Given the diversity of different kinds of reviews, it might make more sense to treat them as relations. Article --> review of --> Book/Play/Film.

I think in general the Zotero crews needs a forward-looking strategy on this, rather than to just constantly add new types. The current list of types is already starting to get a little incoherent.

@TimB -- it's a nice site, but I actually think their categorization isn't appropriate, at least not when you get into the details of their subtypes. I think it's a mistake, for example, to mix media (electronic vs. print) with type with distribution mechanism or format (online vs. physical). That you sometimes see this now is largely a function of policies set during the initial birth of the internet, but which are gradually changing. In a decade, it's likely that almost everything we cite will be available online, with some very rare exceptions.

Ultimately a type policy needs to be driven by a set of requirements; what do you want to achieve? There may be different -- and sometimes conflicting -- requirements here: citation formatting precision, finding records, GUI configuration, etc.

In general, despite common wisdom, citation formatting is actually not terribly dependent on type, but rather on other details.

Ah yes, CVs; yup, that's a reasonable requirement too :-)

The section on citing reviews doesn't really treat book reviews any differently than other types of reviews though.

On this:

"If Zotero is to honestly claim that it outputs in e.g. Chicago style then it meeds to include the main data types that Chicago handles..."

The problem is what constitutes a "main type" is in the eye of the beholder. When I've looked at this previously, I counted somewhere between 50 and 75. And that's just for Chicago. It doesn't cover other styles. So do we really want to end up with, say, 100 types in Zotero, and bloated style files that are more difficult to write and debug?

I'm the author of CSL and the Chicago style that Zotero uses. It's my contention that it can correctly format the vast majority of citations with a small handful of templates. This is because citation formatting is, despite common wisdom, NOT fundamentally about type. There are patterns and conventions that are more important.

I started to try to show this here:

http://www.users.muohio.edu/darcusb/misc/citations-spec.html

Let me be clear: I'm not saying there shouldn't be a book review type in the GUI; just that there needs to be a policy for new types, and that it needs to be integrated into the larger data strategy (with the RDF and such).

PS, I am not intending to be stroppy, but am really keen that we can adopt Zotero and I can recommend it to my students. At present it is really good, brilliant even, but lacks a few "small" things that I need. The ability to output formatted citations of Book Review articles is one, and the ability to import from our library system is another, but since it is not a very widely used system I expect that we would find a way to do that perhaps in collaboration with the supplier.

The easiest way to add book reviews is to add "reviewed title" field and "reviewed author"/"reviewed editor" author types to all article item types--this allows for correct bibliographic formatting for journal/magazine/newspaper books reviews.

I have pretty much no technical help to offer but just some hearty encouragement. A review tag would be great; my own field is classical studies and book reviews are a common bibliographical feature.

Zotero looks fantastic, btw. I plan to recommend it to my students!

I still thank you for a great tool, but still believe that for scholars in fields where published reviews of book length works are sometimes cited, and often used in student essays the lack of a "book review" type is a pity, if you don't know how these should be organised Duke university library has a good summary page at http://library.duke.edu/research/citing/workscited/bookreview.html

I strongly agree with Phil. I'm have several book reviews to cite in my PhD dissertation I'm working on, and it'd be great if Zotero added a book review option.

I usually use a generic document or webpage type to encode the transcript information.

I don't think the zotero model and ui is currently flexible enough to handle this correctly in terms of being able separately represent and link, say, the speech event, and the video and text transcript. But I expect that will change.

I'd like to voice my support for a "Review" item type as well - citing reviews of books is a major part of my discipline, as many of our journals have sections devoted to reviewing the scholarship of the field. These reviews come into broad use in further research and publications.

Having a field for "Reviewed Book" (to go along with ""Reviewed Author") would be vital.

As an example, the review by Joseph Janangelo, itself titled "Rethinking Style and Reversing Hierarchies," is on Richard Lanham's book The Economics of Attention. The review was published in the journal College English. Under MLA style, it would populate in the bibilography as:

Janangelo, Joseph. "Rethinking Style and Reversing Hierarchies." Rev. of The Economics of Attention: Style and Substance in the Age of Information, by Richard A. Lanham. College English 69.5 (May 2007): 49-56.

I'd like a "Book Review" item type as well, and/or a "Reviewed Title" field.

Technical note: in a possible Zotero implementation, this would need to distinguish the review author from the reviewed author, hence another label distinction for the creator field.

no, Zotero is non-customizable by design.
But what exactly do you need, maybe there is some other way to do this - i.e. give us a sample citation for a computer game.

that looks pretty standard, no?
What if you just treat this as a book?
Item types are really mainly to support different ways of citing.
For sorting purposes you can still use tags.

ok, fair enough.
But then the problem seems to me not that you need more customizability (which is often synonymous for "hacks that don't travel between users"), but
a) that the computer program item type can be called using if type= in the csl (which it cannot atm) and
b) that the "system" variable in that item type is mapped to csl (which it isn't atm).

I'd encourage you to open a new thread with that as a request referencing this discussion as well as your citation style requirements. You've kind of hijacked the "book review" discussion with this, which isn't a good idea if we want to keep this forum ordered.