Your encoding rules are very interesting.
I understand your series of <milestone> : folio, page, column… instead
of a more disparate series like <pb type="page"/>, <pb type="folio"/>,
<cb/>, and in this logic, <milestone unit="section" rend="3-lines" />
"EndColums" is an interesting notion. I haven't found an example in your
Livingstone papers to be sure to understand. I'm figuring it like the
section break in a text processor, an event in the flow to change column
layout (here, from n to 1). The word processor section point could be
said "a priori": « From this point, all pages will be in 3 columns ».
Your "endColums" seems less verbose, saying at the end of a set of
columns, no need to count, that's all, and for example, close a </table>.
About <milestone unit="section" rend="line" />, HTML seems here more
accurate to reality. Isn't it an <hr/> ? Sadly, there is no standard way
to say <hr/> in TEI. All TEI applications may have local conventions for
that, making the texts impossible to share without a human understanding
of practices and documentation.
I don't want to introduce my nationalism in my encodings with something
like <milestone unit="section" rend="astérisme" />. An American will
damn me when he will search the 'é' on his keyboard, but I keep your
excellent suggestion of the keyword "section", it is something like a
Because I want to keep the rule that all characters from source should
be kept, an empty element will not work form me. I understand that a
page number <pb/>, or a gathering(?) number <gb/> is outside the flow of
reading, but an asterism is inside the flow, and should have its chars
in a text node.
<ab type="section" rend="center">⁂</ab>
<ab type="section" rend="center">* * *</ab>
<ab type="section" rend="center">*<lb/>* *</ab>
@rend="center" is redundant, but redundancy may sometimes secure
communication of information. A very agnostic TEI application, like for
example Sebastian XSLT
will keep appearance and let the text readable.
>> About milestones for "couillards", I'm not agree. My understanding of
>> milestones is the covering of the text with a sectioning scheme
>> incompatible with the logical tree, like pages <pb/>, or thematic
>> analysis. "couillards" are always compatible with the paragraphs flow,
>> and the sections.
> For what seems to be this very purpose, The David Livingstone Spectral
> Imaging Project uses
> <milestone unit="section" rend="line" />
> <milestone unit="section" rend="2-lines" />
> <milestone unit="section" rend="3-lines" />
> Had Livingstone been French, the project might be using
> <milestone unit="section" rend="astérisme" />
> -- JPM