Using Scrivener with Dropbox

Hey AmberV - thanks for that advice! I’ll give it a go - carefully of course!

Tripper: your tech support conversation regarding indents has been moved here.

Hi,

I’m afraid I don’t understand the instructions below. I’m trying to sync between two macs and an iPad (like a lot of people here it seems). My .scriv file is in dropbox, as is my external folder (with Drafts/Notes subdirectories) for PlainText on the iPad. I can’t seem to get my second mac to sync with the .scriv file/folder combination. Should I just give up-- that is, is this impossible (in which case I’d add my plea for this feature to be incorporated into future versions…)? Or am I doing something wrong? Many thanks!

Norm

Scrivener doesn’t actually do anything here. If you have having difficulties getting one of your Macs to keep up to date with Dropbox in general (files are downloading/uploading), you might want to troubleshoot why that is. For example, maybe you are running an older version on one of the Macs that isn’t compatible with your operating system. The Dropbox forums might be able to help you out better.

Let me know if that isn’t what you meant, though.

Norm, you might also want to try checking that each of your Macs has the most current version of Dropbox - it doesn’t seem to flag you when it’s time to update, and I recently had just that problem.

On another matter: I work in two locations, and use Dropbox to keep backed up and up to date. I of course live in fear that my .scriv file will get corrupted, but if it does, how will I know? Will Scrivener simply refuse to open the file? Or is it a creeping, incremental thing, where when I’m at page 288 of my novel, it’ll suddenly announce to me that I can go no further, since the file is epically corrupted, and has been since page 23?

The term “corruption” is a bit dramatic when used to refer to content. This might sound strange, but it is possible for a project to get corrupted, while I’ve never heard of Dropbox ever corrupting data within the project. Here is what can go wrong: Dropbox, when it gets confused over which file is the latest, will produce conflict copies. The machine that produces the confusing result will have its version uploaded to Dropbox with the filename appended to state that it is a conflict copy, which machine it came from, and the date. Ordinarily, this is pretty obvious when it happens. You have a .doc file you’re working on in a folder, and suddenly a duplicate appears in the folder with “(conflicted copy from blah blah).doc” added to the name. You open both of them, figure out which is best, merge bits of text if necessary, and then delete the redundant copy. Case closed. Synchronisation is inherently risky like this, and so fall back measures are important.

Only problem is, in Scrivener all of the files that comprise a project are essentially hidden from you. Scrivener expects them to be named a certain precise way, so if something else comes along and starts duplicating stuff and renaming things, the program can get mislead. If the main .scrivx file, which controls what you see in your Binder—it’s the primary architect file in Scrivener—states that file #308 should be here and contain this, but in the actual Dropbox conflict merge, #308 is something else, or an older version from another computer that got conflicted, things can get messy. If the mess is bad enough and even the .scrivx file is forked, the project can refuse to open.

But to be clear this is all data level mismatching. That is what I mean by a project getting corrupted but not the data. The actual words you’ve written are all okay, they are just no longer operating as a seamless unit of data within the Project Infrastructure. It means going into the .scriv folder in the Dropbox web interface and teasing these divergent threads apart, figuring out which is which and what is best, and reconstructing a whole functioning project again.

All of this can be avoided by not producing conflicts in the first place. The three point guidelines published on this forum, and in the user manual, demonstrate techniques which should in nearly all cases avoid conflicts. Heed them and this should never be a problem. Ignore them, and you might find your afternoon is wasted poking through conflict files and restoring versions in the Dropbox website.

But you shouldn’t ever lose a single word, or end up with an RTF file that is corrupted. Not with Dropbox anyway.

Fantastic, Amber! I shall continue to heed the Golden Rules…

Thanks for taking the time to explain.

I’m confused. Despite the big warning in the first post, not to use External sync to save to Dropbox, many users seem to do so anyway?

The other part I don’t get is how to get a file from DB onto another Mac. Finding the right version on DB, dragging it to the desktop, then unzipping it, (as described by other users) seems not that straightforward and prone to accidentily opening a wrong version. Why not sync the project from DB directly into Scrivener, work on it, and then sync it back to DB? What am I missing here?

Will iCloud come to the rescue when it is released in a couple of weeks?

There’s no prohibition against this, but it’s main purpose is not to sync from Scrivener on one machine to Scrivener on another machine, but to sync between Scrivener and iPad apps like “Index Card” and “Plain Text”, which can’t read a Scrivener bundle, but can access your dropbox folders.

Being especially careful to follow the instructions in the first post of this thread, simply open the project directly from your dropbox folder.

Not likely. iCloud seems to be designed for syncing individual files, like a Pages or Word document, not a complicated bundle of files like Scrivener’s.

