Scrivener as a project tool for non-writing projects: version 2 is good enough, version 3 is overkill

Thanks for posting your feedback, even if it isn’t the kind of news we’d like to hear, it’s good to know how you found the update. Firstly, if Scrivener 2 works well for you, I hope you find it useful for many years to come. Something you should be aware of, in terms of system updates around it, as that you will probably want to halt upgrading macOS at 10.13 for as long as you continue to use v2. That’s the last OS from Apple that will support 32-bit software.

So to respond to the main bulk of your argument, I hope you don’t take these as criticisms of your feedback, but more my own inability to understand where you are coming from—precisely because I use Scrivener heavily has a notepad type program, and much of my feedback into the shaping of version 3 was built around that motivation for using it in such a fashion. So naturally I am a bit biased there, but I find the many little improvements to the project window user interface to make notepad usage in v3 in entirely different league than v2.

To be clear, v2 for me was always more in the realm of it being a good notepad program in theory. It did work all right, and I did use it that way to some degree, but I never did fully migrate away from other more dedicated programs for that purpose (Notational Velocity, VoodooPad, etc.).

Once the early alpha builds for v3 were tuned and stable enough to use regularly, I never looked back, and for the last few years now I really only use Scrivener for that particular task. But, if I try to use v2, phew—it feels like I’m constantly running into roadblocks. :slight_smile:

So I’d like to have a bit of discussion over this aspect of the update, since it is something I have a bit of a passion about—this kind of software and how I feel Scrivener 3 dips into that role.

It doesn’t add any particularly wonderful features for organizing notes & brainstorming.

I’d be curious to know what that means to you. It’s difficult to take this as constructive criticism since it is vague, and from our point of view there are hundreds of new approaches in the software that make it more viable for this kind of work—so it feels a bit like one spent a long time writing a book about wrenches only to get feedback that the book doesn’t talk about wrenches. First reaction is: say what? :slight_smile:

Here are a few that come to mind:

  • The corkboard and outline now receive content drops from all sources.

    • I can select text from the web browser and drop it into a corkboard to capture some text into a new outline item.
    • Drop a JPEG onto a corkboard to import it as a file to that folder.
    • Drop a file from another project into an outliner to import it. This is really great for those kinds of idea-nexus projects where data may be coming in or out of it from other projects.
    • Etc. Experiment. The idea being if you think it should work, and it is technically possible for us to import or duplicate data via drag and drop, it should work.
  • Much of the content creation capabilities have been tuned to have a more content-centric aftereffect. For example if you hit ⌘N while typing in a document, the result will always be to bring you to the new note with the cursor blinking in the current context. In Scrivener 2 you were forced off into the binder, demanded a name for the new note right then and there (maybe in brainstorming you don’t even have a name in mind yet) and then had to manually get back into the editor. Now it’s the note that matters. You can name it later, and easily so even with binder closed with ⌃⌥⌘T.

  • And I think a tangent in that prior point is worth mention all on its own. I have maintained for many years now that one of the most important things a notetaking style program can do is keep out of one’s way while they are rapidly taking notes. The less interface between you and typing down your ideas, the better you’ll be able to jump from one thing to next, and then back again to whatever you were doing if need be. If you have to stop typing, reach for the mouse, click on a little button, assign something a title, give it some metadata, choose where it will be saved and finally click into an editor (maybe not necessarily all in that order)—well, by the time I’m done with all of that I’m already thinking about something else, and the main thing at the forefront of my mind is frustration over how much I have to do just to start typing already. I’ve used so-called notetaking programs where the above isn’t all that much of an exaggeration—and no matter how much I liked the rest of the program, that one point could spell doom for it in my workflow. I wouldn’t say Scrivener 2 was that bad, but there was a bit of friction, and we’ve worked hard to remove as much of that as possible:

    • Names no longer really matter, and sometimes not using a name can be more beneficial than naming a thing. One’s content can drive the name dynamically if it is left empty, either the synopsis or the main text in that order.
    • Synopses can be deferred if one prefers lead-in text to hand-crafted summaries. In the past you could generate these into static text (and bother with refreshing them if the lead-in text changes), but now you don’t have to bother. Index cards are immediately useful even if you haven’t touched the feature in years.

    Aside from the obvious reduction of UI friction as one moves from one note to the next, there are some impactful implications to this approach. With the ability to rapidly capture your notes and then have the content of your note generating meaningful identifying information about them in the various group views, your overhead in sorting and up-front cataloging is automatically diminished to only what you need to do—and there is no penalty for not doing so. Adaptively named items sort right into manually named items, they can be searched for by name, they can even have that handle compiled as an option.

    Without a penalty for heavy categorisation up front, the possibility of using the software less rigorously, more sloppily and getting similar or maybe even superior recollection results out of it in the years to come is achieved.

    Of course, one needn’t subscribe to chaos if they don’t want to. You still can easily name everything, and give everything a hand-crafted summary, and spend five minutes on custom metadata, keywords and other forms of organisation. That’s all still there (and I would say all improved in their own myriad hundred little ways as well).

