optional timestamps (left over from TimestampVsCreationDateTime)
version identifier, a string that changes each time the entry is modified
contributor, a thing that helped write the entry
(all the same children as author)
Title: Community Gathers to Introspect
Summary: In a move that shook the foundations of the Internet, thousands gathered to decide what, exactly, constitutes a weblog entry. The idea is to build a sort of conceptual model, which would be the underpinning of something much bigger.
subtitle pulled 7/4/2003 -- see SubTitle for a summary of the arguments
Prior Art: Qualified Dublin Core (See Relation refined elements), ResourceDescriptionFramework
Expanding on this thought: Norman Walsh writes blog entries in DocBook and transforms them using XSLT. LiveJournal mainly uses direct HTML markup but has a set of application-specific tags to supplement. Neither is useful to the general public, but is *vital* to the original author who may want to come back later and edit previous posts. So not useful for syndication (would just syndicate the transformed output), but very useful for editing software like KungLog, W.Bloggar type software.
[JamesSnell] Question: Is this more of a type of ContentModule? In other words, the Title/SubTitle/Summary collection taking the place of the complete content representation? A very good question - either make this true Metadata or move Metadata out to a separate block
[MichaelBernstein] technically, a lead is the first paragraph of the story. In many cases, in a well written newspaper story (inverted pyramid style) the lead paragraph is also summary of the entire story.
Metadata is a poor name. Every bit of data about an entry is entry metadata, for example, Categorization. Here are three alternate suggestions:
call it Miscellaneous
call it Entry Description, referring to the entry as a whole. Move (sub-)descriptors of author, content, permalink, date elsewhere.
split the data above back into separate modules
Metadata is a good name for the stuff that *isn't* a ContentModule (as James mentions). Strictly speaking every bit of data about an entry is entry metadata, but in common usage people know the difference. There's all the stuff that's intended to be machine-readable only. This stuff is very important! One conceptual split is done in Unified RSS. I'd suggest renaming the stuff described above something like Summary and using Metadata in its current common usage. Also (where can it go but here?) I'd suggest a <metadata> tag. [DannyAyers]
[Moof, RefactorOk] Having come across this one before in publishing a blog, Metadata should also be extensible. At one point in my blog, where I was publising and entry from was an importan bit of metadata botu the post, but I ended up with the deision of either bundling it into the content, or leaving it out. I don't we know, even now, what people want to do with blogs, so limiting the possible metadata to just a standard set is not necessarily good. Hell, I still know blogs and other published data which has entries with no titles. A representation of a fortunes file, for example.
[BryantDurrell, RefactorOK] Moof: agreed. See NamespaceDiscussion for more on this.