The write function for download currently downloads each part to a temporary file on the disk and then joins them to create the final file. However, it would be much more efficient to do this in memory. I.E writing to a object instead of a file. We would then join all these chunks to create the final file.

An example of this in C can be seen [here](https://curl.haxx.se/libcurl/c/getinmemory.htm