Hi Norm, Some responses below. On 2/8/2019 2:37 AM, Norman Walsh wrote: Bob Stayton <
bobs@sagehill.net> writes: Yes, by default the formatting of legalsection is indistinguishable from section. It is up the implementor to customize it if they want to. Making it an element makes it easier to customize formats, easier to find and maintain, and easier to extract for review by a legal department. I understand those benefits, but allowing it to be randomly interspersed with other ordinary sections seems odd to me. I don t understand what semantics are being represented by a section that is indistinguishable from its surrounding sections but has a different element name. The intended semantics are in the content: legal statements. The content model differs by allowing only legalsections as subsections. We already intermix simplesect, which has even less semantic distinction, differing only by not allowing subsections. I also forsee a day where someone has used cut-and-paste to carelessly spread legalsection throughout their document. I don't know we how can guard against such behavior, but I think the byword here is carelessly . Regarding numbering, the stylesheets currently number section, sect1 through sect5, and simplesect as the same numbering set. We would add legalsection to that set. I'm not clear where the complexity you are referring to comes in? *shrug* It s not a huge complexity, but now counting requires matching across yet another element name. And note that section and sect(n). Correct me if I'm wrong, but I believe you solved the problem of counting section, sectN, and simplesects in the stylesheets. I was planning to extend that to include legalsection, taking care that legalsection can nest legalsection but not other types of sections. Speaking of sect(n), how does legalsection fit into that hierarchy, or does it not? We have not yet written out the schema changes, but I think it fits into the hierarchy the same as simplesect, so wherever simplesect can appear, so can legalsection. After we write and test the schema changes I may have more to say about that. Bob Stayton