- Instead of downloading the whole ~1MB file, download just the first 16KB. With fallback to the old way in case of fail - I've come across some technical difficulties as to how to limit the download (current implementation is too basic/simple for that), but in the end, a simple http header (Range) was enough