Permit @rendition on <handNote> to encode different script type

[danbalogh]

The inscription I'm encoding at the moment changes to a palaeographically later script in the middle of a stanza. It's a set of reused plates, where most of the praśasti (probably from the preceding ruler) has been retained, but after some point the old text was beaten out and re-engraved by a different scribe in a different style ("regional Brāhmī-derived" to "vernacular Brāhmī-derived").
Our existing methods of encoding script class/maturity aren't really suitable in my opinion:

• encoding a default script for the edition, and then adding @rendition on every container after the hand change would be very cumbersome
• encoding textpart divisions for the two parts doesn't sound very nice (although it's doable), since a stanza (and a line within the stanza) would have to be split between the textparts, and the text is coherent as a whole

Add to this the fact that a hand change had best be encoded anyway, and I think it would be best to encode the script type on the hand note in the header (and then just use handShift in the edition at the applicable point, as normally). Thus, I propose

<handNote xml:id="DHARMA_INSVengiCalukya00104_hand1" rendition="class:83225 maturity:83213">human-readabe description</handNote>
<handNote xml:id="DHARMA_INSVengiCalukya00104_hand2" rendition="class:83225 maturity:83215">human-readabe description</handNote>

I think it would be best to retain rendition on the edition division as well (for the "default" script type), to avoid needing another exception in the schema. Also, if rendition is used on a handNote, then it should be mandatory for all handNote elements to exclude any ambiguity.

Any objections to this? Any alternative suggestions or improvements?

[michaelnmmeyer]

I would propose two modifications.

Firstly, to use @rendition on handShift instead of on handNote. The main reason for this is that handNote is a text container and that @rendition is meant to describe the contents of the container i.e. the actual text within handNote. On the other hand, handShift is an empty element, so adding a @rendition to it would not be ambiguous.

Secondly, I think we should only allow a single way to indicate @rendition in a given file: either the "traditional" top-down method (with @rendition on divs), or the new one (with @rendition on handShift), but not both, because it is not quite clear which @rendition should take precedence on conflict.

For the handShift method to work properly, a file should have an initial handShift with @rendition before the beginning of the text. It would then be assumed that everything that follows (in file order) upto the next handShift with @rendition (or the end of the file) is in the given script.

In any case, this is not trivial to deal with (like all the milestone-related stuff: pb, lb, etc.).

[danbalogh]

Thanks, @michaelnmmeyer , your modifications sound perfect to me. Should we then adopt this?
Could you elaborate what you mean when you say it is not trivial to deal with it? I don't think we are planning any machine actionability for these. Note: I personally would be OK with a solution where script class and maturity are not encoded in the XML, only kept in the metadata. As I noted a while ago (#254), the redundancy (script classification in both the XML and the metadata) bothers me. There I suggest removing this from the metadata, but removing it from the XML may also be viable.

[michaelnmmeyer]

We can adopt this. I will add an exception to the schema, and write the processing code later (so I leave this issue opened in the meantime).

[danbalogh]

A slight complication to the matter: since we're dealing with a copperplate set, there are two textpart divs, one for the seal and one for the plates. So in the edition as a whole, I think the best way to encode script type would be as follows:

<div type="edition" xml:lang="san-Latn"><!--no rendition-->
  <div type="textpart" n="A" rendition="class:83225 maturity:83215">seal
  </div>
  <div type="textpart" n="B"><!--no rendition-->
    <handShift new="#DHARMA_INSVengiCalukya00104_hand1" rendition="class:83225 maturity:83213"/>text in hand 1...
    <handShift new="#DHARMA_INSVengiCalukya00104_hand2" rendition="class:83225 maturity:83215"/>text in hand 2
  </div>
</div>

Agree?

[michaelnmmeyer]

Noted, I will address this.