Closed fredjean closed 11 years ago
@fredjean Could you please help me understand the impetus for this? Is the gem currently pulling chunks that aren't related to the original file? Is it ignoring these extra chunks? Is this something we could add a unit test for too?
My original math was wrong on which chunks to pull from the database. It ends up taking one extra chunk per request for byte ranges that are inside the file. The code already only returns the bytes requested, so there is no external signs that we are pulling extra information from the DB.
Thanks. I also found #6, so I better understand now.
Restarted the fix from a clean fork. It should merge cleanly now.