Collaborating With Scrivener

Are there any plans for making Scrivener a truly collaborative platform? I reviewed the little bit you’ve posted about the upcoming Version 3 of Scrivener, and I was disappointed to not see anything about making the software collaborative.

I purchased Scrivener nearly three years ago because I believed all the hype, and while I’ve made five heartfelt runs at implementing the software into my writing routine, I have yet to make it work for one simple reason - I can’t collaborate with my editors. I WANT to take advantage of all the amazing features of this software, but the world of writing has changed so much since you put this together. More and more, writing has become a collaborative affair for many authors, with co-writing becoming a major trend in the indie author industry.

We really do need a file structure that allows for multiple users at the same time, in different locations around the world, to be able to edit projects. For example, I have written four novels in Google Docs. I have NONE of the wonderful features and functions of Scrivener at my disposal, but what I do have is the ability to share my documents with other users (simply by entering and email and sending them a message). They get a link and can edit the same document AT THE SAME TIME I am.

The added benefit is that the work is ALL in the cloud, so I can do all my work from a browser. I don’t need a MAC or WINDOWS machine. There is only one version of the software to support. Sadly, until there is a way to legitimately collaborate with Scrivener, I can’t see using it. My writing would suffer too much with my support team.

No, there are no plans to either make Scrivener a web application or to support real time collaboration.

Both of these are extremely challenging technical problems, requiring more resources than a very small company like L&L can command. There are more people on the Google Docs on-call support team than L&L has employees.


I have the same concern, and it has driven me to explore all sorts of other options. I’m still doing that now.

I get the huge technical challenge that is represented by facilitating online collaboration, but frankly - I also think you’re going to find that this becomes an ever more significant obstacle for people considering your software. There are already a lot of competitors in this space. Their feature sets are nothing like yours, but they’re just getting started. I’m actually just starting to beta test one advanced addon for google docs that is designed to facilitate the creation of full books, called “write with claire”.

For my own purposes, I would be satisfied with a short-term compromise that would enable me to just compile to a format that is easily ingestible into something like google docs - perhaps the google docs file format itself, if that’s possible (to avoid translation errors), get comments, make changes in google docs, export out to a new file and then UPDATE my scrivener project from that exported file, with comments intact.

Thoughts anyone?

Yes, I think that if you need specific features which you find in another software but not in Scrivener, then use that other software.

Or, your collaborator could learn to use Scrivener. That way you could actually collaborate on the same project, stored on Dropbox. Just make sure that you don’t open the same project simultaneously.

There are some additional features coming in Scrivener 3 that will help facilitate serial collaboration.

Real-time collaboration, on the other hand, is a very heavy technical and infrastructure lift. Among other things, it requires that the documents in question be stored in a central “cloud” location, with all the complexity that involves. There’s a reason why an enormous company like Google is the leader in this space.


That sounds like good advice until you realize that other program also lacks features that you need. I’m trying to avoid a tortuous workflow that is strung out between multiple programs.

At the moment, I’m not “collaborating” with someone, but I am soliciting detailed feedback from various people. Asking each of them to buy Scrivener is a barrier to getting feedback. What I’ve been doing up to now is exporting to .doc and uploading to google docs. This, however, has been an agonizing process due to limitations in the export process that make it an ordeal to get a nicely formatted .doc output (scrivener adds unnecessary footnote numbers for example, so you end up with multiple, un-synced, numbers next to your footnote links. Also, its lack of proper styles is an issue), and then there’s the annoying issue of syncing your changes to and from the google doc. After trying A LOT of different options (lots of different export options and formats, post editing in lots of different programs, considering making my own markup style approach (before I learned that was actually a thing)… it’s been incredibly frustrating) I finally discovered multimarkdown and pandoc, which will make things MUCH easier (not only for feedback, but blogging etc - basically getting a clean output that can be easily converted to other formats without all the mess, of say - inline formatting in an html file - aarrrgghhh!!). There is, however, still the data syncing issue for getting feedback, and it would be really nice to be able to work on my project from anywhere (such as my phone).

Okay… If you look at it from the sideline, from a strictly logical perspective, it is just as reasonable to ask Google to adapt Google docs to be able to read and write Scrivener projects, right? Most requests in here assume that Word or Google Docs is some kind of standard that all should adhere to. But if so, why not use Word or Google docs?

So instead you are asking the programmers of all those different softwares to adapt their apps to your needs?

