`content-type` should default to guess
Closed this issue · 2 comments
By now, I've accidentally uploaded multiple days worth of data to S3, only to find it unusable due to the content-type being application/octet-stream. This renders images, html, css, etc, unusable by default.
The default logic should be safe to use and therefore: the default content-type option should be guess.
This will need to take into account the gzip options logic to ensure it doesn't break gzip when no content-type is specified.
+1 I've just arrived at the same problem
@adamlwgriffiths if you're looking for a solution to fix your files in the meantime, I'd recommend:
I had to re-upload the files with --content-type=.
Luckily the files were all the same.
An easier way would be to write a script to traverse the bucket and modify the meta data values, but it was easier to start a re-upload in the background than to spend the time writing said script.