office-collab — archive

[Robin LaFontaine]

In general the independence of the changes is established by the actions of the user, i.e. whether he/she made the changes as a single action or several actions perhaps with other actions in between. Such independence is recorded in the tracked changes because each change is a separate change (CT). The example was intended to illustrate the problem in the simplest possible way, but the problem itself is more widespread. UC6 is a larger example - a text frame is moved, text is edited, then it is resized , then moved again. The changes are independent, from the user's perspective and are recorded as such because they are different CTs. So I believe this remains true: for ECT any two changes that affect one 'bucket' can only be rejected correctly in the reverse order, i.e. if the last change is rejected first. Of course a user does not know anything about buckets so this may lead to confusing behaviour. Robin On 10/09/2011 16:07, Dennis E. Hamilton wrote: 005601cc6fcb$6e471c20$4ad55460$@acm.org type= cite > But of course it must be known that changed attributes are independent in a given situation, yes? In some cases, the interdependence is not known from the schema alone, and so some intelligence is required. (It is potentially even more fun if foreign-attributes or attributes with foreign-values are present in the document produced with change tracking.) I am not sure if there is any guidance in the current implementations or not. Experiments with style changes would have to be made to see if it is even apparent to users that there were multiple, separate style changes in exactly the same span. - Dennis