Best practices for speeding up lag & slow typing in 2-million-word Scrivener 3 project?

I’m in the process of converting many projects from Scrivener 2, which I adore, to Scrivener 3, which is even MORE awesome, with even BETTER features.

However, with one large Scrivener 3 project I’m having major lag problems – especially typing lag. I don’t know what’s causing the problem or how to troubleshoot it.

I created the project by merging what were 22 separate volumes / projects in Scrivener 2 into one Scrivener 3 project. I separately converted the Scrivener 2 projects, set up my new project, imported tags and labels into it, then dragged the contents of each volume into a folder in the new project’s Binder. The resulting Scrivener 3 project totals 2 million words and 93 MB.

I merged the projects in order to (a) sync tags and labels across volumes, and (b) test whether Scrivener 3 is fast enough I can keep 22 volumes in one project. (In a 2015 test, I merged 10 volumes / 1 million words. The result was far too slow to use, hence the separate projects.)

At first, the 22-volume Scrivener 3 project worked great and was nice and fast. Being able to use direct document links rather than project links as I update and cross-reference the volumes is immensely faster and less error-prone than keeping 22 projects synced. But…

The Scrivener 3 project soon developed lag problems, especially typing lag. The lagginess seems to be progressive: The longer I work on the file without closing it, the slower it gets. Typing can get VERY slow. Sometimes it takes 1 second or more for text I have typed to appear on the screen. Since I do a LOT of writing in this project (roughly 100,000 words per year), lag is a huge problem. Especially because lagging often keeps my text-expansion program (Typinator) from expanding shortcuts, or mangles its results.

I suspect the cause is in the project itself, since my smaller Scrivener 3 projects don’t get slow even if I work on them for hours. Also, Scrivener 2 handles a 4-million-word, 81MB project with negligible lag. (That project is almost entirely text.) Rebuilding the Seach index, or quitting and restarting Scrivener 3, temporarily gets rid of the lag… but it soon returns.

Lagginess occurs whether or not I do much typing in the project. If I mainly move files in the Binder, update keywords in the keywords list, apply keywords to documents, do Searches, or apply labels, the project still develops lag. It might develop faster if I type a lot, but I’m unclear on that.

The project’s contents are mostly text, with a few illustrations, PDFs, and web pages. Chapters are mostly divided into short sections of text. They still use Scrivener 2 formats; I have not yet converted to Scrivener 3 styles. Checking the project in macOS, “Show package contents” found 11,500 RTF files inside the Files > Data folder, with a further 923 inside Snapshots folder. The project contains around 2950 keywords.

What do you recommend as best practices to (a) troubleshoot and fix this large project, and (b) keep large projects fast and responsive?


Mac Pro (mid 2010) with 32 GB RAM
macOS 10.13.6 High Sierra
Scrivener 3.3.1

The number of words in a project rarely matters, but any of these very well may:

  1. the number of words in a document
  2. the number and size of images embedded in a document
  3. the number of documents in a scrivenings
  4. the number of words in a scrivenings
  5. the total size of the project

Of those, #3 and #4 depend on the way you use scrivenings, independent of the project’s content, and #5 depends mostly on images and other source material (such as web pages, etc).

Some of this can depend on the number of documents in a scrivenings in a non-linear fashion, I think (twice as many taking 4 times as long to react, for instance).

I doubt it’s a good idea to include this much in a single project, even if performance were good. You’re risking too many eggs in a single basket.

You shouldn’t have to do anything special.

The size of a project does not matter if you have backups. However, you should really do that. A project should be stored on different hard drives. At least that’s how I do it.

In my experience Scrivener handles very large projects (2 GB) well. Very large documents may have small delays when writing, as @drmajorbob says. And the project search can take a little longer.

Your project doesn’t seem very big to me. I’m afraid I don’t know why it’s still causing problems.

1 Like

Don’t know if this is relevant to you, but I have found using Scrivenings mode on a large number of (even small) documents that each contain a substantial number of inspector comments causes typing lag; I assume it is something to do with redrawing the screen with the comment markers, as using footnotes instead with the footnote marker set in Project Settings cured it.



drmajorbob, thank you for your help! That gives me a list of things to check.

The problem occurs typing in short, 1-document Scrivenings with no illustrations and no other documents open. I also routinely open 1-chapter and 1-volume Scrivenings (volumes are 50,000 to 180,000 words). Although a few things take slightly longer, typing doesn’t seem to be one of them.

Typing slowdowns also happen outside the Editor, for example when typing Keyword names.

Document safety: My partner is a former IT guy, so I’ve got the niftiest multi-disk backup system you ever did see. Daily backups (which include Scrivener’s automatic save files); weekly disk changes; the works. So not a lot of worries there. (Also, in my experience Scrivener is outstandingly reliable. I started using Scrivener 1 in 2007, and through millions of words written, I don’t think I’ve EVER lost data. It’s amazing!)

fto, thank you. 2 GB projects!! That’s good to know. Keeping all these volumes in one project will make managing them immensely easier as I create and cross-reference more material. I’ll treat this as a project-specific problem.

Backups are well-handled – see my reply to drmajorbob above. Also, I do remote backups.

xiamenese thank you, I’ll have a look. That whole class of problems had not occurred to me.