Warning for Scrivener+Zotero (2.1.x) users

Hello everyone,

I made the mistake of upgrading my Zotero to 2.1.1. Fortunately, I was able to downgrade back to 2.0.9 (since I made backups) but I am posting this as a warning to other Scrivener users who use Zotero for footnotes using the RTF Scan function (I think that there are many of us in the academic world). Beware of Zotero 2.1.1: the RTF Scan function does not work properly!

Here is what I posted in the Zotero forum last night:

Hopefully this message will save some people from some serious troubles.

Thanks for posting your findings. I have changed the subject name to include the Zotero version that gave you problems, so that future searches for Zotero and Scrivener are clear that it isn’t a general warning.

Fair enough. I just hope that your addition of 2.1.1 to the title does not dissuade users of other versions of Zotero from reading about the problem. It is important that people using RTF Scan do not upgrade either their Zotero or their Firefox, since Firefox 4.0 does not work with Zotero 2.0.9, which, as far as I know, is the last version to have working RTF Scan capability.

At the Zotero forum, some have suggested that the Zotero development team has little interest in fixing the RTF Scan capability. Is there any chance, then, that the Literature and Latte team could write a Zotero plugin that would link Scrivener to Zotero and thus provide Zotero citation functionality that would be equivalent to that of the Zotero plugin for MS Word or Open/LibreOffice?

Thank you very much.

It’s a good suggestion, but Scrivener would probably first need AppleScript support, which I don’t think it has. I guess support for fields and/or citation objects. Doesn’t have that either, I think, so it would require a major change in the software.

In any case, if you search these fora, you will probably find this suggestion at least a few times, and I would guess it hasn’t gotten much in the way of a ‘yes! we’ll do it!’–but to be fair, if I remember correctly–too lazy to look–these requests do at least get a response from developers here, whereas they are totally ignored at Mellel and Nisus–the one wp for which I suspect it would be more or less plausible to write a plugin. In other words, in the foreseeable future, you’re probably dealing with the imperfections of RTF scan or trying another citation manager. There is a lot of information about that in the forums.

To be fair to Keith, he has always said that AppleScript is something he would like to implement; to that end he opened a thread on what users would like to see. If I remember rightly, his last words on it were that he hoped AppleScript was something he would include in version 2.5. But of course, as a sole programmer, it is an issue of how much he can get done at any time, and prioritising development needs as he sees them in relation to feedback.


Fair enough. Should I post this feature request in another forum, or do you think that the Scrivener team would notice it here?

Leave it here; they’ll see it. If they think it would be better in another forum, they’ll move it and leave a “Thread moved” notice.


I don’t really see how a Zotero plugin would benefit, seeing as you would need to scan the exported RTF document anyway. And yes, AppleScript is on the list, hopefully for 2.1 later this year.
All the best,

For cross platform purpose a scripting api for python would be cool.

If I understand correctly Keith is pointing out the way in which any document produced, given the more fragmentary writing environment that is one of the strengths of Scrivener, would need to be recompiled after export. That makes a lot of sense, of course, and is another objection that occurred to me. But what if it were possible to script the citation recompile on Scrivener’s end as part of the export? That would make the day of not a few dull academics.

Agreed, this is how I envision it working. Either that, or it works something like how it does with the Word plugin.

I have no idea how this would work, sorry. If any experts in Zotero know how another program could automatically call on Zotero and have it can an RTF file that has just been produced, I’d certainly be willingly to consider it, though.

I think soon enough this will be easy to do with Zotero (and easier to understand how to do it), so maybe sometime soon down the line it would be good to revisit this. Zotero has only a couple of paid developers, so I doubt they would have the time or the inclination to take this on, but there are many Zotero users enthusiastic about Scrivener (at least if Profhacker and the like is any indication), so hopefully somewhere in the community there is someone with the know-how who might be able to collaborate with you.

It’s a tough problem. I like Bookends a lot, but doing research in Zotero is much easier and more intuitive, but, while its citation formats are great, the plugin only interacts with two word processors that I don’t much like using (Word and the now three different Opens) in the face of such good alternatives (Nisus and Scrivener). But I hesitate to maintain a separate database in Bookends just to interface with Nisus and Scrivener, since I can’t see it saving any time over just doing some of the possible work-arounds I’ve come up with: manual in-text citations and export from Zotero a works-cited listed, or using the still imperfect rtf scan for articles requiring Chicago notes and taking a half hour to go back and correct the finished product (which one must do anyway, of course). And of course if I’m dealing with this three or four times a year, then that’s a pretty prolific career!