Hi Georg

<g> and <c> are really animals of very different kinds!

tei:c is as you say a "linguistical entity" -- it represents an interpretation of the content. Like tei:pc and tei:w and others defined in the analysis module, its purpose is to delimit strings of unicode characters which the encoder wishes to consider/analyse as segments of some kind

tei:g however is not markup at the same conceptual level: it is the way we get around the fact that tei users often want to mark up non-Unicode glyphs or characters. It should not itself contain anything other than Unicode characters -- and may be empty.  This is because it is really just standing-in for a Unicode character which doesn't actually exist yet.

does that help?

On 26/06/14 21:11, Georg Vogeler wrote:
[log in to unmask]" type="cite"> By the way:

I see that the linguistical entity "character" can be represented by (one or more?) glyphs, which allows tei:c/tei:g.

I see that the glyph can represent graphical phenomena like ligatures (example in doc/tei-p5-doc/en/html/ref-g.html).

In my mind that would mean, that one glyph can contain several linguistically defined characters. But the Guidelines/the schema has choosen by some argument, that it might be a bad idea to do that by including a tei:c element into the tei:g.

Again: What have I missed?



Am 26.06.2014 20:28, schrieb Lou Burnard:
[log in to unmask]" type="cite">
This looks to me like a corrigible error. If someone would like to submit it as a bug report on sourceforge, I will fix it pdq.

Currently zone and line have a content model of

( text | model.global | model.gLike | model.linePart )

this should be

( text | model.gLike | model.global | model.gLike | model.linePart )

On 26/06/14 18:56, Matthew Davis wrote:
[log in to unmask]" type="cite">
Dear list,

I just want to piggyback on Georg's comment that I've noticed something similar with <line> -- if I try to put my text in directly I get an error, but if I wrap it in <orig> first I'm ok.  I haven't really understood why this should be the case.

All best,

On Jun 26, 2014, at 1:53 PM, Georg Vogeler <[log in to unmask]> wrote:

Dear list,

I've certianly missed an argument or appropriate bug report, which would explain, why the g-Element isn't allowed in tei:zone. I'm preparing a repository of letter forms connected with zones of a manuscript image. I put my zone around the letter and would like to encode it as a specific glyph. I feel completely in line with the guidelines by placing my "transcription ... directly within the zone element" <file:///C:/Tempdat/TEI2.6/doc/tei-p5-doc/en/html/PH.html#PHZLAB>, but the 2.6.0 rev. 12802 tells me, that I can't put a tei:g inside tei:zone. Do I really have to put a tei:seg or tei:c before entering my text?