Doing the unthinkable

I’m going to do the unthinkable. After 14 years with Scrivener, I’m going to move my stories into Ulysses. This decision is a journey rather than a snap decision, and certainly one that won’t conclude in a week. I’ll follow up on my progress in the coming weeks (and months), but for now, I wanted to share my reasons for making this seismic shift in my creative process.

Halfway through 2018, I began using Ulysses, to streamline my blogging process while being drawn to its outstanding design that uniquely married features with a focussed writing experience. For blogging, Ulysses met my exceptions and more. I loved the library — keeping all my writing in one easy-to-manage location. Textbundle export took some of the pain from static-site authoring. It was the move from Pelican to Ghost, which improved my publishing process three-fold with the ability to publish directly to my blog with a click.

Use breeds familiarity and familiarity breeds comfort. I became comfortable and proficient using Ulysses after six days as I was using Scrivener after 12 years and three major releases. The more I used Ulysses and came to appreciate its focussed minimalism, the more Scrivener began to grate on me…slowly but surely.

Scrivener’s UI is dated and stuffed with distractions. There are large parts of the app’s feature set that I never bother with; such is the behemoth it has become. Ironically, Scrivener’s appeal to me in the early years was that it wasn’t as bloated as Microsoft Word, but that’s no longer true. Scrivener tries to be all things, to all writers, not just novelists but screenwriters, academics, and technical writers. It’s a tremendous achievement, but it’s come at a cost. Scrivener’s a malleable beast, but one you must tame with a lot of configuration, the willingness to learn, and the discipline not to turn your manuscript into a tangled mess.

After adopting Ulysses for blogging, I quietly began using it for short stories. I loved the quiet minimalism and the ease with which I could export the story for my website or newsletter. Yet, I continued to use Scrivener for writing my books — big ones, epic fantasy novels that are part of a six-book series. I defended Scrivener loudly and often to myself and others, citing the app’s organisational power and lauding its feature set — and complexity. I even started a tutorial series to help other authors flatten the learning curve — a series I wrote ironically in Ulysses.

My persistence with Scrivener, and recommendation to others, was out of the belief I needed Scrivener as a novelist. I told myself I needed the Binder, I needed the Outline mode and custom metadata, I needed its frustrating and complicated compile system.

In reality, I didn’t need any of it — or at least not how Scrivener would have me work.

The crunch came last weekend when I spent two solid days restructuring my latest fantasy novel (and its place in my series). It’s already 130,000 words in size — and that’s only 2/3rds of the planned story. Had I stayed on my trajectory, I wan on track to hit 180-200k, but I knew there were severe structural flaws that would make revision a laborious and time-consuming slog. So, I needed to act now.

That I reached an ‘oh shit’ moment where I needed to right the ship, is of course, on me as a writer. Yet, it’s telling that I did not restructure my novel in Scrivener, at least not immediately. My approach was to position all the pieces of my plot on the expandable canvas of a light-weight mind mapping app (Scapple). When I had a bird’s eye view of the structure and fixed the problems, I wrote plot-points out, row-by-row, in a spreadsheet (Apple Numbers). There was a lot of back and forth, working between the two 27inch monitors I have hooked up to my MacBook Pro, with each app doing what it did best.

The problem became apparent when I went to recreate this structure in Scrivener. I converted the spreadsheet columns into custom metadata fields so I could display them in Scrivener’s Outliner. Only, as I started doing the process, I realised that not only was I duplicating the effort, but I was losing something in the process. In Numbers, I had two tables on the page to track my story’s core narrative conflict, and the other followed my subplots. I also generated pie charts showing a breakdown of various elements of my book’s structure, including POV characters and the relative size of each story arc in the book. You can’t do this with Scrivener’s outline view, and I wouldn’t expect to. Scrivener’s outliner behaves more like OmniOutliner, not like a data-crunching spreadsheet.

It occurred to me that I had found a better way of working, or at least one better suited to my mindset -- and therein lies the source of my disquiet.

Scrivener is an integrated writing environment modelled on the IDEs programmers and web developers use to develop apps. But for each integrated feature of Scrivener, there are better options out there. Ulysses is better for drafting and syncing. OmniOutliner or spreadsheet is better for outlining. DevonThink is better for managing research. AeonTimeline is better for timelines. Vellum is better for creating ebooks and paperbacks. Scrivener’s appeal isn’t that it is the best at those things, but you can do them in one window.

And yet, despite this kitchen sink approach and convenience, I no longer work in a way that plays to Scrivener’s strengths. Part of this is born out of a desire to work how I want, not how Scrivener tells me I should. Scrivener doesn’t allow me to break the app across different windows. I cannot build custom outliner views or derive characteristics from the data like with a spreadsheet. Despite its organisation power, Scrivener doesn’t like you including an entire series in the Binder, preferring the one-book-per-project approach. Yes, you can do it, but once you compile a single volume from the series, your headaches compound with the technical debt of a system that was never meant to work this way. I could name a dozen more niggles that have rubbed me the wrong over the years until I can no longer ignore the blister it’s created.

So, as I realise that Scrivener’s organisational and compiling features no longer serve my needs, what does that leave? For me, that would be drafting and world-building. I’ve already admitted I prefer Ulysses to draft, but what about world-building? Scrivener’s research collection abilities are quite formidable. Coupled with templates and you have a powerful world-building solution.

Yet, for reasons I can’t quite articulate, I’ve never liked Scrivener for world-building. It’s probably down to my preference for separating concerns, along with the tedium of writing structured documents using rich text. I’ve done all my world-building in markdown. I use a static site generator to turn them into a website or PDF document on the rare moments I feel inclined. These are practices I’ve introduced to my creative writing from professional life as a technical writer. But that’s a post for another day.

Concluding thoughts

I’ve used Scrivener for over a decade and written well over a million words in the app. It was the reason I first migrated to the Mac, and it was the perfect antidote for the horrors of writing novels in Microsoft Office 97 and 2003. I am enormously grateful for what it has given me.

Since 2006 I have changed, my writing has changed, and what I need from my tools has changed accordingly. For many reasons, personal and professional, I find more than even I have to shift contexts; Scrivener hasn’t kept pace with my needs.

I’ve come to appreciate that Ulysses is much more convenient than Scrivener. Ulysses outperforms Scrivener in performance, synchronisation, feature parity between macOS and iOS, and document export. Most important, Ulysses’ drafting experience is sublime. What Ulysses lacks is easily added by augmenting my workflow with other apps I like that works in ways that suit my creative processes better.

Time and experience will tell if I’m making a mistake. Perhaps Scrivener’s integrated approach is better for the laborious and mentally draining task of growing a novel. Nevertheless, I owe it to myself to try, confident in the knowledge Scrivener isn’t going anywhere. Regardless of the outcome, I’ll chronicle my journey here — be it a resounding success or an abject failure on my part.