What's the appropriate Disaster strategy ?

Is there a recommended best practice for protecting against loss of access to the project file ? I see the backup action also creates a proprietary file and that suggests to me some risk of loss in the event of project corruption.

I imagine there is a best-option compile action that will dump all the text into an open and editable format so that at least the raw material words are preserved (and in the right order !).

What’s generally considered best practice ? I intend to write in markdown.

Scrivener projects may be a proprietary format, but they’ve based it on open standards like plain text, XML, and RTF. At any time, you can burrow into a Scrivener project using just standard system-level tools and extract all of the text from your project if you need to – no Scrivener executable required. The bulk of your content is in RTF; the XML and text files hold metadata and the relationships of the various pieces of the project to each other.

A Scrivener backup simply creates a separate copy of the your project – optionally it will timestamp it and package it into an archive. You can backup all of your Scrivener projects to a single folder, or you can have a per-project backup location. You can let Scrivener manage the number of older backups for you, or you can set it to unlimited and take of it manually. Since you get to pick the backup folder location, you can easily specify that this be on a separate drive location.

So really, it comes down to what kind of risk are you trying to protect against?

Loss of a single disk – backup to another drive.
Loss of your system/easy re-install – backup to another folder and rely on Time Machine
Loss of your location (so your system and external drives) – use sync software like Dropbox (for your live projects) or OneDrive/iCloud/GoogleDrive (for zipped backups) to store your projects and/or backups in the cloud. This may also allow you to sync your projects to more PCs/Macs, and using Dropbox, to iOS devices.
Loss of data within your project – don’t rely on sync software for this, as it will faithfully replicated deletions. Use proper point-in-time backup solutions (like Time Machine) or, if you also want to have an off-site copy, some other cloud backup solution.

Depending on your needs, relative risk levels, etc., you could end up setting up an elaborate strategy.

I use Mac, Windows, and iOS. Here’s my personal choices:

  1. I store my live projects in Dropbox\Apps\Scrivener and this is the folder I use on Windows\Mac. if I lose a system drive on one of my PCs and have to rebuild, DropBox sync will take care of restoring my data. This also lets me use iOS to edit my projects on the road.
  2. I store my archived projects in OneDrive\WritingArchive. This is not accessible from the iOS Scrivener app so it limits the amount of data I have to sync with my mobile devices, but my Windows and Mac devices all run the OneDrive client, so they are accessible as well. Since these are archived projects, not projects I use on a regular basis, on my systems with smaller drives (like my Surface Pro 3) I have configured OneDrive to not keep a copy of this folder on the PC. My Mac has a 1TB drive, so it gets to keep a local copy. This lets me keeps copies of them around where I can get to them whenever I need, from whatever device I need to, and get that copy to a system where I can drag the project back into the live project folder if I need to work on it again.
  3. I store my backups in OneDrive\WritingBackups. Everything is zipped, includes the timestamp, and is set for unlimited copies. This way if something breaks with Dropbox, I still have all my backup copies where I can get to them, unarchive them, and drag them into my working folders.
  4. I use Time Machine on my Mac and Windows Backup on my Surface Pro 3 to backup to external drives.

The only piece I am missing is a cloud backup solution.

If you want to protect against the loss of Scrivener itself, probably the best solution is to File → Export → Files with the entire project selected. That will dump the entire contents of the project to a folder, with file names based on the Binder names of the individual documents.

You’ll lose all the metadata this way, so this should be treated as a true backup of last resort, for when your computer is a smoking ruin and your only way to get to your data is via the Dropbox (or other) web interface. If you have access to any device that will run Scrivener, using Scrivener to open the actual project (or a backup of it) will be a better solution.


Excellent post Devin! Very through and informative.

Yep - that looks like I wanted, to get all the source files back out


The XML file is very readable. If you have a Scrivener project you want to turn into separate files without using Scrivener, it’s not that hard to do by hand.

One of many reasons to have confidence in Scrivener.

I just went through this using Onedrive as a live storage container.


Among other things Windows Update will reassign your Onedrive to the default C: location regardless of where you had it assigned. I learned this the hard way about a week in. I’d load and save from what I thought was the location (G:) and nothing was syncing. Local as 1D on demand files were mismatched, corrupted. A real PITA. I went back to Dropbox for live syncing and run a zip backup once a week that I leave out on Onedrive as archive. The safest bet is Dropbox with about a 10 minute autosave. Local isn’t safe, 1D isn’t safe, Google is a mess. Dropbox and flash drives are your best friend.

First part of any disaster strategy, if something goes wrong, Don’t Panic! Don’t start clicking on files, opening, closing etc.

Sit and think over what led up to the disaster. What backups do you have? Possibly rename/move the disaster file so when you start opening files from cloud storage/backups etc they are not overwriting that file. ( it may turn out to be the best you have)