[tei-council] Further update on PH

[Arianna Ciula]

I know we shouldn't wake ghosts up right now and concentrate only on the 
hight priority things for the release, but I wanted to say that after 
Syd has expressed the remark that <geo> and the definition of lat/long 
within it are rather unspecified, I have contacted one of the people - 
Leif Isaksen - that came at the meeting in London last May to discuss 
the issue further and make sure our specifications were refined enough.

here is part of our email thread:
-----------------------------
LI:
As far as I recall, my
 > primary concern was also to record the coordinate system/datum that any
 > listed coordinates/degrees/etc are based on ( i.e. they have no semantic
 > content without them).

AC:
Sure and there is an attribute for that now.

LI:
The point made above [AC: he was referring to a snipped of Syd's message 
here] seems to be that a
 > controlled vocabulary and standard format should be used to allow for
 > better interoperability ( i.e. they are only human readable otherwise).
 > I'm not sure how far TEI involves itself with such things, but it would
 > definitely seem to me to be a good idea.

AC:
I would think so, but I am not sure how much we can restrict the values
of the coordinates per se without considering the datum, that is in
combination the two attributes (coordinates and datum) could allow us to
create a rule, but could the vocabulary for the coordinates be
restricted without considering the datum?

LI:
The last is a good point. I think you might be able to get away with 
just two types - decimal (in which case you just need a floating point 
number of arbitrary precision), or degree (in which case you probably 
need tags for degrees, minutes and seconds). You might want to swing 
that past Tom E. as well though, in case he can think of any other 
variables.

---------------------

We may ask Tome Elliot's opinion as Leif suggests and consult other 
people in the future about this...but these are just notes for future 
releases.

Arianna

Syd Bauman wrote:
> LB> It seems a little misleading to describe this as
> LB> "underspecified".
> 
> Then why doesn't it tell me how to specify the latitude and
> longitude? I don't even know what units, let alone what notation, to
> use. (Afaik WGS 84 doesn't say.)
> 
> Currently the fact all of the following (and lots of other variants)
> are correct impairs interoperability in order to allow flexibility
> that, I bet, no one cares for.
> 
>    <!-- DM: -->
>    <geo>41&#xB0; 48.474' N, 71&#xB0; 15.562' W</geo>
>    <geo>41&#xB0;48.474'N 71&#xB0;15.562'W</geo>
>    <geo>41d48.474m N, 71d15.562m W</geo>
>    <geo>41&#xB0; 48.474', -71&#xB0; 15.562'</geo>
>    <geo>41&#xB0;48.474' -71&#xB0;15.562'</geo>
>    <!-- DMS or sexagesimal: -->
>    <geo>41&#xB0; 48' 28.44" N, 71&#xB0; 15' 33.72" W</geo>
>    <geo>41 48 28.44 N, 71 15 33.72  W</geo>
>    <geo>41&#xB0; 48' 28.44", -71&#xB0; 15' 33.72"</geo>
>    <geo>41 48 28.44, -71 15 33.72</geo>
>    <!-- DD (all with commas; could be without): -->
>    <geo>41.8079 N, 71.25937 W</geo>
>    <geo>41.8079&#xB0; N, 71.25937&#xB0; W</geo>
>    <geo>41.8079, -71.25937</geo>
>    <geo>41.8079&#xB0;, -71.25937&#xB0;</geo>
>    <!-- Radians (I think I have this right, but not sure): -->
>    <geo>0.72969 rad, -1.2437 rad</geo>
>    <geo>0.72969 rad N, 1.2437 rad W</geo>
> 
> (Anywhere I've used "'" above could also be U+2032; anywhere I've
> used '"' above could also be U+2033.)
> 
> Indeed, one could express the details of the format in prose inside
> the associated <geoDecl>, but it would be much more helpful if it
> were a controlled vocabulary.
> 
> 
>> It ought to say something like "within the section of the parent
>> declaring element" rather than "within a document" though, I
>> suppose, since geoDecl is declarable.
> 
> True enough.
> 
> _______________________________________________
> tei-council mailing list
> tei-council at lists.village.Virginia.EDU
> http://lists.village.Virginia.EDU/mailman/listinfo/tei-council

-- 
Dr Arianna Ciula
Research Associate
Centre for Computing in the Humanities
King's College London
Strand
London WC2R 2LS (UK)
Tel: +44 (0)20 78481945
http://staff.cch.kcl.ac.uk/~aciula/