  • You no longer have to Reveal in Binder, or manually find a thing in a corkboard or outliner view, to move or copy that item to a different location. All you need to do that is have any icon representing that item visible, even the icon in the header bar. This makes the process of filing things out of inbox or scratch pad style folders so much more effortless, and in return that means less friction toward using those kinds of central inbox folders in the first place. In Scrivener 2 I avoided those constructs because filing things took a lot of effort unless that was what I was doing—in terms of just sitting down for fifteen minutes and doing nothing but filing. Now, filing is an afterthought. I can type, and at the conclusion of having done so, I can say to myself, you know what, I want to save this beyond today, and drag its icon from the header bar into a folder in the sidebar.

    I’ve done that with just this very document. It started out like all messages I write on the forum and in tech support, in a notepad style Scrivener project. But once I realised I had a few things worth making note of for future conversations, I hit the shortcut to give it a name in the title bar, then dragged its icon over to the current quarterly folder, out of the disposable “Today” folder that gets cleared every morning of the thousands of works I write about Scrivener.

    In essence, clicking on a document so that it is shown in the editor is an automatic and implicit pinning of that item for future use in all drag and drop uses. If I have a document I want to cross-reference to in five or six other documents, I can just split the UI and visit those six documents in turn, dragging its icon from the other split into the editor to drop a link in. If I need both splits, I can drag the icon from the editor header bar into same editor header bar with the Option key held down to open it as a copyholder—and now both splits can be freely used to find content.

    As I put in the user manual: if you can see an icon on the screen, you can drag it. And if you drop it in a place that would have a logical outcome, like a list of Bookmarks, or a collection tab, or a binder folder, or within a document’s Notes field in the inspector, or in a corkboard, etc., it’ll probably do what you expect.

    This simple capability even comes in useful for routine things as well. Ever had the problem of needing to drag an item from one end of a very long binder to another? You can sit there with the mouse at the bottom of the sidebar for a minute waiting for it to scroll and hoping you don’t accidentally release the button somewhere random—or in v3 you can click on the item, inertially scroll through a thousand items in a second, and drag from the editor into the desired location. Done.

  • Navigation by title is more direct. If the type of note taking you do results in frequently looking up information by name, then the Quick Search ability is like Spotlight for your project. There is nothing quite like that in Scrivener 2, other than a weird approach using the index card finder tool and manually dragging and dropping search results into header bars. A far cry from hitting ⌃⌘G, typing in “name of no…” and punching return.

    Again this folds right into the earlier discussion on the reduction of mandatory routines for establishing identity for notes. If it is easier to jump straight to something by its name, then the necessity for excessively organising works into topical hierarchies can be scaled back to only what is necessary. And meanwhile a tool that treats an item without a formal name at the same level of priority as an item with a name again helps bolster that sloppy-is-as-good approach.

  • Corkboard and outliner views are searchable now. Hit ⌘F in an outliner to check it out. This means you no longer need Project Search to find an item in a large folder of notes that you’re already looking at, and it also means that if your project search only created a broad list of options that needs further narrowing down, the means to do so is more direct and less disruptive to the project window configuration than running sequential searches was in v2. Of course those methods still exist as well, and can thus be combined with outliner/corkboard filtering for even more power. For me, searching and navigation are integral to note taking.

  • The net effect of many of these adjustments I’ve been referring to above is more subtle: it means you can minimise how much user interface is going on in the project window, without losing core functionality. In the past, if you turned off the Binder and used a simple Notational Velocity style outliner + editor in horizontal orientation, you could do that fine for a while—but eventually you would have to open the binder to do certain things. It was unavoidable, and sometimes even the software would force you back into the binder, opening it for you as a result of some action taken. Thus collapsing the interface was a bit like putting the top down in a convertible car—glorious, until it starts raining. With the new design, I have notepad style projects that consist of a window about as complicated as VoodooPad’s window, or TextEdit for frame of reference if you aren’t familiar, and the project has been like that for years. I strictly and 100% use it as a wiki type hyperlink driven notepad. I don’t even really care what the binder hierarchy looks like. I suppose it is there if I ever need it.

  • And on the other side of the coin, there are some note taking projects of mine that are the opposite. They benefit from having lots of stuff going on at once. The new Copyholder system in conjunction with Bookmarks in the inspector in conjunction with the better integrated Quick Reference window (again, remember that icon is functional!), I can reference a huge amount of networked information at once.

  • Bookmarks record hyperlinks and bookmarks targeting an item. This was true in v2 as well, as References, but given what Bookmarks now provide to you, it means all items have a back-referencing listing in the sidebar where data can looked up, edited and copied and pasted out from the larger network of concepts surrounding the direct material being working on. One can acquire a broader understanding of the context of the ideas and notes as a result. The only program I’ve used that approaches this level of fuzzy awareness of the bigger picture is Tinderbox. DEVONthink is up there as well, but Scrivener is more like Tinderbox in that the fuzzy network is hand-crafted, while DTP’s is algorithmic for the most part.

