General goal: encode the sequence in which a text has been written down.
Texts usually allow to gather two forms of time-related informations: 1) exact dates e.g. 12.2.1893 or Febr. 1902 or 1709 or 1719-1722. 2) relational information Either in the form a) before or after a given date e.g. 'after 1718' or 'before 2.12.1934' or 'after Febr. 1730 and before 1734' b) before simultaneous with or after another source or entry e.g. 'after H1' or 'after H2'
These informations can be given explicitly or implicitly (based on inferences by the editor) and the information can be given fully or partially
This kind of information can be mixed, for example: 'after H1 and before 1734'. The lack of an explicit encoding implies a lack of information and must be handled accordingly in the sequence and timeline.
The encoding should allow to encode these informations as a sequence. In some cases it is not possible to determine the exact sequence because more than one sequence can be detailed based on the existing information. In some cases it won't be possible to attribute a source to a definite part of the sequence.
Should allow to encode two views: sequence and timeline (calender) Not all elements of the text can be attributed exactly to one sequence, some can't be attributed not at all. Only some some elements can be dated exactly. Sequences are not monolinear, but can branch etc. It should be possible to make implicit sequences explicit
Sentences become agrammatical by corrections; marking the level of correction where this becomes necessary
We have to allow editors to choose what elements they want to choose
Sequence / time information should be kept on an own level as standoff markup
Especially the dating of manuscripts or passages of texts is often hotly disputed. It should be possible to a) give the reasons for an editorial decision like attributing a certain date. b) document discussion and alternative arguments