Upgrade StorageIdentifying citation notes in word processor
To some users, particularly ex Endnote users, there is a significant difference between a citation that is added to an existing foot/end note, and a citation in a foot/end note that is created 'automatically' by the word processor plugin, when using a 'note' based citation style.
The distinction becomes important when changing from a note based citation style to an in-line citation style.
To me the distinction between them is if any other text has been added to the foot/end note. Already if the citation text ends with a full stop that full stop is removed for ease of adding other text.
There are currently 3 proposals for identifying a foot/end note that was created by the plugin to contain a citation:
1. Enclose the foot/end note mark in the main body of the document inside a Zotero 'field'.
2. Use a different 'field' name for the citation.
3. If the foot/end note has any other text then it stays a foot/end note.
Pros and cons for the above numbered options:
1. Ideologically sensible. What to do when user edits foot/end note - remove the identifying 'field'? But this is tricky to do without cutting and pasting and thus the user losing what is on the clipboard.
2. Using a different field name would be easy to implement, but what to do if the citation is copied into the main body of the document - simply rename the field?
3. Simplest but lacks the semantic difference of how the citation in the note was created.
The distinction becomes important when changing from a note based citation style to an in-line citation style.
To me the distinction between them is if any other text has been added to the foot/end note. Already if the citation text ends with a full stop that full stop is removed for ease of adding other text.
There are currently 3 proposals for identifying a foot/end note that was created by the plugin to contain a citation:
1. Enclose the foot/end note mark in the main body of the document inside a Zotero 'field'.
2. Use a different 'field' name for the citation.
3. If the foot/end note has any other text then it stays a foot/end note.
Pros and cons for the above numbered options:
1. Ideologically sensible. What to do when user edits foot/end note - remove the identifying 'field'? But this is tricky to do without cutting and pasting and thus the user losing what is on the clipboard.
2. Using a different field name would be easy to implement, but what to do if the citation is copied into the main body of the document - simply rename the field?
3. Simplest but lacks the semantic difference of how the citation in the note was created.
1. In any case, brief instruction in the differences between citation-in-a-note and citation-as-a-note has to be part of user instruction (induction) into Zotero's auto-cite capability. It's easy once you get it, but a person might not appreciate the importance of the distinction without help. Since that's the case (and if #1 seems the best option on other grounds), it doesn't seem too much of a stretch to incorporate strong advice on how to deal with changing your citation-as-note to a citation-in-note, or vice-versa.
2. Perhaps this would work as well. The other usability element which impacts the choice is that it might be nice to make it relatively painless to change between the two types. I particularly would be glad to have a minimal-work path to change citations-as-notes to citations-in-notes, if I decide to add a brief comment.
3. First I thought you could just say "if the citation has anything added at all, it needs to stay a note," but that doesn't exactly work since a citation preceeded by "Cf.", "See", "contra", or the like really *can* be put into parentheses --for inline citation types as well (contra BadAuthor, 2006).
So that's doesn't get us closer, but it does bring the following into play: however the distinction is applied by Zotero (and I'll be very happy when it is), it would be nice--for those writers who don't expect to change bibliography style, but who appreciate the idea of being able to-- if (1) the insertion of both kinds of citations were as easy as we can make it and (2) it were easy to change a citation from one 'type' to the other. Even if that means a crude Word/OO macro to delete the citation and replace it, I just will be glad if it's easy.
2 is possible at least now, though I still think 1 is the forward-looking solution. You ask: I think we need to step back and examine the issue scot raised, because it's an important one WRT to this issue. You tend to be assuming text is just text, but it's not that straightforward.
The question is how to model this kind of parenthical captioning in general? It's also used in other styles (author-date, author, and I guess in-text).
There are three options there:
1. it's just text, like any other text
2. each reference AND each citation supports explicit prefix and suffix captions
3. perhaps 2, as well as supporting different classes of reference (see, see also, contra, etc.)
1 and 2 are the most flexible, but also the most problematic for automation, because citation styles typically define rules for sorting within citations; rules which a user shouldn't have to worry about. If they do add just text, what happens to sorting?
3 is therefore preferable, though there'd need to be more work on the modeling and making the GUI simple/elegant.
So to answer your question above Ian, I think we need to distinguish different kinds of text (or textual rendering of different kinds of references) a user might want to add. If scot wants to do as above, it should stay as a footnoted-citation.
If he really wants to change it to a footnote with commentary, he should be able to do that. It probably suggests a contextual menu command or such (automatation?) where a user can convert between the two forms.
BTW, if we agree on the best way to do all this, I'd like to code up some examples in the new OpenDocument metadata support and document it as suggested best practice. There this stuff is planned to be encoded in RDF.