I recently updated to Scrivener 1.2.1.0 (on WIndows XP / SP3)
Today I converted a folder in the binder to a file.
The folder contained several text files (seven, in fact).
I converted the containing folder back to a folder.
I noticed that the icon of one of the text files in the converted/unconverted folder now differs from the others and the text that was once in the file seems to have disappeared.
I’m sure, or rather, I hope, that the text hasn’t been deleted.
Could anyone tell me the procedure for recovering the text?
Do you have any snapshots for this particular item? If not, I’d visit the Backups tab in Tools/Options and click on the button there at the bottom to load your backup folder. You should find at least five (by default) .zip files corresponding to your project. I prefer to disable the automatic backup system while going through backups looking for a recovery copy as closing copies of backed up projects generates new backups, rolling off older ones prematurely. Just remember to turn it back on after you close the last backup project. Anyway, once you find the file, drag and drop it from the backup binder to the working project’s binder to bring it back in.
That’s for recovery—it would be nice to figure out what happened though. Is the project stored somewhere other than an internal harddrive, or in an location that is subject to synchronisation?
I’ve located the zip files (oddly, the middle of five seems to have the most recent date stamp!)
Do you mean I should drag the zip file into Scrivener, or open the zip file, drag the contents out to another location, and open the scriv file via Scrivener’s “Open” menu?
(The working directory is a Dropbox folder, the aim being that the project can be opened on more than one machine – the project is not currently open on the other machine, which is on the other side of town and is currently switched off)
I’ve explored the two most recent zips and have dragged the Docs folder out of the zip file and used the WIndows search facility to scan the files therein for a phrase from the file that I’ve lost (it seems the only other option is to methodically open hundreds of rtf files, each apparently named with a random number!)
The search doesn’t find the word or phrase.
Am I searching in the wrong place, or are more sinister forces at work?
You should just open the backup project, and use Scrivener to locate the text you’re missing in the main project. That will be far easier than manipulating the individual files inside your project.
Windows Exporer can open zip files as well natively. Anyway you need to extract the entire “my project.scriv” folder of the .zip before it can be opened. Scrivener will not open .zip files. It’s just a full copy of your project in there. So ultimately you want to have the backup copy opened at the same time as the current copy. Then you can drag files in between them from binder to binder.
Yeah, that’s why I had you check modification dates. It rotates through five backups by default (you can increase or decrease how many it keeps). If you’d like a more logical (if messier to read) order, enable the date stamp option in the backup tab.
I know that now, although it took some trial and error (hence my second post), which was compounded by the fact that I was using 7-Zip, which appears to be somewhat limited in functionality due to the difficulty it has in working with the zip format backup files created by Scrivener: it just couldn’t extract the contents of the zip files. It took a little while before it occurred to me to try a different unarchive programme. As previously noted, WinRAR extracted the contents of the archives without any difficulties.
Perhaps that could be a feature to add to the wish list.
It would be quite handy if Scrivener could leverage the functionality of working with the archive files it creates in order to provide users with a one-click solution to accessing such an archive: select the zip from a dialogue box or menu within Scrivener, click the “open backup” button which prompts the user to specify a destination directory into which the zip is extracted before Scrivener simply opens the project in a new Scrivener window/instance.
This clarifies what I didn’t quite glean from your previous post but, after previously noted trial and error, managed to work out.
I missed the bit where you got me to do this. I sort of had an inkling though that I would need to go far back enough that the backup file I opened had the material I needed, but not so far back that I had opened a backup that had been created before I had actually written the missing text.
All I need to do now is restore the “automatic backups” option that I turned off.
Thanks for your help in getting the missing text restored.
I’m ultra-paranoid. I turned on time-stamps so that when flustered, I can easily sort the backup files by date, and I also increased the number to be kept to the maximum. Also, I send the backups to my Dropbox folder, so that they’re somewhere besides just my internal hard drive.
Come join me in my paranoia. I wouldn’t say it’s “fun” here, but it’s safe. Ish.
