Alexander Shorin created COUCHDB-2353:
-----------------------------------------
Summary: Normalize special CouchDB headers name
Key: COUCHDB-2353
URL: https://issues.apache.org/jira/browse/COUCHDB-2353
Project: CouchDB
Issue Type: Improvement
Security Level: public (Regular issues)
Components: Database Core
Reporter: Alexander Shorin
So far we have:
- X-Couch-Id
- X-Couch-Update-NewRev
- X-Couch-Full-Commit
- X-CouchDB-VHost-Path
- X-CouchDB-WWW-Authenticate
- X-CouchDB-Requested-Path
- X-Couch-Request-ID
- X-Couch-Stack-Hash
- X-CouchDB-Body-Time
- X-Couch-Test-Auth
- X-Auth-CouchDB-Username
- X-Auth-CouchDB-Roles
- X-Auth-CouchDB-Token
First, I'd like to see them sharing common prefix: X-Couch or X-CouchDB, not
both. For consistency. Personally, I think X-CouchDB is better name since it
completely matches with the project name.
Second, some name cleanup would be nice: it doesn't oblivious what X-Couch-Id
belongs to while there is X-Couch-Request-Id and X-Couch-Update-NewRev is too
verbose: X-Couch-Doc-Id and X-Couch-Doc-Rev would be more clean IMHO. Or
X-CouchDB-*.
Third is an idea to drop X- prefix as per
[RFC-6648|http://tools.ietf.org/html/rfc6648]. Personally, I don't like this
RFC as like as idea of pollution standard headers name space with custom ones,
but what do you think about? Is it worth to keep or drop X- prefix?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
