On 3/21/17 10:49, Pete Resnick wrote:
1.
The 666 mechanism leaves the decision of what is meant by
"unwanted" to the provider, which will inevitably cause data loss
through the use of heuristics. Adding a header to the 603 response
is an extensible mechanism that could capture the single semantic
bit of 666 /and any future semantics that one wishes to add/,
without requiring the provider to guess. The provider gets
definitive information that it can act on.
The problem is that the semantics of 666 are backwards from the
semantics of 603.
603 means "there is an issue with the disposition of the *called* party
that prevents completing the call."
666 means "there is an issue with the disposition of the *calling* party
that prevents completing the call."
You're proposing conflating the two semantics, which necessarily
introduces more ambiguity, not less.
Carrier determination of how to handle these codes is, as with any
nuisance communication mitigation, going to necessarily be based on
heuristics to avoid accidental blacklisting, intentional blowback, and
the inevitable gaming of the system by bad actors. You imply that the
application of heuristics is an unwanted side effect of the system, when
it in fact the primary goal.
In terms of being able to provide gradation between types of unwanted
calls, the application of a header -- as you propose -- to provide such
indication seems like a fine idea. I will point out one caveat and make
one observation. The caveat is that it would be harmful to add this
header in a way that reverses the sense of a response code, such as
making 603 bear on the calling party rather than the called party; so
such an indication would need to be on a new response code, such as the
one currently proposed in draft-ietf-sipcore-status-unwanted. The
observation is that this mechanism can be added to this new status at a
later date, in a modular fashion and in a backwards compatible way. In
other words: your proposed enhancement is *nice*, but it does not
preclude publishing the mechanism currently described.
2.
The 666 mechanism limits the implementation of the mechanism to a
1-button choice and requires additional standardization work to
use a different UI. Clearly 666 was designed around something akin
to a "SPAM" button in the UI. But what if I want a field in my
address book that says, "Permanently block this particular caller"
(say, my ex-boyfriend or my annoying neighbor)? I don't want to
indicate that these are spam calls because I don't want my
provider applying heuristics to them. I want to send back a 603
with a header that says, "Block these whenever you see them, but
only for me." I can think of all sorts of other-than-one-button
UIs that someone might want to implement. 666 is tied to a
particular UI. 603 with a header is extensible.
That kind of persistent, hard-state user account configuration *really*
isn't a reasonable kind of thing to do with a SIP status code (not least
of all because you would need some mechanism for reviewing and managing
such a list; and whatever that mechanism is could just as readily be
used for adding entries). If you feel the need to pursue the somewhat
related problem space of a network-assisted blocklist, we have a variety
of tools we could bring to bear, such as XCAP. That's a separable
problem, though, and I think that tangling it up with spam mitigation is
a huge mistake.
I'll also point out that devices are perfectly capable of handling this
feature locally without any assistance from the network, so I'm not sure
how much value would be added by defining a network-hosted version of
the service. The overarching point, though, is that a SIP response code
is absolutely the wrong approach for doing so.
3.
The 666 mechanism will be treated as a 600 "Busy" error by
un-upgraded callers and/or providers. That might imply to some
implementations that they should try to re-dial (e.g., the
provider using the auto-redial feature) or to engage in other poor
behavior. 603 with no Retry-After header already has a semantic of
"The call was rejected and I'm not telling you when a good time to
call back is", so there is some hope that an un-upgraded caller is
more likely to do the right thing.
This is really reaching, and I think is based on a misperception of how
SIP clients actually work when they receive 6xx responses. Taken to its
logical conclusion, if we accept your assertion then we can't ever
define a new 6xx-class SIP response code for fear that such codes "might
imply to some implementations that they should try to re-dial (e.g., the
provider using the auto-redial feature) or to engage in other poor
behavior."
/a