Duplication: Eratic, Flakey and Buggy

Scrivener Scrivener for Windows
1

I am finding this latest version of Scrivener for PC very flakey and buggy when it comes to duplicating folders.

Even when I right click on a folder and click on “Duplicate”, the results are not predictable.

The folder “appears” to have duplicated ok. But closer examination of the file reveals that old images have been used in various places.

I noticed this because I am working on a book and happen to use a pen name image. At the last minute I decided to use a different name, and so was surprised to see that when I duplicated the file over it was using the old image (with a different name!)

Having used the file (assuming it was a duplicate) in Smashwords, I am now going to have to spend the rest of the day remedying files in various places.

This needs to be urgently fixed.

Also the way this thing compiles .mobi files is all over the place. I optimized a file in Photoshop and then brought it back in (300kb smaller), and the file size went up!

The .mobi files that the windows version produces are VERY bloated.

My book file is 2300kb when saved as a .mobi from Scrivener.

If however I import that same .mobi file into Calibre and then convert it to .mobi (sounds odd I know, but Calibre must run its own conversion algorithms); then it will convert it into a 500kb .mobi file that looks the same.

So, presumably, Calibre is able to cut down the file size by over 75%, simply by removing the bloat that Scrivener is saving with it.

Can you have a look at what Calibre is doing to produce .mobi files that are so much smaller?

I have also noticed that the saving is eratic. it doesn’t always save the latest version.

This program is very buggy for me, and I am starting to lose faith in using it, because I am never sure what the final outputted version of the file will be. The final one that I “think” I am using, or some other “older” version that Scrivener randomly pulls up.

2

That duplication problem sounds pretty nasty, do you have a set of steps I could follow along with to see it? By example, here are the steps that I tried:

  1. I created a new blank project

  2. Clicked on the Research folder

  3. Pressed Ctrl-Shift-N to make a new folder; called it “Test Folder”

  4. Pressed Ctrl-RightArrow to indent the folder beneath Research

  5. Pressed Enter to create a new text file and called it “Test Text”

  6. From Windows Explorer (just using Windows sample pictures for simplicity), dragged int “Jellyfish.jpg” into the “Test Folder”.

  7. Click on Jellyfish in the binder to verify it had imported correctly

  8. Pressed Ctrl-F5 to open the file in Paint (my default for JPEGs in this test account)

  9. In Paint I just scribbled a bit on it and pressed Ctrl-S to save and Alt-F4 to cose. I also verify in the title bar that the picture’s internal name is “7.JPG”.

  10. Back in Scrivener I use the UpArrow/DownArrow sequence to select another file and then come back to the picture, I verify it has been defaced.

    All right, at this point I presume I have something similar to the setup you describe. A file that you have in a folder, that you’ve since edited to change the name.

  11. I select “Test Folder” in the Binder, right-click, and choose Duplicate. I name it “Test Folder 2”

  12. In the Corkboard I double-click on the picture’s icon to open it in the editor and I verify that it has the defaced version of the Jellyfish picture as well. I rename this “Jellyfish 2” to avoid confusion and press Ctrl-5 to view it in Paint. I do this strictly to note down that its internal name is 9.JPG and not 7.JPG. So I close it and go back to Scrivener.

  13. I press the History Back arrow button three times to get back to the original “Jellyfish” picture. It remains the defaced version. I press Ctrl-F5 and verify that it is 7.JPG and not 9.JPG. I draw some new squiggles on it using a different colour.

  14. I do the UpArrowDownArrow dance again to verify the green squiggles are now on “Jellyfish” (7.JPG)

  15. Press Ctrl-] three times to get back to Jellyfish 2 (9.JPG) and verify it has no green squiggles. Ctrl-F5 confirms in Paint that I’m looking at a unique file.

But that is only one interpretation for your description. You may be doing something else entirely (and probably were not doing as many double-checks as I was). If you can provide a similarly detailed list that results in duplicate linking two binder items to 7.JPG that would be very much appreciated and would go a long way toward helping us solve the problem.

If you examine your project folders with Windows Explorer, do you see a lot of conflicted files? If so the project has been or currently is sitting in an automatically synchronised location and has gone out of sync with itself. This can happen fairly easily, but it’s also extremely easy to avoid it happening. I would encourage you to read the “Scrivener Everywhere” chapter in the user manual in chapter 13 for tips. You can disregard that if you do not use Dropbox or something in that vein.

The .mobi thing is easy to answer. Scrivener doesn’t actually have much to do with the creation of the .mobi file. What it provides to Amazon’s KindleGen software is quite similar to the .epub output it creates, which you can verify should be size-wise more what you expect. KindleGen takes these sources files and is responsible for creating the .mobi file from start to finish, and at this point in history, it currently producing two e-book files in one container. One version is the KF8 format that Amazon will be gradually transitioning all of the devices to over the years and will be much better for authors in general. The second version is the old legacy .mobi data which is compatible with older Kindle devices. Thus the file can be opened on a first generation Kindle or a Kindle Fire and look good on either one. If you use Calibre to generate the .mobi file, all you are producing is the old legacy .mobi data, which at this point in time can still be read (though at the cost of quality in some cases) on newer Kindles, but this will most likely not always be the case.

So it’s your choice, just know that if you’re using Calibre instead of KindleGen to create your e-book you aren’t using Amazon’s official tool for doing so, and your book may not display well or at all on future Kindles, depending upon where the dice roll.