>>>>> "Allan" == Allan Rae <[EMAIL PROTECTED]> writes:
Allan> I don't have a problem with the <before> and <after> sections Allan> only the part you labelled <natbib>. I can leave with that too. Allan> I have a problem with saving data that is easily regeneratable. Ditto. And also with saving data that does not exist (Caesar et al.) for a pure old numerical \cite thingie. Allan> JMarc may have concerns about the <before> and <after> handling Allan> but I don't. I see, you try to put words in my mouth to defend yourself... Allan> As I see it: We know natbib support is turned on/off. We know Allan> what \cite format the user wants. We know the key. We know the Allan> bibtex file to search. Therefore, we can reconstruct the label Allan> string just the same way it was generated originally. We don't Allan> need to save the part you labelled <natbib>. There is probably a problem of having access to the right data at the right time, though. Also, the fact that most of the code is in FormCitation creates some problems that look fishy (described in one of my earlier mails on the subject). JMarc