Personally I avoid a tortuous workflow by accepting the limitations of different softwares and trying not to develop any advanced workflows. Each task is kept within the software that is best suited for that task. :slight_smile:

Alternatively, if you find that no software does what you want, you could do what Keith did and write it yourself! Legend has it that is how Scrivener came to be. Who knows, you might just launch an unexpected career for yourself.


p.s. I am sure there is a workflow for collaboratively working on individual scriv docs. The external sync function can externalize the text files for manipulation elsewhere, such as uploading to Goolgle docs and then bringing back down to text doc. Running sync files then reintegrates the changed text files into your project. It is not your dream of dreams, but it gets you both working in scriv at the project level and collaborating at the doc level. Disclaimer: I don’t do collaborative writing like that, but I bet someone out there in the Scrivosphere is doing it. (I do know a pair of writers who mean to try working out something like this – will report if something interesting happens there.)

For an idea in Scriv 2, see @nontroppo’s post here:

Another alternative could be Fade In. While it is primarily a screenplay writing application, it also has a template for a Novel Manuscript.
Fade In features real time multiuser collaboration.
It’s Elements (= paragraph styles) can be altered to the user’s liking and/or new ones can created. It also features a Index Card View akin to Scrivener’s Corkboard. Feature-wise, Fade In is somewhere between Ulysses and Scrivener.

All the best.

As a long time user of Scrivener for novel writing, I can concur with folks looking for a collaborative feature. Having sent novel drafts away as word docs, I have to go back through the resulting edits line by line, searching for where I’ve misspelt “you’re” as “your” in the middle of a five thousand word chapter is no fun. If I could send my editor a link to an online edit space, or have them download a collaboration only copy that can only see my work, but then the edits are sync’d back to my original, it would save a lot of time and hassle.

As an IT professional, I can fully understand Katherine’s point about the difficulty of implementing this. One of the greatest features of Scrivener is it’s reliability. I will not tolerate any software that loses something I’ve written, Scrivener never has, even after the introduction of the iOS sync feature.

So mixed feelings. Would love to see it, but not at the cost of losing the integrity of the core program.

I posted the original “wish” here a few weeks ago. Just read all the replies. Here’s my take away:

  1. Saying it’s just too hard is a cop out - clearly customers want collaboration, they NEED it, and co-writing is a thing of the future. Telling people to get over themselves is not helpful. It’s a legitimate request.

  2. Clearly people are willing to play outside the Scrivener box if necessary. Make that easier is all we are asking. For example, if I need to export a chapter or full document to GoogleDocs for collaboration, fine. But make that SIMPLE, and make it SIMPLE to get the document back into Scrivener as well.

  3. Cloud based file management is not new, and DropBox is not the only game in town. Yet the Scrivener ecosystem appears to recognize only DropBox as a viable option. At the very least, it should not be hard to enable Scrivener to open a full Scrivener project file from a cloud based location - after all, it’s just another storage drive. Online storage is the future.

I don’t hear anybody directly involved with Scrivener development telling people to get over themselves. But - although I know next-to-nothing about software development, my understanding is that to plug simultaneous collaboration into Scrivener would indeed be very hard. (Sequential collaboration, with each user taking turns to work on a project in a central location is a different matter and very possible now - see below.) The problems with coding for simultaneous collaboration in Scrivener are at least twofold: first, people to do the work - within L&L Scrivener for the Mac is coded by one person, and Scrivener for Windows by two, whilst outfits such as Microsoft and Google which do offer simultaneous collaboration apparently have more people in their PR departments than L&L has employees. Imagine how many coders they must have! Second, each Scrivener package is made up of a folder containing hundreds, perhaps thousands of files (unlike, as I understand it, Google Docs, say). This protects long Scrivener projects from being totally destroyed by the corruption of one part*, amongst other advantages, but even a layman like me can see that getting simultaneous access and editing by two or more users to work with such a file structure would be, well, a very hard thing.

Although I’ve never used Scrivener’s feature that enables syncing of a document with an external folder, what you have described is my understanding of exactly what that feature will do. (See Section 13.2 in the Scrivener Manual, if I remember correctly.) I see that you work on Windows - if that particular feature hasn’t yet arrived on Scrivener for Windows, it will do before too long.

Dropbox isn’t the only viable option; other cloud options are also viable. It’s much more that certain cloud options aren’t designed for Scrivener’s multi-file project structure and will pretty definitely screw it up for you, and therefore are wise to avoid - the advisories in the FAQ here are well worth reading, if you haven’t. As far as Dropbox is concerned, I keep all my Scrivener projects there - at no cost to me - and open at least one of them from that cloud location on most days. Dropbox has never to date let me down. And I could work collaboratively with a writing partner on the same projects (see above) in the same cloud location, as long as we worked sequentially.

