Should I place my main scrivener project files in Dropbox?


I’ve currently got Scrivener setup to automatically backup when I open or close the project, or when I hit Cmd + S to save, but I was wondering if it’d be a good idea to also place the main files into Dropbox just for the extra security?

The only concern is that including pictures, my Scrivener project is over 200mb, so I’m worried that could cause problems if it’s constantly trying to upload that to Dropbox?


Yes, having a backup other than what Scrivener does automatically is good. No, Dropbox is not necessarily the best solution. I’ve discussed backups in a bit of detail here:

Generally speaking, Dropbox (or any other backup software) will only need to upload the files that have changed. After the initial upload, a 200 MB project shouldn’t be particularly onerous.


if your project is 200Mb, it’s worth asking why – is there something you can do to compact it?

For example, lots of web archives contain huge chunks of spurious graphics – ads sourced in via ad exchanges and cdns, random clickbaity images, and so on. Copying the text from a web page and saving that instead of a full web page can often squish down 1-2Mb into 2-6Kb (text is incredibly compact, images … aren’t). Again: PDFs with images embedded tend to be huge.

(For reference, I’ve got one Scrivener project that runs to about 66Mb. It contains an entire nine book series of novels, over a million words of prose, cross-referenced and tagged, and supplementary research material. It’s gotten a bit bloated over the years! But because it’s primarily prose it’s still a lot more compact than the monster you’re talking about.)

Second point: it’s worth moving your Scrivener backup directory onto Dropbox (Preferences -> Backup). If you do this, give serious thoughts to switching on compression (to save space), and also to putting a numerical limit on the number of backups to be kept; if you have a Time Machine volume this’ll keep earlier backups until it runs out of space (as long as you haven’t excluded your Dropbox folder from Time Machine backup). You may want to tweak the backup frequency so that it only backs up when you tell it to, not every time you save the project or quit Scrivener.

Thirdly: if you intend to use Scrivener on iOS, you will need to sync your projects between the desktop and the iPhone or iPad using either iTunes or Dropbox. And guess which is less painful? Especially if you switch on the dropbox option “Enable LAN Sync” (under Dropbox->Options->Network) – it will then sync over your local wifi network rather than spewing everything out across a broadband connection.

The biggest argument against using Dropbox is simply if you have terrible broadband internet access: a 200Mb project is going to hog your line if you can only get 1mbps of throughput, but is trivial if you have an 80mbps fibre broadband link, for example.

Note that we do not recommend keeping both backups and the live project in Dropbox. Dropbox synchronization errors are common enough that having a local backup that Dropbox can’t touch can save a lot of pain and suffering down the line.

Also note that a backup is, by definition, a complete copy of the project, while the normal save operation only saves the things that have changed. So the backup – to whatever location – will take more time than simply saving an incremental update to the project.


Lots of my scrivener projects are well over 200mb (PDFs, illustrator files for figures etc), and I’ve never had problems (even now that I have to go through a VPN and very unreliable internet in China)…

The other big advantage, although I love Snapshots I often forget to snapshot, make an epic change i regret and don’t have a rewind mechanism. BUT using dropbox, it is easy to use their versions feature to call up the lost text. I wouldn’t use Scrivener any other way.

I’d love to have a time machine type automatic snapshot system in Scrivener: every hour for a day, every day for a week, every week for a month, every month for a year, this would limit the total number of snapshots and remove operator error

Or some sort of user-defined triggers on when to take a snapshot (X number of new words, etc.). THAT would be majorly cool.

Sometimes the biggest issues I have is when i overzealously remove text , so your system would need to ignore the sign of the word difference: ±100 words :laughing:

Personally, I’d be reluctant to place both my live files and my back-ups inside the same system - single point of failure dependency, and all that.