On 11/14/17 14:17, Eliot Lear wrote:
Hi,
On 11/14/17 11:42 AM, Mark Nottingham wrote:
- HTTP header insertion / "enrichment" (in many scenarios)
- Content filtering (to some people, in some situations)
- Performance-enhancing proxies (in some deployments, as they add to ossification)
- Customer access monitoring (to some people, in some situations)
- Content compression (depending on how it's interposed)
Perhaps the devil is in the detail in each of these cases, but in
sorting through them there are probably authorized ways to accomplish
certain goals, while others might need to be left behind. It seems like
in most of these cases it comes down to a matter of authorization, and
the relationship between parties. I don't think that's for this
document to delve too deeply into, but as one decides which problems to
work on later, at that point, teasing out those aspects may be
worthwhile. Starting with a statement of what the issues they are
facing are is pretty much factual. Whether we choose to address those
issues over time is separate.
And the goalposts go shooting off into the distance.
Mark is right. The questionability of some of these practices in some of
the ways they have been deployed argues against the use of the word
"ideally".
/a