Need help - backing up Scrivener files w/ Chronosync to .mac

From long habit, I’ve been keeping my Scrivener files in the folders I’d set up for the various book projects I’m working on.

I’d like to set up an autobackup system where Chronosync would back up all my Scrivener files to .mac.

I tried using a Smart Folder to gather all my .scriv files together, but Chronosync wouldn’t recognize either the folder or the files in it (probably because they weren’t really files, just aliases).

So I’m at a standstill. Any ideas? I do back up to a separate hard disk but I’d like an off site backup as well.

How familiar are you with the command line? If I wanted to back up all of my scrivener projects, where is how I would approach it. I would open up Terminal, change directory to ~/Documents, and run a find piped to tar to create an archive.

% cd ~/Documents % find . -name *.scriv | tar czf scrivener_backup-DATE.tar.gz -T -

This will create a file called ‘scrivener_backup-DATE.tar.gz’. Naturally, you’ll have to substitute ‘DATE’ when you input the command. So, to do this automatically, you would need a script that is executed periodically with cron, preferably just before Chronosync is schedule to run.

Oh, and the reason nothing can read Smart Folders is because they aren’t actually folders; they are just XML files describing the search conditions. I’m guessing that whatever triggers the search is something esoteric, because if it were as simple as reading an XML file, everyone would have figured out how to handle Smart Folders in their applications; instead even Path Finder cannot do anything with them. Apple kept the whole thing secret.

Hi, Amber,

Thanks so much for your response! :smiley:

As to my comfort level with the command line… I grew up (computerly-speaking) with DOS, so that sort of thing doesn’t bother me, but I’ve never used it with the Mac. (And I bought my first Powerbook - a 540 - new a long long time ago.)

Will what you wrote go through lower level folders? I keep my writing in a separate folder (not ~\Documents), with each project given its own sub-folder, meaning that my .scriv files are in several different sub-folders of higher level folders, all under ~\Writing. Does that make sense?

Thanks for the info re: Smart Folders. I knew that the contents of Smart Folders weren’t actually there, but I didn’t realize that SF themselves weren’t ‘real’.

I should probably learn a different storage structure, but some old habits die very hard indeed.

Again, thanks for your help.

Yes, you’ll want to change directories to the highest point in the file system where you can still hit all of the .scriv files below that point, to reduce processing time. You would want to type ‘cd ~/Documents/Writing’ because searching the rest of Documents would be a waste of time. The UNIX ‘find’ command which I used to assemble the file list for ‘tar’, will automatically descend from the current point as deep as it needs to go to find everything that matches ‘*.scriv’.

So the nice thing about this is that tar will preserve the Finder structure involved. Not only will every Scrivener project get saved, it will save the folder layout, too. This will adapt to whatever storage structure you prefer.

You can easily create the same kind of thing with Chronosync by having CS search the highest ranking folder with scrivenings in it and setting the extension to ‘*.scriv’.This will preserve the folder structure just as well and should also preserve the metadata which regularly get messy when using the commandline (unless you are using Leopard).
If you don’t use metadata (tags, colour labels etc) the script AmberV describes does the job just as well, it just boils down to preference I guess.

Prion

I know of at least one other user who uses Chronosync to back up. He had one minor issue with is easily fixable, so I ought to warn you about it:

Be sure to check whatever option is available in Chronosync to delete files from the backup that have been deleted from disk for your Scrivener projects. This is very important, or your backed up projects will start acting very strangely.

The reason for this is that .scriv packages contain lots of RTFD files. Each RTFD file’s title is a number that corresponds to the internal ID of a particular document in the binder. Whenever you open a project, Scrivener cycles through every available document in the binder (but it does not check the hard drive, which would take significantly longer) for the used IDs, and from this takes note of the next available ID, which will be used for the next document you add.

Now, if you backup a project from which you have deleted some recently-added documents, and the backup is merged with an older backup in such a way that the binder information now has no knowledge of those file IDs, but RTFD files still exist in the package with those IDs (because you didn’t check the option to remove deleted files from the backup), what will happen when you add a new document to the project is that Scrivener could very well allocate an ID to that document for which there is an associated RTFD file on disk. So, you will create a new document and find that it contains text from an old file.

Of course, there is nothing particularly dangerous in this - you could just delete the text and carry on with no adverse effects - but it would be pretty annoying.