Yeah, and if the .zip backups bother you, you can switch that off in the preferences as well. It will just save the backups as regular project folders—naturally we have this off by default as it’s a more advanced state of operation, it requires you to be careful and not open/edit the backups directly.
I also use the datestamp option, and have my limit set to 25. In addition to that I have my backup folders all located on Dropbox, so that every computer I own has every copy copied redundantly. But that is getting off topic.
Your idea of allowing Scrivener to open a Zip file in an extraction mode isn’t a bad one on the surface. Personally it would never cross my mind to consider opening a .zip file in a program that isn’t a .zip file management tool. That just seems odd to me, but I suppose I see the root cause of such an inclination. I’ve used a number of multi-file format applications over the years, so I’m kind of used to the idea of a “thing” that the program works with, being lots of files in a structured folder. It isn’t terribly common though.
Anyway, all of this trivia aside, the important question, were you able to recover the data?
Yesterday with v1.2.1 I watched as eight of nine long chapters “vanished” in the Draft folder; their file icons went blank as I clicked to open each one, and there was no text in the editor window. A string of seven files in a “notes” folder in the Research folder did likewise … their icons blanked, and the editor window was empty. Only one file in each folder retained their contents.
Previously, I had broken two large note files into chunks and drug them into new, smaller files in my notes folder. Then I drug that folder out of the draft group and dropped it into the research folder. I closed the project, did some other busywork, then later reopened the Scriv project. The scriv folder/package was in Dropbox. There should have been enough time for synch & update to occur.
Using win explorer, I opened the corrupted project folder (on a Mac its a single file called a ‘package’) to examine the contents. Both the draft folder chapters and the notes folder files were actually missing. Only one .rtf file survived in each folder.
I turned off Scrivener’s backup function and opened that last available (datestamped) backup file. I keep ten date-stamped, auto-zipped backups in a “Scrivner BU” folder in Dropbox. Everything in the last backup file was normal, all contents intact.
Lessons learned: I had just recently decided to put the working scriv project in Dropbox for convenience between my other machines (I won’t do that again). I have never had a problem with zip backups auto-saving into Dropbox (I will continue doing that).
The built-in zip facility in Windows works well for opening the zip-compressed backup and extracting the project file. Moving zip files into and out of Dropbox seems to be a safe approach.
I am baffled by the disappeared files.There was no other clue, and the backup saved the project. First time in three years I’ve had a project go corrupt. Thank heavens for backups. Maybe … I should increase that backup count to fifteen …
It didn’t occur to me either until I found that XBMC could handle zip and rar files; VLC can also handle zip and ISO files so, while it’s not terribly common, it’s not unknown. Also, as the default backup behaviour in Scrivener actually uses (writes) zip files, I think providing the complimentary read function has an elegance that fits right in with the whole “why doesn’t it just work” agenda.
Yeah, and another one is the .epub format, which is just a .zip file with the extension changed to .epub. Handy trick if you ever want to examine the guts of an e-book is to change the extension to .zip and decompress it.
We’ll give it some thought. The main thing to avoid is people editing their backups directly. A “Save this where?” type set-up would avoid that—though it might not help with the ultimate problem of alleviating confusion (or to put it another way, your suggestion requires another type of technical competency so it might just be swapping one problem for another). But yes, making this a little less file and folder based is something on our list of things to think about, already. Obviously, the most seamless solution would be a Snapshot-esque roll-back window for the project, where the filesystem was completely abstracted. Unfortunately that isn’t terribly feasible though without sacrificing file system flexibility for those that do know what a folder is and like being in control of it.
In an OCD-ish attempt to put things in the right place, I’ve also posted the suggestion that Scrivener should include functionality to read .zip files in the Wish List forum:
I think I can anticipate the other problems that might arise from presenting the hapless user with two or more instances of a project, and suggested in the Wish List post how further problems and confusion might be alleviated:
Anyway, once you find the file, drag and drop it from the backup binder to the working project’s binder to bring it back in.