Thanks for the answer!
Hopefully sooner than later, because ONE single mistake during a sync might mess up everything and force you to start from scratch. 
There is a simple solution to save your project. There is an option in Scrivener “Backup on project close”.
Yeah, you’re right, and I’m already making backups that way. I was lazy in my answer.
I meant “restart syncing setup from scratch”, if you made lots of changes to your sync settings.
I see there is a similar backup feature in Aeon, I might give it a try, to avoid these worst case scenario.
Thanks for the answer,
Matt
Aeon makes backup each time you save your project
Well, A backup stored in the same location as your main project file is not much of a backup…
If your file system, or disk get faulty, you have lost both your main project and you backup… If you don’t also use another backup software/regime in addition.
I store aeon project file in the folder of scrivener project folder. So when I start syncing prosess both files are backed up in that directory that I chose in Scrivener settings.
You can store your main project where you want
1 Like
And the backup is stored at the same location…
I’m not sure how much of an issue this really is.
Most people have only one disk. So their realistic protection from disk (or file system) failure would be a cloud drive. Their second option is a deliberate manual save to a memory stick or other external drive. And location makes no difference.
The backup provides a degree of version control enabling rescue from software or user issues.
I’m sure some of us have a multiple hard drives and formal overall backup systems with extensive version control. I certainly don’t need individual programs muscling in on that.
Though for something as complex as this, I usually like the ability to set my autosave frequency, though Saving doesn’t take much effort, and Save As allows you to save a copy anywhere you want.
And when you have done a “saved as” the project you works on is the “saved as” project, not your original, so you will need to close that and open your original…
None of this is a good practice for backup and safeguarding data…
Luckily I have some other mechanisms running
That was my main point - those that understand the need already have systems set up to do it.
No, you simply alternate two locations. One will have the most recent save, the other will have the next most recent. Simple manual system.
I store my projects in Dropbox.
Seriously not a good practice in any thinkable way regarding safety, it has nothing to do with safety, and if you work that way, you will have different “internal application backups” for the different files.
And at some point a user might lose track of what file they actually work on… so I really hope you don’t recommend that to anyone as a “backup solution”.
But of course, if you don’t like a good backup feature, any “workaround” will do…
The “backup” in AT3 is a false safety point, and is actually nothing but a “version point” of the project, it’s of course really great to have version control in a content creation software, but that do not equal a good backup function.
Just like the “Save restore point every x min.” in Word and other Office tools is not the same as a backup of the document.
I tend to also think about those who don’t know much about backup and content safely store their content… I have helped to many people (including overconfident writers who thought their Mac was “doing it for them”), trying to recover lost data the last 20-25 years or so…
Some people need a little “extra help” to get things set up a good way, and a good backup function in a software is one step on that road.
So for me, the only good backup solution for any software, is to have setting for where a user save the backup, it should be one of the mandatory settings for any software that creates content… and it should be possible to set the intervals for the backup… Regardless if its a full or incremental backup…
If people then choose to backup to their working directory, regardless if it’s online or not, it’s their active choice to do so…
A “backup” is not a backup if it by default is saved to the same folder as the “original”, it is only a copy, or as in AT3 a versioning function, nothing else, it will never be anything else, until it is saved in another location than the working project, and that location should preferable be on another physical storage device… and it do not help anyone to speak heresy.
Actually I disagree completely.
This is not a backup solution. A backup solution is a process/programs(s) independent of any individual software that produces content. It should cover the whole system.
Having individual apps doing ‘backup’ really is a recipe for a memory nightmare.
I don’t know about false safety point. I agree it’s not really a backup - it is a limited number of version points. This is sufficient to help people with the problems they encounter more often. And more explicit that Word’s restore points.
Can’t lose track of the file - creation time in the OS gives you the sequence. It supports the possibility of different physical devices. Allows users to know explicitly what they are doing and where the data is. It supports a planned programme (eg memory stick every day, external SSD once a week). It’s not as good as a proper backup program and system, but it is capable of being applied systematically within the limits of the system the user possesses.
not everyone is as clever as you…!
I have cleaned up in companies where their users had 50 or more copies of the same document spread around and where they open and edited different documents on a daily basis, ending with a massive complain about “unstable systems” and “other users delete my edits”.
I have been working with this the last 15-20 years or so as a IT consultant and network administrator.
If a software say it has “backup” it should be a backup as I described earlier, if not, they need to call it something else, because when users read “backup” they think its a magical save all to a magic place where its safe forever…
When a software say there is “backup” in the software, it needs to be a REAL BACKUP function, else, they need to call it something else… Not everyone that use AT3 are as clever as you or has worked with computer systems for most of their adult life, so try to put yourself in their shoes!
