[tei-council] date-stamping <desc>

[Syd Bauman]

> Using @version would be fine. But @lastUpdated buys you the same,
> and more, viz the distance of the relationship not just the
> cardinality. The word "lastUpdated" is not perfect, should be
> "lastSignificantlyUpdated", of course.

Yes, I suppose you're right. But again, it requires good rules on
when they get updated, and adherence to them, I should think.

  The following is a thought-experiment which lead me to the
  conclusion that it's not the date vs non-date that counts, but the
  semantics that are important, as I said in my previous post. So if
  you read and agree with my previous post on this subject, you can
  skip the rest of this one. Unless you want to be assaulted by my
  bad French :-)


It also buys you some potential confusion when a translator may not
have the most up-to-date source. [lsu is "last significant update".]

--------- 2006-11-02 ---------
<gloss xml:lang="en" lsu="2006-11-02">wife</gloss>

--------- 2006-11-04 ---------
<gloss xml:lang="en" lsu="2006-11-02">wife</gloss>
<gloss xml:lang="fr" lsu="2006-11-04">sa dame</gloss>

--------- 2006-11-06 ---------
<gloss xml:lang="en" lsu="2006-11-06">wife, girlfriend, or mistress</gloss>
<gloss xml:lang="fr" lsu="2006-11-04">sa dame</gloss>

But the French translator still has the 2006-11-04 version on her
desk when she updates it:

--------- 2006-11-08 ---------
<gloss xml:lang="en" lsu="2006-11-06">wife, girlfriend, or mistress</gloss>
<gloss xml:lang="fr" lsu="2006-11-08">sa bonne femme</gloss>

Which is no longer true. If that 2nd gloss had
"translationOfTheOneDated='2006-11-04'", all would be keen, and some
other French translator would notice the discrepancy and translate it
to "femme, copine, ou ma&icirc;tresse" or "&eacute;pouse, petite
amie, ou ma&icirc;tresse" or whatever. So it doesn't really matter
whether one uses version numbers or dates, this problem occurs
because we're just giving the version (date), not the version (date)
of the source of the translation.