The concept is that a save triggered by autosave could record just the changed tiddlers (so it's very low bandwidth compared to a full file save), similar to the way autosave works when running with a TiddlyWiki node js server. Tiddlyhost could record those saves as-is, then apply those changes in chronological order before serving the site. Clicking the save button could perhaps save the entire TiddlyWiki as it does now, and clear any incremental save records. There might be some interesting multi-user implications.
The concept is that a save triggered by autosave could record just the changed tiddlers (so it's very low bandwidth compared to a full file save), similar to the way autosave works when running with a TiddlyWiki node js server. Tiddlyhost could record those saves as-is, then apply those changes in chronological order before serving the site. Clicking the save button could perhaps save the entire TiddlyWiki as it does now, and clear any incremental save records. There might be some interesting multi-user implications.