Multi-fetch across partials, better performance Now multi-fetches across partials better, enabling setups that use a single network call for entire collections of chunked caches. Also vastly improves performance when writing the chunks back into the template string, to the order of 10-20x faster.