  • I value Scrivener’s multi-file folder structure particularly - I once irrecoverably lost 40,000 words in MS Word.

The reasons for this have been discussed in detail elsewhere. Briefly, (1) other cloud services do not support the level of access to the file structure that is needed for Scrivener projects, and (2) what happens to your cloud-stored project if you lose access to the internet with 4,000 unsaved words in your computer’s memory? Personally, I have no interest in a writing program that requires an internet connection to use it.

If you find that Google Docs is a better fit for your needs, feel free to use it. I wonder, though, why no one is pointing out that, after years of development with a much larger team, Google Docs has nothing that even resembles Scrivener’s organizational and editing tools for large documents. They are different tools with different goals.


What you’re asking for is the equivalent of making a Lamborghini into a helicopter.

A Lamborghini is designed for a specific purpose. Specific requirements (“can do X miles an hour”, “can handle forces of X magnitude when cornering”, “does not need storage space”, “does not need space for more than driver and passenger”, etc.) are determined, assumptions made, and the entire design hinges around those requirements and assumptions.

Now, sure, you can go all Mythbusters on that Lamborghini and figure out a series of modifications that could be applied to the car that would allow it, technically, to be a helicopter. You could brute-force it and get something that manages to get off the ground. However, it’s not going to be a GOOD helicopter. It won’t do the kinds of things that you can do with a real helicopter. It won’t have the range, the safety features, the cargo capacity, the passenger capacity, and all the other features and requirements a real helicopter is designed to meet. It’s a one-off hack that is not usable on a regular basis.

It’s probably going to suck as a high-end race car, too, because you had to compromise all of the original design requirements and assumptions to get it to the Frankencopter state.

All because you tried to make something perform a task outside of its design requirements.

It’s the same thing with software – good software design includes requirements and assumptions, whether there is a formal process to identify them or not. KB had a very specific set of goals in mind (still seems to!) for what he wanted Scrivener to be. The design goals and requirements he made determined the technology choices he made (which, BTW, is one of the reasons why the Windows port has always been farther behind, as it doesn’t have key pieces of the technology the Mac version does because the Windows OS doesn’t offer them the way Mac OS X does). The vision of “multiple little documents compiled into final output document” has HUGE advantages for the process of writing, but it massively complicates (I would go so far as to say “makes impossible at this time”) things like real-time multi-user collaboration because the requirements just don’t add up.

For the record, one-way sync can be a more difficult problem to solve than you think, even for single files. Every sync solution I’ve used has issues, at first, as the dev teams figure out what kinds of errors and oversights are common and learn how to solve them. Very few sync engines can handle multiple files that are all related, which is how ANY sync solution sees the Scrivener package format. Of the key players in this space, only DropBox both handles that complexity safely AND gives the necessary level of access to its routines to permit app authors to wire it up correctly with their iOS and Android apps to handle that package format. Writing mobile apps is a whole new set of assumptions, requirements, and constraints.

Give up that package format, and a lot of the features that thousands of current Scrivener users rely on TODAY suddenly go away. Plus, you essentially throw out all of Scrivener’s code and start from scratch.

As Katherine noted, multi-user realtime collaboration solutions like GoogleDocs or Word Online don’t have the full range of organizational features that a tool like Scrivener does. They are designed to do one thing – permit collaboration on a SIMPLE document that is a subset of functionality of a full desktop word processor. They can’t even handle all versions of their own format – take a complicated traditional Word or RTF document through either solution and watch it get mangled!

If LamboCopters were easy, they’d already be a thing. In software, as in the physical world, design is a series of compromises. A lot of times the features you leave out are more important than the features you include.

Moreover the initial commitment to a cross-platform framework and Linux releases impaired the deployment of Windows DLLs and ActiveX components in SFW’s development. So the team couldn’t really make the most of Windows’ capabilities.

I haven’t always agreed with you, but I LOVE this. I’m telling you it will be shamelessly stolen out of sheer lack of any better explanation of exactly why requirement creep is the spawn of all evil in the universe…

LamboCopter… Whuummpaaummmpaaummmpa


You can’t steal what is freely offered, my friend. :slight_smile:

“Let me introduce, the Lambocopter! The Swiss army knife equivalent when it comes to vehicles!”

… which means it isn’t really good for anything, which Devinganger explained in a fantastic way. Specialized tools always out-perform multi-purpose tools when it comes to really quakified work. Ask any carpenter…
Including more and more specialized tools in a single software is like buying a carpenters belt filled with tools. It’s likely to hold some of the tools you need but not all, and definitely also a lot of stuff you don’t need.

Lambocopters - I want one!