Save or Save a Copy?

[EditorOne]

Originally Posted by JSWolf

Yes, email programs and folders.. Perfect use for the term folders. But when it actually is a filesystem directory, then call it a directory because that is what it is.

Some good examples...

Calibre stores each eBook in a separate directory.
Collections on your Reader are folders

Thanks. I've been properly schooled now. I guess because I began computing with an Apple II, I can't lose the old terminology. Plus, I continue to see a directory as a "container" for files, much like the paper folders in my desk drawers. Now, if only someone would apply some hierarchical structure to my desk . . . .

[pypo]

After reading this thread, it is apparent that Calibre is lacking a decent workflow and documentation for people who must edit and save versions of their ebook. I'm talking about authors. This is currently very unsophisticated. Calibre should think along the lines of a repository, where everything changed in the repository is clear and committed. You can also revert changes. The problem now is that the writer must fool around with the file system and guess where things are and what they might be called, at the same time being tricked by complied versions that come up during preview or reading a book, but which don't match the most recent edits. Also, I have saved versions and those saves cannot be found. I have also saved versions and those versions are found, but upon reading the book the older versions (but uneditable versions nevertheless) are coming up in the viewer. Now I'm sure with this I will decipher this mess of file structure, saves, versions and so on. But it is not easy and it involves a lot of time-wasting guesswork and trial and error. On the other hand, Calibre could go for Full-Idiot mode like Apple has done with Garage Band, so that everthing is saved and you have no idea anymore what the relationship is between the application and the file system.

[eschwartz]

Currently there is little demand for repository-style control over ebooks.

My $0.02 -- save an original in case of disaster and do edits. Once edits are confirmed working and in good order, delete backups.

The editor will be most likely used for typography/layout after a final draft is submitted -- no need for authors to keep copies of drafts and ideas in a code editor.

[pypo]

Okay. So I create a directory called "backups". And I just "save a copy" versions under whatever (name-version) with .epub extension and anytime in the future can just open that up and continue. Fine. Makes sense.

so...

1. "epub" is the Calibre default file format

2. "save a copy" is how I make versions

3. "save" saves everything in the current open epub, including CSS

4. When Calibre says you are removing a book from the library, and warns you that the book will be deleted, this is simply not true. The book is still very much on the hard drive. You just have to know where to find it. Calibre does not delete the book from the hard drive in this case.

[eschwartz]

when calibre says you are removing a book from the library, and the book will be deleted, it is absolutely true.

The existence of a backup is entirely irrelevant. For instance, I keep rotating timestamped hardlinked backups of important data via rsync, a common cli alternative to Apple's Time Machine feature -- does that mean if I delete a file the program is lying when it warns me the file is deleted?

Note also, calibre usually deletes to the Recycle Bin anyway.

[BetterRed]

pypo - FWIW - if I need to do progressive saves that map to workflow steps, particular if its likely to be over a period of hours/days, then I don't edit in the calibre library folder, I edit in a separate editing folder. I only edit in the library folder if its a low-risk, quick-fix.

BR