On 11/16/14 10:58 AM, Christian Huitema wrote:
By creating a standard, we would be creating a social norm. It would not take long for regulators to mandate "safe" behavior for web sites, or to enforce the safe bit in various kinds of "great firewalls." It will all be in the name of protecting the children, but we all know that the real target will be dissent and free speech. By offering this setting as a standard, the IETF would become an accomplice of repressive regimes and other religious dictarures.
The Internet routes around damage. Let's say you're in one of those repressive regimes. The web sites in your country are already heavily regulated, so this feature is of no consequence. Any sites that express dissent about your regime are going to be in regulatory domains where free speech is permitted, and are free to ignore the safe option.
The only thing I see wrong with this is the one bit. I would prefer to see one byte, with a standard meaning developed for the bitmask. Something like:
1 Filter pornographic images and language 2 Filter violent images and language 4 Filter offensive language 8 Filter everything I'm not married to this scheme, but I think it's a good start.
Some features do not need to be standardized.
Interesting. I use this same argument in the DNS protocol world for things like identifying the requesting client's subnet, and negative trust anchors. I'm regularly shouted down and told that since there is already running code we MUST document it for interoperability purposes. Can I quote you? :)
Doug