David Carlisle writes: > What is the main purpose of a standard markup convention for journal > articles? > add * to allow sensible conversion to other systems > Some journals give full postal addresses for each author. > Some just give an `affiliation' for each author and highlight one > `corresponding author' for whom full address is given. ... > or some other markup scheme. The question is, does it make sense to > try to have one preprint class that covers both schemes. If such a > class is to guarantee that documents can be run without error on > either production class, then it seems that authors will be asked to > provide lots of `redundant' information such as full address and > affiliation for each author, even though a typical class will only use > one or the other. my immediate reaction is that the pain of doing this is relatively trivial, compared to the advantages. i would expect to supply the full affiliation and address for every author (allowing for 2 or more at the same address, of course), and to mark which is the corresponding author. the production class can throw away the extra addresses if it likes. (by the way, one should consider more structure in addresses, to allow for elision of common elements. consider 4 authors from different departments in a hospital - its nice not to repeat the whole hospital address each time. but i dont have a suggested markup) ... > is cross referenced back to the authors. Does it make sense to try > to capture all these systems with a standard markup scheme. Especially > as author order in some disciplines involves political implications of > `seniority' (In others, authors are always listed alphabetically > irrespective of seniority). > i wouldn't even consider dropping the concept of a single common markup scheme. without that, whats the point of the exercise? i would say that author sorting is up to the production class, god help it. obviously as a minimum you have to mark affiliation differently from postal address (and email); my only question is whether you have to break address down further (see above). so i'd be expecting to see something like \author*{...} % * means main author/corresponding author \affiliation{...} \address{...} \author[a]{...} \affiliation{...} \affiliation[Also affiliated to ]{alternate affiliation} \address{...} \address[Temporary address to June 1999]{alternate address} \author[b]{...} \affiliation{...} \address{...} \author{...} \affiliationlink{a} \addresslink{a} where the notation for re-using addresses is a bit Elsevier-like anothe thing to consider is structure within author. consider the paper by "Professor Hans Eysenck and Dr Sir Lord Dennis The Menace, DD"; it would common to have a running head of Eysenck and Menace which means that the markup should possibly be \author{Professor Hans \surname{Eysenck}} \author{Dr Sir Lord Denis the \surname{Menace}, DD} i'd rather see a specific isolation of known useful elements (ie surname) than any more structured setup. and Hans Eysenck is dead. better make that \author#{Professor Hans \surname{Eysenck}} or ... or ... but as has been said, notation is not so important as function. i would claim we have to distinguish: - author full name from summary form, ie surname - dead authors (commonly marked with a \dag) - corresponding authors - multiple affiliations (with annotation) - multiple address (with affiliation) - common addresses and affiliations sebastian