Don’t run around give bad advices, or actually, just do that, because then those of us working with data restoration and helping people with good backup solutions will keep our jobs for a long time.
And as I stated before, and it is still valid even though you try to argue against it, the “backup” in AT3 as it is now is only a “versioning manager”… and as an digression, that can also be set in the Windows file system, without any problems… for those who knows how to do it…
You clearly just argue to argue and try to get the last word…
Software with a good backup feature is nothing like that, it is actually the opposite, because if you have a good backup in software, you will be able to save all backups to one location and set your backup software to backup that folder, instead of having to backup and keep track of 10-150 or more different folders where users save their work…
And if a user don’t know how to install and run a backup utility, at least they can copy that one folder to an external location or 3.
You can disagree as much as you like, it’s a free world, most of it at least…
Jjust do not run around and advocate heresy!
Yeah, I can confirm this happens to me too. Each time I perform a sync from Aeon, Scrivener tries to resync the external folder as if each file was modified. Obviously it’s not the case, as you have already pointed out, but I suspect “last edit time” or something like that comes to play (considering that’s the only way Scrivener has to detect external changes, I guess).
Matt
And for those arguing over backup… just work on the file within a cloud synced folder (OneDrive, Dropbox, etc…) and boom, you’re done (for the majority of folks out there. The paranoid like myself will make an extra backup elsewhere, but that’s another story. This is not the place to flex individual tech knowledge.)
Sure, but there are more and more people that don’t want to save their work on cloud services anymore…
Personally I never recommend storing data, and specially not database files like sqlite or similar in any of the sync services…
For most people it goes just fine, but the minute you get an sync conflict, getting that fix can be a lot of problems…
Hi all,
I think at this point, this thread has veered away from a productive conversation, so I will be closing this thread to avoid the discussion escalating further.
Please try to keep discussion civil and on-topic, and avoid any unnecessary antagonism or name-calling.
We are a small development team and an even smaller support team, so we are not equipped to moderate off-topic conversations that don’t meet the aims of this forum (assisting users and gathering feedback).
Note that I have chosen to delete the last few off-topic posts as they don’t contribute towards those aims.
Scrivener Syncing Issues
We do appreciate the feedback regarding Scrivener syncing. We have tried hard to be flexible in how we fit into different Scrivener project structures, we have run 6+ months of beta tests to try to ensure we have covered any bugs or issues, and we have tried to provide enough guidance to help people understand the process.
But syncing is often a complex area to both implement and document, so I fully accept that there is more we can do on both documentation and better guidance and protection within the app itself, and we will continue to work to improve these issues as they are reported to us.
If you experience Scrivener syncing issues, please start a new thread to continue those conversations.
It is generally easier for us to deal with each issue independently rather than having a large number of issues accumulate into a single catch-all thread.
Automatic Backup
I believe it may have been on another thread, but I have previously acknowledged the request to be able to allow users to choose their own backup file location, which is something that is on our list of tasks we are balancing and prioritising.
I agree that backing up files onto the same computer does not provide the same protection or perform the same function as backing them up elsewhere - whether that is off-site, in the cloud, external hard drives etc, and I would always encourage users to make those separate arrangements also.
But the mechanism we have does provide a level of protection against both inadvertent user error (i.e. accidentally mucking up a file), or software error (i.e. a bug causing data to be lost from a file), which are the primary motivators behind the backup system we have implemented – and generally more common than a house fire or complete disk failure.
This won’t protect against every case, but it will protect against more cases than if it did not exist, and so I would prefer to offer some level of protection than to offer none.
Personally, I think the debate around “versioning” vs “backup” is splitting hairs a little. At this stage, I prefer to continue to use the term “Backup” simply because this is a term that is better understood by more people (which is the most important thing in the moment a user has that sudden panic about potential loss of work).
Sorry to post a response and then immediately close the topic, but I thought it was important to at east respond to the points made before closing it down.
Matt
2 Likes
Hi Matt,
Each time I perform a sync from Aeon, Scrivener tries to resync the external folder as if each file was modified. Obviously it’s not the case, as you have already pointed out, but I suspect “last edit time” or something like that comes to play (considering that’s the only way Scrivener has to detect external changes, I guess).
Just to update on this particular issue:
I have implemented a change for our next build that will avoid updating the “modification date” inside Scrivener unless the content actually changes.
Unfortunately, on Windows, this still results in the modal popping up and it being a several click process to sync everything and close it again, but the sync itself should usually be faster as it will only sync the files that have changed somehow. Note that metadata changes will count as a change, as this mirrors the behaviour Scrivener has implemented.
That fix will be rolled into a future update with a bunch of other small fixes (most likely the next update, unless we have to push out any hot fix for a pressing issue).
Matt
1 Like