#422 1.3.2.1 needs repair

GREEN
closed-accepted
Lou Burnard
5
2012-09-23
2012-07-24
John P. McCaskey
No

Guidelines 1.3.2.1 needs repair and clarification.

It is not clear what elements fall into each of the five classes and even how one would find out.

There is a reference to model.inter, but there is no such model.

In the phrase, "within such divisions," it is not clear what "such" refers to.

Are the elements populating the eight classes model.divLike, model.div1Like, model.div2Like, . . . , model.div7Like the only divisions? Or is the "etc." more open-ended than that? (I suspect divGen would be included.)

Which elements are chunks? The language suggests all and only those in model.divPart, But the language also suggests that model.divPart has several subclasses, but there are only three and model.entryLike isn't one of them, though the language suggests it is.

Elsewhere (http://www.tei-c.org/release/doc/tei-p5-doc/en/html/DR.html#DRBOD) the guidelines indicate that chunks are "paragraphs, verse lines, stanzas or . . . ab elements". But there is no <stanza> element. In the guidelines, "stanza" is generally a @type of <lg> but is at least once a @type of <div>.

The Guidelines say that over two-thirds of the 500+ elements defined in 5.0 are in one of these "groupings." Are those "groupings" these "basic model classes"? But the opening sentence set the expectation that all elements do. Once the reader sees that they do not, then the relationship between the "basic model classes" of 1.3.2.1 and "model classes" of 1.3.2 becomes unclear, since the "basic model classes" have no unambiguous definition.

Discussion

  • Lou Burnard
    Lou Burnard
    2012-09-05

    • milestone: --> GREEN
    • assigned_to: nobody --> louburnard
     
  • Lou Burnard
    Lou Burnard
    2012-09-12

    The five "basic model classes" (BMC) are not quite the same as the other "model classes" and so maybe a better name should be found for them. In particular, not every model class is a subclass of a "basic model class". To answer your other comments:
    - elements are not assigned to a BMC, but rather to a model class. To find out what elements are in a BMC, look at the classes specified here.
    - model.inter is a model class like the others; please explain why you say "there is no such model"
    - the "divisions" BMC probably should include divGen (though this is an empty element and hence a bit un-div-like). Like other BMCs, it's an open-ended list
    - I agree that the specification for the "chunks" BMC is misleading : although model.entryLike is not a member of model.divPart, both are chunks. nmodel.divPart, for the record, has three subclasses and several other unclassified members.
    - I am unclear why the sentence you quote from #DRBOD is a problem. There isn't a <paragraph> or a <verse line> element either. It's true that a stanza may be represented by a <div type="stanza"> but a paragraph represented by a <div type="para"> is not inconceivable.
    - "groupings" in the penultimate para of this section is definitely wrong. The point is that "some elements and element classes" are not themselves classified, and thus cannot readily be assigned to any BMC.
    Thanks for the careful reading: I will try to produce a revised text in the light of these criticisms, after discussion with the TC.

     
  • Lou Burnard
    Lou Burnard
    2012-09-23

    • status: open --> closed-accepted
     
  • Lou Burnard
    Lou Burnard
    2012-09-23

    I have revised the text in an attempt to reduce the ambiguities. Please check revision 10866 and reopen this ticket if it still needs work.