Scrivener users, where do you backup your files?

I first owned Scrivener on my Windows PC before transitioning over to the MacBook. I tried to open the Scrivener files I backed up onto Dropbox, but it wouldn’t work. I’m assuming it might have to do with my second Scrivener being a Mac version - there was no solution I could find on the Internet for it. I’m going to stick to the Mac version, but the problem with opening the file made me skeptical about how reliable Dropbox can be.

What’s your preference for Scrivener files?

I keep the primary files in Dropbox because I use Scrivener on both my Mac and my iPad Pro. Backups go to iCloud.

It has been said many times that Dropbox is a synchronisation service, not a backup service. There is a difference, and it is as well to be aware of it, and to understand it.

Having said that, I have kept my working Scrivener projects on Dropbox for many years. Backups (which are created automatically by the program) are saved as zipped files to OneDrive. In addition to that, I have Time Machine making backups of my hard drive, Carbon Copy Cloner making backups every evening onto a separate partition of an external hard drive, and I use BackBlaze for off-site backups of my hard drive. I feel I’m reasonably well covered, but I await the day when they all fail together …

Dropbox absolutely is a backup location, especially so for zip backups. Even if you delete a file on purpose, it’s still there for 30+ days. More than once, after buying a new Mac, I’ve restored from Dropbox. I’m puzzled what else we’d want from a “backup”.

I send the zips to my NAS and any ‘live’ files are written locally and then backed up to a cloud service as applicable.

Belts and suspenders.

I use Time Machine, Dropbox, and ChronoSync.

I can only refer you to the advice given by @kewms in this post: [url]most of my projects are lost! - #8 by kewms].

I agree with mbbntu

as for drmajorbob -

  1. A complete image backup
  2. Keeping files for years, even if I didn’t know they got screwed up.
  3. Keeping multiple previous versions automatically
  4. NEVER EVER changing my working files

Sync is fine, but also, have multiple backups onsite and offsite.

I suggest, dropbox all you want but also:

  1. Make zipped backups in scrivener at least 10 maybe more.
  2. Backup the zipped files and your regular files every night (at least) to an external drive (NAS or USB)
  3. Backup again at least every day to the cloud.
  4. Make images of your entire working drive regularly.
  5. Be sure you can get your backups back. Are they so big it’ll take a week to download them, if so, what are they worth? Will the backup company make a disk and express it to you?
  6. Check your backups regularly to make sure they are working.

A syncing service is a fine repository for zipped backups. But placing a live project on a syncing service is not a backup of that project.

A backup is a static copy of your data. A syncing service, any syncing service, maintains a living dynamic copy of your data. One is a snapshot, the other constantly changes. These things are polar opposites.

Yes, you can use the syncing service copy to restore your data to another device. Yes, you can use the syncing service copy to help you resolve certain types of data loss issues.

But a syncing service can also destroy your data. A true backup copy can never cause issues for you.

A poster here recently wrote how they inadvertently deleted a chapter from their novel. Didn’t realize they’d done so until months later. Their syncing service couldn’t help them, in fact it made matters worse by propagating the changes to all of their devices. And they had their backups set to the default of 5. Oh well. So they had to rewrite the chapter.

Bob, I know you probably understand all of this, but there are many posters here who throw their live projects onto Dropbox or some other syncing service and think they’re covered. They’re not.

See this post from Devinganger, a frequent poster here. What he is does and what I do are identical: [url]Usefulness of the iOS app? - #5 by devinganger]


Three different TimeMachines, connected to three different Macs, and all backups also stored in another cloud service than Dropbox.

  1. In 50 … no, 51 … years writing code and otherwise using computers, I’ve never needed an image backup. I need my files … and I don’t like the ungodly amount of time it takes to make or restore an image backup. The primary reason to need an image backup is Windows … which I haven’t used in almost 20 years.

  2. Dropbox and Time Machine keep files longer than I’ll ever need them.

  3. Every zip backup is a new version, and Dropbox keeps all of them.

  4. Dropbox never ever changes my working files.

That is multiple backups onsite and offsite.

The IT department head at a pharma CRO where my wife worked deleted their main database … then deleted the backup … and wasn’t fired until he reconstructed as much as possible. No one is covered if they don’t pay attention and don’t know what to do when (not if) things go wrong.

Without further information, it’s impossible to say whether your experience can be blamed on Dropbox or not.

For example, Scrivener cannot open ZIP files. If you make a ZIP backup on your PC, upload it to Dropbox, and download it to your Mac, Scrivener will not be able to open it until it’s unzipped. That’s nothing to do with Dropbox, and you would see the exact same behavior no matter how you transferred the file.

