I am trying to use Scrivener on a laptop and on a desktop computer (at various times) and ensure the copies are synchronised. Scrivener 22.214.171.124 is loaded on both machines and both machines use Win 7. As such, I can integrate them into a ‘HomeGroup’ network and can see the network computer’s drives etc from each. To prevent synching issues, I intended to use the laptop as the live version (ie either use the laptop version standalone or via the networked drive from the desktop computer) and back up on the desktop computer and the cloud.
However, I am having issues loading / accessing project files through Scrivener when using the desktop (ie via the networked drive). I appreciate that there is a solution for sharing / synching files with Dropbox (shared on a forum), but I’d rather not use this method if I can get around it.
The alternative method using zipped backups described in the KB article about syncing projects should work for you. Basically, each time you finish working in the project, you close it and make a zipped backup to a generally accessible location, then when you next work, you pull the latest backup, unzip it, and work locally. That avoids the usual synchronisation problems of only pieces of a project syncing from one machine to another, and it should also give you a much faster load and save time than trying to work from a networked drive.
Hi. I was just looking for info on this very thing.
A followup on this (emphasis added), please.
I am trying to visualize that. But I am not quite seeing where the weak link is. Does the problem stem from network transfer time issues, in relation to the timing of Scriv’s save operations, which puts the .scriv file at risk of getting out of sync with the files it’s indexing, if the timings line up just right? Or is it that the files being worked on are not local (which of course is what accounts for the timing?)
Is this the same problem as with Dropbox? I don’t think I’d want to work over Dropbox. But being able to do it over my own local network would be neato.
One more: What makes accessing a project, over a local network drive, any different from having a really slow local drive, conceptually, I mean?
I’m about to begin passing the baton from an old and beloved Windows 7 lappie to a hot young Windows 8.1 box. I’d like to be able to work on either device, and access the very same project instances (obviously not at the same time), including the creation of backups, preferences and template paths, etc. over my Windows Homegroup. I realize that the two Sriv installations would be fully discrete.
So I’ve got networking on my mind. And would like a better understanding of the issues.
Read/write time is the bigger factor on networked drives than the bits and pieces that is the common problem when syncing. In the case of the network you’re dealing with only one copy of the project rather than two (i.e. the local and the one on the server, plus any on other linked machines) but it’s on a remote drive, so you’re limited by the speed of your network. If you bump up Scrivener’s auto-save interval you won’t be contending with that as frequently, though you’ll be writing bigger chunks when it does run and you’ll be flying with unsaved work for longer; if something goes wrong with the save because the project location is unavailable, you may lose some data. (You should get a warning that Scrivener isn’t able to save the file and then have the opportunity to copy your document text so you can save it locally, but meta-data changes could still be lost, and it’s still going to be inconvenient.) You’ll also have the usual network concerns to watch out for–a computer goes to sleep, someone turns it off, the wifi goes out, etc. depending on your particular set up.
So it’s not that you can’t do it, but just that you need to be alert to potential pitfalls.
The sync issue is a little different, because there everything is saving locally, but you can run into a problem of not everything syncing back to the server and the other machines. Scrivener’s package format means that it only needs to save the files that are being edited, which is much faster, but means you have a lot of little moving parts to content with rather than a single giant file like you get with a zipped archive. So you can edit documents A, B, and C and then shut down the computer before C has finished syncing back to the server, and when you go then to open the project on your second machine, C is out of date, even though the copy on your first machine has all the latest work. The problem at this point of course is that machine two is syncing back to machine one, so as soon as you touch C there, it updates to overwrite the “older” version on the first machine. Working straight from a network drive won’t run into the same case since the project only ever exists in the one location. It either saves or it doesn’t.
Thank you, MM, for that reply. It gives me a much better sense of how it works, and of what I’m seeing.
Thanks again for the reply. it really helped clarify things.
Yup. I think what I’m discovering is that it’s not really worth the additional mental bandwidth to do it over a network. And since I would never need access to the same project by more than one network “user” at a time, it’s easier to just use an unzipped local backup when working on a second computer.
What I did was create the zipped backup for transfer to the other computer, then created a page in the master project indicating that it is no longer up to date, and closed out the project on that page (without generating another backup) so I’d be immediately alerted should I open it in error.