Benjamin,
> On Feb 16, 2024, at 1:04 PM, Benjamin Gordon <bmgordon at chromium.org> wrote:
>> I read through this and had a couple of quick thoughts:
>> 1. On page 7, step 4 uses a different verb tense from the other steps.
Thanks, I've made note of this.
> 2. This specifically talks about IPP-USB. A lot of printers
> broadcast some kind of initial setup SSID when they're first plugged
> in. Does it make sense to allow this mechanism when connected to that
> SSID as well? This would be an improvement in convenience for the
> user, especially in cases where they might not be planning to put the
> printer within USB range of a computer.
>> Obviously it's not as secure as being physically plugged into the
> printer, but in most home networks, the window where an attacker is
> going to come connect a new printer to a rogue network is going to be
> small. Since that SSID is specifically used for setting up the
> printer, an attacker could potentially just connect to it and set up
> the printer anyway.
>> Does this deserve some discussion in the document, either as a
> possible implementation or as a reason why it's a bad idea?
Probably worth discussing. FWIW, such usage would be trivial to exploit, from a distance, through walls, etc. With proper authentication it would be fine, just not without some access control.
________________________
Michael Sweet