This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
from "SML supports three attributes: " to "SML defines three attributes: " from "...the same name as P must specify the same set ..." to "...the same name as P MUST specify the same set ..." from "...as P and these attributes must have the same values..." to "...as P and these attributes MUST have the same values..." from "...all of the following must be enforced:" to "...all of the following MUST be enforced:" from "The above conditions on the use of sml:target* attributes..." to "Note (non-normative): The above conditions on the use of sml:target* attributes..." from "...implementation burden on model validators for verifying..." to "...implementation burden on model validators for verifying..." (remove 2 extra spaces around "on") from "... and they MUST be supported on global and local element declarations. " to ". Model authors MAY specify them on any <xs:element> in a schema document. " (steals more precise language from acyclic and separates concerns) move "Model validators that conform to this specification MUST enforce the following:" AFTER paragraph break that follows this text today. Recommend taking all the text starting with "Model validators that conform to this specification MUST enforce the following:" and ending before "4.3.2.1 sml:targetElement" and making it a new h4 "Derivation by Restriction". I would also recommend placing that new section AFTER the 3 target* attributes it applies to instead of before.
Made all of the suggested changes except the following one: from "... and they MUST be supported on global and local element declarations. " to ". Model authors MAY specify them on any <xs:element> in a schema document. " IMO, what model authors may choose to do is not the same as what an implementation must support. If I am wrong in this assumption then feel free to reopen the bug.
The part of the change that mattered more was missed in comment #1. The existing version says local and global, the new version implicitly says global (omitted local consciously, since I believe we currently say elsewhere they are only allowed on globals). No quibble with the model author vs implementor part.
Mistook xs:element for saying not local, which is incorrect. So fine as fixed, closing.