination block" the user is currently looking at. (Discourse has a nice implementation of that; example and explanation.)
See also: