How do I get my list of substitutions from 1.9 to Scrivener 3?
I found, how and where to transfer the wordlists, but for the substitutions there doesn’t seem to be something similar.
that was discussed here and a method given.
https://forum.literatureandlatte.com/t/any-way-to-import-custom-dictionary-words-and-substitutions-from-scrivener-1/44246/6
I’ve decided to simply give up on the idea and assume substitutions will need to be re-written on upgrades, perhaps even from one beta to the next. This means that essentially, Scrivener does not have a substitutions function.
Perhaps after 3.1 comes out, they’ll reconsider the design decision to use the registry and switch to a file (best option), or write a backup routine for it.
The substitutions are saved in the registry. Designers alone know why.
I figured it out a few days ago. It requires messing with the registry, which can be intimidating, because you can screw up your computer if you do something wrong. However, if you follow the instructions exactly, and only mess with the Scrivener entries, you can transfer them.
See this thread: [url]Any way to import custom dictionary words and substitutions from Scrivener 1?], third post.
i saw those topics about messing with the registry.
That, i certainly won’t do.
That’s why I opened this new topic about only substitutions: hoping someone from the team comes along with a better answer.
I’m confident that there is a solution. Rewrite the whole thing ugh.Over the years I’ve got quite a lot of them.
It would be good enough (for me) , if there were a solution in the final 3.0, as I still use both Scriveners side by side.
I didn’t know that Scrivener for Windows used the registry for storing “operational stuff”. If I had, I would have been much more reluctant to buy it. I have a favorite photo management program, ACDSee, that uses (misuses, IMO) the registry in this way. It makes moving to a new PC much more complicated than it should be. I say this as a retired Teradata DBA whose career consisted of ensuring data integrity.
PLUS, it makes backups incomplete if your backup routine does not include the registry. Let’s say you regularly back up the Windows C:/Program Files (x86)/Scrivener folder, thinking that is where your parameters and operational ‘set up’ information is stored, but you don’t know that important parameters and data is ALSO stored in the registry.
If your registry information and your Scrivener backup information aren’t in sync, you run the risk of erratic behavior, that would be eliminated if the registry and the C:/Program Files (x86)/Scrivener were backed up at the same time.
My advice to the Windows developers at Scrivener would be to finish Version 3, as is, but to start plans for V3.5 or 4 to move that data out of the registry. Certainly, add nothing more like that to the registry. Once that stuff is out of there, I suspect you will see a significant reduction in support costs.
And to the users, I’d say start backing up your registry when you do a system back up. Well, that, and make sure you do a system back up now and then. When it comes to computers, a little paranoia is something of a survival skill.
I wonder if this is what’s happening with updating in place? Have to check at the next update.
I was a bit put off when I discovered the registry entry for substitutions.
I wonder why they put them in there to begin with. Must have been some reason. The registry is a good place for identity information and the like. While it can be used for data storage, I think it’s a bad idea.
A way to export the information from Substitutions/Autocorrections into an XML file (those work great for stuff like this), and then a way to import the information from one, would be an excellent addition. And if they can do that, why not simply store it in one of those, instead of in the registry?
I suspect the registry gets used because when you need a QUICK sequential lookup, the registry seems perfect. As a part of the basic operating system, it gets a higher priority than most higher level read routines. Plus, it is tested and maintained by Microsoft.
You don’t have to load the look-up values into an area of RAM, that you might want to use for something else. However, with system resources getting ever larger, ever more available, even on laptops, I think it is less necessary than it used to be.
I can’t speak for the Scrivener developers, their reasons may be perfectly valid and appropriate. But at first glance it is curious.