    And the mechanism we use for that means less core navigation is necessary to dip into that fuzzy meta layer around content. You don’t have to go to the thing that cross-references to what you’re working on, it is right there and fully editable with a full-featured editor in the sidebar. You can go if you want to, just like you could with References in v2, but you don’t have to. The history feature can thus be more tightly bound to what you are working with directly as a side-effect.

  • And speaking of linking, it no longer requires up-front investment to jump between documents based on key phrases. If you right click on a noun for example, one that is discussed in depth in another binder item, right in the contextual menu you’ll be able to jump to that item as though it were hyperlinked. It’s just as easy to actually make a link right then and there if you want to.

  • If categorical and spatial note taking are two meaningful axes to how you work, then the new Arrange by Label feature is amazing. You might, when first looking at it, be understandably confused into thinking it is a tool for novelists—but really all it is is a dual axis system for assigning category and placement within a context. For getting your thoughts in order and at the same time thinking about those thoughts in terms of taxonomic properties, Scrivener blends the two tasks into one simple drag and drop and event. And hey, even if you don’t need a dual axis movement system, even just dragging cards around to assign labels is a bit easier than clicking on each card one by one and using the inspector to assign a label.

Well, one could go on :!:. Point being, while not all of the things above might be integral to how all people who would like to use Scrivener for note taking, I’m less trying to get across details here and more trying to paint a picture of the design philosophy behind this program as a note taking tool. The adjustments made from v2 are many and small. If you go hunting for them looking for branding-friendly marketable Wow! Biff! Bam! style feature bullet points, you might miss the subtle fact that we’ve spent years using this thing as an unstructured note taking environment, and in doing so, have built countless routes throughout its UI to facilitate that usage so that it is at the core already that way without requiring expensive and heavily structured features to do so.

Scrivener has always been that kind of program: where you are given lots of little ingredients and asked to combine them together into larger recipes. Scrivener 3 just adds many more ingredients, and makes those that were there a little more seamless and a little more integrated than they used to be—without heavily disturbing how they were used in v2.

A classic example of a good ingredient tool like that is Collections. You can use Collections much as you always have, if you did before. But where they have been refined in their usage? It’s now far easier and more intuitive to load a collection into the editor, for further work as a pseudo-text in Scrivenings, corkboard, filtering etc. Nowadays, you don’t even need to load up the collection in the binder (and thus have the binder open!) to do that: just right-click on the header bar and “Go To Collection”). What does that mean for note taking? If you used collections to automatically gather notes of a particular topic in the past, the overall UI investment toward viewing that list at will is dramatically lowered. You can even build keyboard shortcuts to open commonly accessed collection lists into the editor, now.

I went ahead & bought it a few days ago - mostly because I had never been disappointed by a Lit. & Latte product before - but having looked it over, I realize I will not be using it any time soon.

  1. I’d have to invest time learning how everything works again for those features I do want.

I don’t see how that can be fully understood in a few days, nor do I understand how Scrivener 3 would immediately block you from working as you did in Scrivener 2 in such a way that within hours you couldn’t figure it out and found the whole thing too confusing and too much of a burden to learn? At the entry level, it would help to know what it was you were doing in v2 that you now find so very difficult to do now.

We did of course strive to keep v3 feeling like v2 at a basic level, but obviously we missed a spot or two if your first impression was that everything had to be learned all over again! It would be good to know where those spots are.

I mean I guess there is compile—that’s the one feature that was given a more extensive surface layer of control (the guts are still much the same once you edit a format though)—but compile isn’t terribly salient to notetaking and brainstorming. So I need some help on what you’re referring to.

  1. I’d have to redo not just presets for styles, but the macros I built using Keyboard Maestro to trigger those presets.

I’m a bit like a broken record at this point, but again without knowing where you are coming from it’s hard to say what’s going on or whether you would even need to fully port over these old ways of working. For example, why would you need a macro to apply a preset in Scrivener 2? A simple system shortcut usually suffices—so you must have something far more complicated going on, something that cannot be easily adjusted via a menu call or something.

For example, to pull from historic posts of yours—you once wrote an Applescript to make it so different projects could use different default font settings for new notes. Of course, with Scrivener 2 you probably abandoned all of that complexity since the capability was built right into the software. Might it be that styles are doing things you’ve had to build a lot of custom macro tech to accomplish in the past?

The one thing that would have made it easier for me is if importing old presets had been possible.

Which presets were you trying to import? I can’t think of anything off of the top of my head that wouldn’t work outright. Even very complicated presets like compile settings can be imported, and some haven’t changed enough to even require importing and work in both programs equally, like preference presets, themes and script settings. I might have forgotten something I never make use of though.

Well whatever the case, if you don’t have the time to flesh out your critique that’s fine. We put a lot of time listening to feedback and pushing Scrivener’s feature set so that it works better for precisely the type stuff you use it for, so it would be nice to see where we might have missed things. If not, well again I hope v2 serves you well for years to come. I’ve even met people still using v1. There’s nothing wrong with that, and we are grateful for your support even if you aren’t finding it useful (hopefully yet ;)).

2 Likes