Moral of this prolix proliferation of words: check that box. Whatever it’s called.

Best,
Keith

Colours should be fine, as should most other meta-data. If you use some sort of tagging system which utilises Spotlight Comments, then that could cause problems. But most of the issues that the command line used to cause in the past were resolved with Tiger (or some update that came a bit later).

As with all backups, you’ll want to test the integrity of your archives before trusting them.

But if Chronosynch has some built in way of doing this, that would be dramatically preferable. Something could also probably be arranged using Smart Folders and Automator; I don’t know much about Automator though.

The thing is called “sync deletions” and makes an appearance in form of a checkbox in the middle of the topmost Chronosync screen (the one you use to choose source and destination).

I keep that checked all the time because otherwise I’d get spammed with all the rubbish I may have written in the past and had enough sense to throw away at some stage later :smiley:

Prion

PS: Amber,
I was talking about spotlight comments indeed. In fact I am unsure about the details of what will go wrong precisely with the current 10.4.9 system. If you have reliable information, I’d love to hear about this.

Studio717,

are you familiar with Quickpicks? I see you have a .Mac account. What I do is use Apples Backup application to save all my files on my idisk. Backup allows the use of Quickpicks, who will find and archive all Files on your disk of a certain type - eg .doc, Omni Outliner, iPhoto Pictures - whatever. I made one for my Ulysses files, it could easily be made into taking Scrivener Files instead. I find this is the easiest way to save your work (only drawback is of course the cost for .Mac-membership)

edit:
You know what, I just created a Scrivener-Quickpick. If you want, you can get it from my iDisk:

idisk.mac.com/tfunke-Public/Quic … ckpick.zip
(only about 9kb)

More on Quickpicks and how they work can be found on this site:
johnhugg.com/quickpicks/

I have just been testing Scrivener for a few days and am very impressed. Though I own a copy of Ulysses and like it very much I might spend the extra money. Only thing I like more about Ulysses is the Full Screen implementation.

th.

Hi, Thorsten,

No, I didn’t know about Quickpicks, thanks! I’ve had .Mac for years and of course know little about it other than for email and backup. :blush:

Thanks for taking the time to create the file. Unfortunately, when I try to download it, it says that it could not decompress it successfully. Perhaps I need to do something I’m not doing?

It does look like Quickpicks solves my problem for me, though. Wow, .Mac is finally coming in handy! :laughing:

I compressed it with Mac OS X built-in archive function so normally this should be no problem. I am not at my May today (will be tomorrow) and tried to download and unpack it from my father’s windows machine - and indeed it cannot be decompressed. I will take care of this tomorrow.

You’re welcome, it was only a matter of two minutes to modify it.

One more note: This quickpick leaves out the folder /Library/Application Support/Scrivener so you should not put your files there. But anywhere else inside your homefolder, no matter how deep, should work.

th.

it seems there is something wrong with downloading it via direct link into the web browser from the iDisk. It should be 9 kb big, but when I download I only get 4 kb. Unpacking then does not work, of course. Try downloading by using the Finder (connect to iDisk public folder of user tfunke).

Or point your browser to

idisk.mac.com/tfunke-Public/Quickpick/?view=web

and then download the file via the arrow on the right.

Let me know if you still encounter problems or don’t know how to install it. Feel free to email me.

th.

This may be useful.

I use .mac and Chronosync to keep my Srivener files in sync between 2 Macs.

I could just use Chronosync to do the same thing, but that means mountin the other mac each time and I like automated processes.

I have a Chronosync file on each Mac set to sync my main writing folder (which contains my Scrivener files) to my iDisk. Chronosync is set to do this every few hours. This means that anything I have wrien is backed up to iDisk automatically. In addition, Chronosync compares the contents of the folders on the Macs and iDisk and keeps them identical.

This works perfectly for me, without error. if anyone needs more clarity or detail please just let me know.

I would like more explanation, just a basic step by step. I just got a mac TODAY and have .mac . I am fairly familiar, as I have a mac desktop at home, but am in DC doing research so the laptop will have nothing on it. I am familiar with dropping things on idisk that you want saved.

Thanks for the explanation of how Scrivener works with backups, it seems that one would be better off with the “Mirroring” option, rather than the synchronized option, based on my FTP program that I use to backup to my website online.

I will experiment.