[COUCHDB-1128] Erroneous "identity" Content-Coding supplied to Content-Encoding header for uncompressed attachments

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 1.0.2
• Fix Version/s: 1.0.3, 1.1
• Component/s: HTTP Interface
• Labels: http

Environment

CouchOne

• Skill Level: Regular Contributors Level (Easy to Medium)
• Assignee: Unassigned
• Reporter: Andrew Gleave
• Created: 18/Apr/11 17:04
• Updated: 19/Apr/11 20:40
• Resolved: 19/Apr/11 20:40

Description

According to the HTTP / 1.1 spec, the "identity" content-coding is to be used only in the Accept- Encoding header, and should not be used in the Content-Encoding header. Currently, any uncompressed attachment which is retrieved from CouchDB will have Content-Encoding:identity specified in its response header.

http://www.w3.org/Protocols/rfc2616/rfc2616-sec3.html#sec3.5

All browsers and most proxies seem to ignore the header and function correctly. However, Microsoft's ISA 2006 proxy seems to reject the entire request, meaning all attachments are inaccessible from behind the proxy.