Skip to content

protocol: Update secure-output protocol description

Scott Anderson requested to merge ascent/weston:secure-output-protocol into content_protection

This changes the wording to be more precise.

  • "Secure output" is defined, and other parts of the text now refer to it.
  • force/relax were changed to set_enforced/unset_enforced, to make it more clear that they are affecting the same state.
  • set_enforced/unset_enforced are now defined in terms of a "secure output enforcement".
  • A disclaimer was added to highlight the limitations of this protocol for general-purpose digital rights management.
  • The language was changed to more consistently talk about "content protection" instead of "security".

This is to try and help this along this protocol by giving it a more precise and clear description, and hopefully preserve the intent of the protocol. Most of the changes I've made are supposed to not require changing the implementation massively, but I haven't really checked; this is just supposed to be about the protocol itself.

I haven't done anything to address wayland/weston!48 (comment 143961) in particular. I think a new "unsupported" event or advertising them up front would be required, but it may be difficult to enumerate what's supported ahead of time.

There is no distinction between censoring content and hiding content; it's simply left as compositor-defined. I don't know if such a thing is wanted/required.

As some extra bikeshedding, a few of the names I haven't changed don't really match the new descriptions that well now. E.g. "set_type" would probably be better as "set_minimum".

Edit: Changed to set_enforced/unset_enforced, as it's more correct English.

Edited by Scott Anderson

Merge request reports