[tei-council] proof of concept of i18n work

[Lou Burnard]

I think it would be more sensible to trawl through the valDescs for any 
that are actually useful before doing any work on dating them.

I strongly suspect the number is very small indeed -- most of them are a 
hangover from the days before we had datatyping



On 17/11/13 20:27, Sebastian Rahtz wrote:
> On 17 Nov 2013, at 19:08, Kevin Hawkins <kevin.s.hawkins at ultraslavonic.info> wrote:
>> a) Add @versionDate to each <valDesc> (in addition to <desc>).  That way
>> we'll be consistent in using @versionDate on all parts of our specs.
>>
>> b) Merge <valDesc>s into <desc> and just do away with them.
>>
>> My preference is (a).
> the reason I suggested throwing them away is things like
> 	<valDesc>any string of characters</valDesc>
> or
> 	<valDesc>a name of module</valDesc>
> which just annoy me. There is certainly useful material
> in there, but it needs a trawl.
>
> but yes, adding in the @versionDate for the English
> (and indeed for the French too, which are undated) would
> be useful.
>
> Running Utilities/addDateToGlossDesc.sh in the P5 source tree does the
> job (though for sanities sake on might run it only on files which actually have
> valDesc; that script takes a parameter ($1) which is the list of XML files
> to operate on).
>
> Note that the script generates a series of shell commands,
> so usage (for just one file) looks like this:
>
> 	sh Utilities/addDateToGlossDesc.sh classCode.xml | sh
>
> Someone may like to amuse themselves by running this for real,
> for English and French, and checking the results do what is expected.
> --
> Sebastian Rahtz
> Director (Research) of Academic IT
> University of Oxford IT Services
> 13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
>