Brian Lloyd wrote:

...

What if we change this to:

if a 'content_type' attribute exists, it is used

else if the object's id seems to have a file extension (. in the id), try to use guess_content_type

...

else if there is a 'default_content_type' attribute, use that

...

else fall back to the hated application/octet-stream (or, I've seen x-unknown-content-type used before, but I don't know if this is really any better).

FYI, for those interested I've checked this and another fix for HEAD handling on collections into CVS... Brian Lloyd brian@digicool.com Software Engineer 540.371.6909 Digital Creations http://www.digicool.com