So just to make sure I understand, just plain open and save from/to my DB folder on both Macs?

Yes. Never at the same time, of course (see page one of this discussion)*, but it’s that simple. I do it between my beta Windows Scrivener and my Mac’s Scrivener projects frequently.

*I’m sure you’re taking that into account, but I want to make absolutely sure that Keith’s advice on the first post is heeded by anyone reading this far back in the conversation.

Edit: wasn’t paying attention, could have sworn that Ioa had started the thread, not Keith. Edited for accuracy.

You mean this post: Working off of network drives (MobileMe, thumb drives...) ?

There it says at the bottom:

Confused again…

No, I meant page one of this thread. Scroll down to the bottom or the top of this page, and look for “xx posts Page 1 of y 1 2 3 4 …”, and click on the 1. that will take you to the first page of this discussion, where the instructions I’m referring to are.

I mistakenly thought that Ioa (not Keith) had started THIS thread. Sorry about that.

If you are careful, you can ignore the advice you quoted and use Keith’s instructions instead.

Edit: Though it feels wasteful to repeat Keith’s words at the top of this thread, you should read it (and everything that follows in that initial post:

"Using Scrivener with Dropbox

Postby KB on Sun Feb 13, 2011 12:11 pm
Hello,

Our support thread on “Working off of network drives” has understandably left many users concerned about using Scrivener with Dropbox. However, the problems with storing .scriv files on Dropbox have been somewhat overstated, mainly because we want to ensure users know that no syncing method is 100% safe. So, here are some guidelines on using Scrivener with Dropbox that should keep you out of trouble."

To add to Robert’s wise advice above – bear in mind two things: the dates on the posts and your own personal appetite for risk. The post from Ioa/AmberV that you’re referring to predates the post from Keith that begins this thread by – I think – two years. In those two years, I guess people became more comfortable about putting their Scrivener project directly into their DropBox folder, rather than a zipped backup.

But in Keith’s post, notice not only his specific caveats but also his warning that no syncing is ever going to be 100% safe. And this warning is what relates to your appetite for risk. I have a low appetite for risk with any writing that might earn or lose me money, so I follow the Ioa/AmberV advice for this type of writing and only ever put zipped back-ups into my DropBox; the original project folder remains in “Documents” on my machine. Of course that’s not 100% safe either, but obviously that’s slightly safer than putting the project folder directly into DropBox. If I were writing purely for fun and personal enjoyment, I would probably be more relaxed about using the method Keith outlines.

Thanks again guys, I get it now :slight_smile:

I think the confusion for me also comes from the naming of the menu items: External Sync vs BackUp vs Open/Save.

Not sure why syncing with Dropbox should be construed as risky. Even if things go wrong, they can be fixed. First, you always have your backups. Second, you can always roll back to before syncing. It happened to me last night (and obviously it was my fault, I should have known better). I just rolled back to before syncing.

Using DB (carefully) to sync between two Macs, I note that if I save the project to the DB folder, when I reopen it (whether on the system used to create the project, or the system I am syncing with), the coloured icons and saved searches generated using metadata no longer show in the Binder. Otherwise, the project seems to be entirely intact (for example, the metadata used for the saved searches, and to colour the document icons, is preserved). Any ideas why this should be? It is a small inconvenience, but I have found the saved search facility particularly useful.

That information is saved in a file called “ui.plist” in the Settings folder within the .scriv package. Check in there and see if Dropbox created a conflicted version of this file. It might be you made those changes on the machine that generated the conflicted copy, and so Scrivener loads the one without those UI settings since it isn’t going to load the conflicted one.

Thanks. The problem was exactly as you described, with the ui.plist file marked as conflicted. Not sure how the conflict arose, however, as I was not aware of opening that file on a machine other than the one used to create the file in the first place.

HI there

I am trying to work on my book on both my Mac and my PC machines. At home my iMac is my main machine (except when I am in bed) and away from home my PC laptop is my main machine. I spend about equal time on them.

I have a folder I save to on dropbox and which I access from both computers, making sure that the book is only open on one machine at a time. However the Mac and PC versions of Scrivener do not seem completely compatible with each-other, as evidenced by weird things happening like all the quotation marks disappearing throughout my novel (Yes this happened last night when I opened my project on my PC).

What I would like to know is, is there some sort of compatibility mode I can save in on the Mac to make the Scrivener file 100% compatible with no issues to the PC version? I don’t want to just use external folder sync and Word on my PC, I paid for my Scrivener version on the Mac and I’m planning to pay for the PC version once Nano is over. Right now I’m only working on NanoWriMo but once that’s over I will be preparing a previous novel for publication and won’t be able to risk these sort of problems.

Thanks for your time! Love the program and couldn’t imagine using anything else now that I’m used to it.