On the PC, you double-click the .scrivx file to launch a project. But the .scrivx file is only the index. If you upload only that file to Dropbox, then download to the Mac, Scrivener will report an error. But that’s nothing to do with Dropbox, and you would see the exact same behavior no matter how you transferred the file.

Do you see the pattern? Dropbox can facilitate a variety of user errors which can lead to data loss. But Dropbox itself is in my experience pretty reliable. I see it as a useful part of a comprehensive backup strategy. It should not be your only backup. A damaged Dropbox archive will propagate to all connected devices at internet speed, potentially making a bad situation worse.

IMO, the first line of defense for all Mac users should be a Time Machine backup. The software comes free with your Mac, is stupidly easy to use, and only requires that you buy and know how to plug in an external hard drive.

I also use and recommend BackBlaze, an offsite backup service, and SuperDuper, which copies a disk image to an external drive.

Getting beyond generic advice, it’s important to think about exactly what the threats to your data might be. Do you live in a wildfire zone and worry that your house might literally burn to the ground? Are you worried about curious toddlers, your own clumsiness, or malicious adults? Is Scrivener your only critical application, or do you also need to protect financial records, family photos, communications with political dissidents?


Time Machine repeatedly failed on three USB-C drives and I never found an explanation or fix. I had to abandon those drives for that purpose and switched to using ChronoSync to save all my files on them. (All my files are in the Dropbox directory, and that’s what I sync with ChronoSync as well.)

Time Machine works on the current external drive … as far as I know. Like any backup, you don’t know it works until you need it. Even then, how can you know everything was properly restored?

I can’t see any reason to save a disk image. Maybe that’s because, since moving to Macs over 15 yrs ago, I haven’t worried much about hard drive integrity, even less since switching to SSDs. It would have been nice to have a disk image when I bought the new M1 in January, if it would have prevented me having to reenter licenses for every damn thing. Would it? If so, I’ll reconsider my position.

And aren’t we all political dissidents?

My husband recently found a disk image backup very handy when he had to revert a Big Sur installation back to Catalina. A friend was delighted to discover, after a catastrophic hard drive failure, that BackBlaze would mail her a physical drive with a complete archive of her data.

I think the thing we can both agree on is redundancy. No matter how rare failures of any given method might be, it’s no fun to find that you’re an outlier.


I keep only certain live Scriv projects in my Dbox folder — the ones I want access to on ios Scriv.

I have Scriv set to automatically make a zipped backup when I close a project. These are stored on my Mac’s hard drive.

I use an external hard drive to back up my mac. (Okay, actually there are two: one is a time machine backup (which has a wayback function) and the other is a carboncopy backup (and kept at work rather than home). I do this sort of backing up when macos reminds me to, so every ten days or so.

This provides me a good amount of assurance and is easy to maintain. Minimum outlay: one good sized external drive.

A small point of order: that’s written as if “in the Dropbox folder” and “on the hard drive” are mutually exclusive. Not so. If there is a Dropbox folder, that folder is on the hard drive. There’s a “smart” sync feature now that moves some files off the drive leaving only links to them, but that’s not the case for anything recently created or modified … and I haven’t seen much change at all since activating the feature. Dropbox takes up 204GB on my drive.

True dat! I was an outlier twice, once when a virus ate my Windows hard drive and again when I simply forgot to save my Quicken files when moving to a new machine. That was 20-25 years ago, and my habits (and favorite OS) are a lot better now.

No, Bob, it is not. It is a sync service with versioning. What Dropbox does is sync changes to files and keep versions of any of those changes depending on storage space and speed of resolution. However, it is primarily designed to keep multiple copies in sync with the latest changes – if I delete file A on computer 1, if everything has adequate network access I can reasonably except it to be deleted on computer 2 in a short amount of time. Likewise if I add or edit files.

With the versioning added, I can go into the Dropbox website and download the past versions of discrete files. But I can’t say, “Give me this project as it existed at X point in time” and have it automatically restore that version of the entire project – I’d have to manually re-assemble it from the versions of the various component files.

It is not a true point-in-time backup service, where you have a discrete set of files that were all backed up at the same time (or close to it, depending on whether your underlying file system has something like a snapshot capability) and then those files are all clumped together.

There are some use cases where Dropbox’s capabilities are reasonable enough to work, but that does not make it a true backup solution any more than the fact that you can strap small table to the top of a VW Beatle makes that Beatle a moving van.

Wrong. I can restore any folder as of a point in time, and a project (to Dropbox) is a folder. I can restore a folder that contains all my projects as of a point in time. I can restore Dropbox itself as of a point in time. That’s a backup. I may not be able to rewind more than 30 days in all cases, but nor do I need to. If I did, I have Time Machine.