DntResponseHeaderOrURI
Scale used:
- -- = Not supported
- - = Barely supported
- + = Supported with limitations
- ++ = Unlimited summort
Comparison:
- V01: At this point, the table summarizes mschunter's personal opinion (2012-02-24)
Criteria | Response Header | Well-known URI | Notes |
Transmits tracking status for retrieved object at URL | ++ | +(may need dynamic generation) | MSchunter believes that sending individualised (per user) status is easier with headers since they can be locally generated and attached. This seems easier than dynamically generating a centralised resource at a well-known URI. |
Enables enforcement by regulators | ++ | + | Headers are slightly less ambiguous due to their narrow scope |
Granularity | ++ (per request; very fine-grained) | + (per sub-site; coarser-grained) | To support DNT for each URL, the site needs to be replicated under the well-known URL. |
Simplicity of user agent | ++ (info along with page) | 0 (separate lookup) | |
Traffic generated | ++ (info along with page) | 0 (separate lookup) | It seems as -worst case-, the URI proposal may generate 2x the number of requests? |
Robustness wrt caching | - | ++ | Some cache implementations do not properly handle headers while all caches can cache a well-known URL. |
Tracking protection on info resource | + | - | URI can be declared 'no tracking area' and retrieved before actual request. For headers, tracking may have happened while returning a header? |
Maintainability of site | + | - | URI requires all site stakeholders to agree and populate the well-known URI (difficult); header allows resource-owner to just attach a header |
Ability to transmit complex data | - | + | Status object can transmit larger data sets (partner-sites, options, ...) while this would overload header. |
[More criteria needed...] |
Comments / Questions for Response Headers:
Comments / Questions for Well-known URIs:
- Is there a way to prevent that each URL needs to always be checked at the well-known location? E.g., retrieving foo.com/bar/one requires checking foo.com/.well-known/dnt/bar/one. If I now want to retrieve foo.com or foo.com/bar/one/sub, I need to re-check. Don't I? Wouldnt this double the web traffic (sort-of?)