Annotations | Round trip Scr. -> Word -> Scr. fails

Here’s the desired scenario:

I work in Scrivener. When the time comes I export to an RTF file and my editor annotates it in Word. (Yes, yes, I know. Publishers. What can I tell you?) He sends the Word file back, I bring it into Scrivener and carry on.

Here’s the reality:

My Scr. annotations go over to Word immaculately. But both my and his annotations do NOT make the round-trip back from Word, whether I use RTF import, RTF drag-and-drop, .doc import or anything else.

In RTF, they just vanish. If I import the file as a .doc file, his annotations are gathered up at the bottom of the file, e.g.:

PAGE # “‘Page: ‘#’?’” One problem with starting off with the experience of loss as a leitmotif is that a large part of your readership … [blah rant yadda]

… which doesn’t allow me to see them in place.

This is a big workflow PROBLEM. Does anyone have a SOLUTION? Please say yes…

First, I will wait to hear from Keith to see if he has anything in the works for this. If not, I can maybe work something up that will convert RTF comments to Scrivener annotations. Are you at all comfortable with running a script from the terminal?

I agree that this is problem but I am afraid that I do not have a solution at present - currently, annotations (and footnotes, and images in non-RTFD files) are one-way only.

Honestly, please, please post an enhancement request to the OS X text system here:

Scrivener currently hacks the OS X RTF exporter to add these features. Doing the same for the importer is way, way less trivial (er, more complicated), because of way RTF syntax works. Essentially, because of the fact that you can vary the order of where certain control characters appear, it is very difficult to know where the annotation/footnote text actually begins. This is why, to date, it is only dedicated word processors that support this (notably, even Nisus and Mellel don’t support RTF annotations just yet…).

This is something I will review in the future, but I want to wait to see if Leopard introduces RTF footnotes etc “for free” - there is no point me spending weeks trying to code this right now if in a couple of months it turns out that you can do it anyway. If this doesn’t come with Leopard, then I will definitely add it to the list of things to address around 1.5-2.0.

All the best,

Ah, well, then it is probably more than I could handle in an afternoon, then! Sometimes RTF looks like it was designed by a committee, and then I remember that Microsoft publishes the specifications – so, it probably literally was designed by many committees!

Here is an example of an RTF annotation in RTF syntax:

An example of a paradigm might be Newtonian physics or Darwinian biology.{\v\fs16 {\atnid bz}\chatn{\*\annotation \pard\plain \s224 \fs20 {\field{\fldinst page \\#'"Page: '#'\line'"}{\fldrslt}}{\fs16 \chatn } How about some examples that deal with social science? That's what this paper is about.}}

As you can see, it would be very difficult to parse through this to see where the text of the annotation actually begins. Dedicated word processors handle all of their own RTF parsing (or at least, I assume they do - I don’t know how Mellel and Nisus do it, to be honest, but I doubt they just hack the default importers), so they will interpret all of the other syntax (“pard”, “plain” etc) and use it, and in so doing - that is, in being aware of all the syntax - they know which part is actually the annotation (or, if they don’t support annotations, they can just skip everything contained in the curly brackets that begin with the “annotation” control word). It would be very, very difficult to piggyback the Apple RTF importers to find the annotation without the hack being aware of all this syntax, too - and given the large amount of syntax, that would be tantamount to writing your own RTF importer!

Piggybacking the exporter is much easier just because Scrivener only needs to use a limited amount of syntax. It places markers in the text where annotations should go, uses the Apple exporters to create an RTF stream, and then inserts the annotations, surrounded by the appropriate syntax, directly into that RTF stream before saving.


That’s pretty nasty looking.

I agree, though, that this feature would be terrific to have for those of us who work with editors who are locked into Word. And even for those who are not, some sort of back-and-forth annotation capability is a must for serious collaboration. I’m finishing a draft of a book this month and am already wondering how I’m going to work with my editor effectively. (He uses Word on Windows or else I’d try to persuade him to use Scrivener!)

I don’t suppose the initial attempts to write a Perl RTF parser at … arser.html

would be of any use?


Good luck persuading an editor to work outside his company-mandated work environment, and one with which he or she has become cybernetically familiar via involuntarily constant use. Remember: you’re one author with one book, and it’s very likely that your editor is dealing, at any given time, with 20 authors and 20 different books. If all of them demand collaboration in a different program, what a tower of Babel we should have.

Authors adapt to editorial needs; not the other way around. And Word, as irritating as it may be, is a near-universal language in publishing–in large part because of its collaborative abilities via revision marks.

So write your book in Scrivener, but if you expect to collaborate with your editor on-screen, versus on-paper with a forest-full of Post-It notes, expect to bite the bullet and buy a copy of Word.

@Ahab - oh, indeed. I’m an editor myself and Word is the lingua franca of our field. (Or something stronger . . . more like Latin in the church in the Middle Ages . . . one is inconceivable without the other.)

But I am so truly, truly glad to be set free from Word for my own writing. It hurts every time I have to open it up again, which is, unfortunately, every day. (I do think that Word on Windows is less annoying to its users than Word on the Mac, though then again, they use Windows to begin with. :slight_smile: )

Anyway, the ability to get annotations back into Scrivener would be marvelous, even if no editors ever agree to use it.


Like I say, I agree it would be marvellous. I even spent a couple of hours last night looking into it, experimenting, but it really is a minefield and my experiments were not exactly successful, so I doubt this will happen for some time, I am afraid.

Hi Keith,

I hope you manage to figure something out concerning this workflow issue of importing rtf-files back into scrivener. I would make Scrivener, an already fantastic app, remarkable since it could seamlessly connect with the world of editors and reviewers using ms word.

For me, this issue could be solved if Scrivener printed the text + the footnotes (when selecting file->print), and left out the annotations. In that way, I could save the file as a pdf and send it to my editor (I am aware it would be very “draft-like”, but that’s sometimes the point). Now, when printing from Scrivener it outputs the text + the annotations

So, if there is any quick workaround of changing the way Scrivener handles the printing of drafts, it would really make my day.

Thanks, and keep up the good work.

  • Magnus

Magnus -

Good news: you can already choose to remove the annotations from printing! If you are printing a single draft, go to File > Page Setup and then select “Scrivener” from the Settings pop-up. You will be presented with various options, one of which is a checkbox that allows you to “Remove Annotations”. For printing the draft, you do this via the Export Draft sheet - again, just choose to remove the annotations under the Text Options tab.



I should have checked the Help viewer in Scrivener, since the fix is also documented there (I just checked). But, I really appreciate your help. Next time I’ll consult the documentation.