#1 Temporary file problem

[philosophy]

Temporary files are created using os.tmpfile() . On my system this points to a fast RAM file system that runs out of space for large downloads - resulting in http-replicator crashing.

Http-replicator should use an explicit directory for this purpose (the cache directory would be fine - since this would typically be a large disk-based file system).

A possible solution would be to use the python module "tempfile", which provides the function tempfile.mkstemp - this allows "dir" to be specified.