Thanks for sharing a bit of how you use the program. I always find that stuff interesting. I wouldn’t say that most people do not work in small chunks. It’s probably safe to say the split is roughly equal. There are many people who prefer to work in larger chunks because that is how they write—they don’t need anything smaller, or that is what they are used to, working in Word in chapter length files. Then there are those that have grown used to how adept Scrivener is at handling smaller pieces, or those that take a liking to it immediately like yourself. I myself write in very small pieces as well. Larger works can number way up in the hundreds, even approaching 1,000 pieces in the draft outline, many of them only a few paragraphs long.
I think the main difference is less that, and more how people assign meta-data and how they end up with small chunks in the first place. I’d say most people who are starting a new work in the software end up with pieces that don’t get split or merged at all. They build pieces at the size they need in the first place. Just to apply this to an example, when I build out a new section of a book, I tend to either create a series of cards for each topic I know I need to work on, knowing ahead of time that each card that I create is for one atomic piece of information. If I need more pieces, I make more cards. When I’m working this way I have no content yet, so splitting is wholly irrelevant—I just have bunches of empty cards. The other way I tend to work is when I’m actually writing without a pre-existing structure. Then organising takes the back seat. When I’m working that way, when I reach a point where I’m not longer discussing the topic at hand, I create a new card and continue writing. Either way, the concept of applying meta-data is particulate to the pieces I create, less so than being in batches. Batches do happen, but when they do I assign them after I’ve created a bunch of pieces.
Splitting implies that (a) you either already have a huge chunk of text you need to break down or (b) you don’t break into new cards as you write. I’d say the latter case is something a veteran user might run into more often if that is how their mind works. But give how pieces = atomic topics more often than not, the concept of having meta-data carry over is, for me anyway, more often an inconvenience. Why? Because the reason I broke off to a new card is because I’ve shifted to a new topic—hence, new meta-data. Of course, what you are describing sounds entirely different to what I’m describing. It’s a usage I did not anticipate when considering the importance of this feature. To me, carrying over meta-data is one of those very small conveniences, gravy if you will, not a core component of splitting, let alone of the whole software platform.
Another difference between the workflow you are describing and how I think most people use the software is the importance of the meta-data. I don’t think most people depend so heavily upon it because their structures are a lot more simple than what you are describing, which is almost more like a database. Most people are just putting down 25 chapter files in their Draft, or maybe breaking that out into scenes or sections as the case may be, applying a little meta-data to pieces as necessary. They aren’t using meta-data as a tracking system for origination, bread crumbs, and so on. That’s really cool what you were doing, wish it worked out for you.
And for that I do apologise! Like I said before, there are so many different topics in the manual, and it is literally impossible for me to completely audit the software constantly (keeping in mind I have to do this for both platforms!) so sometimes errors like this creep in.
Yeah, it seems to me you’ve taken the software in a very specific and specialised purpose and unfortunately without auditing whether or not is was working before getting hundreds of items down the road. That’s really frustrating for me because I try to avoid that situation when documenting the software.
Well, I hope you can find a way to salvage the problem with the bulk meta-data tools I described and maybe find a way to make the software work better for you. I’ve done what I can to elevate the issue to the developer, and I’ll be fixing the manual as well. That’s the most I can do. I’d appreciate if you are just too frustrated to continue working with the software though.
it is unfortunately the case for me 
