Ungroup in Corkboard: Retain Nesting

Dear Keith,

Ive been working through your fantastic Interactive Tute for the first time, that’s a real work of art!

I’m really interested in the corkboard options, and the tute revealed lots of exciting possibilities. One thing that puzzled me and which I think, if changed, could lead to Scrivener being adopted by professionals in the industry ( :wink: that was a very trenchant observation of yours!!)…

…is the undoing of folder/doc nesting when one uses “Ungroup”. I have some usage cases where retaining the nesting is desirable (especially: mapping an argument) and can imagine others where it would also be useful (e.g. a family tree of characters; basic visualisation of folder/doc structure).

And I find it hard to see the value in the nesting being dropped. It seems equivalent to undoing the nesting when you expand an outline.

The stacked corkboard comes closest to what I would like to see but it does not arrange nested items in the order which is visually most logical (for example, that represented by the Outliner).

All the best


I don’t think I quite understand what you’re after…“ungroup” is essentially an undoing of the “group” command, so the whole purpose is un-nesting. I’m tired, so I’m probably just misunderstanding. But would “Open>With All Subdocuments” be helpful to what you’re trying to do?

Oh, also collections, if you haven’t checked those out yet. It won’t have the hierarchy, if you’re trying to see that (but then outliner really is a much better place to take in hierarchy than the corkboard–as the metaphor suggests), but you can easily pull in documents from all over and view them all together on the corkboard; in freeform mode you can even group these however you like.

Yes, I don’t understand this request either - the whole point of “Ungroup” is to get rid of the nesting. :slight_smile:

Thanks MM for taking the time to respond in your tired state!

I agree that what Im after isnt really ‘ungrouping’ because as you say that presumes some loss of order, and perhaps that comes in handy sometimes. I guess what I’m after are ‘Expand’ and ‘Collapse’ commands for freeform, nested corkboard cards.

What I would like is to be able to see is something like that shown in the picture below:

That’s actually two screenshots pasted together, because as far as I know, you cannot achieve simultaneous expanded views of the binder and corkboard. But in this fictional case the corkboard view represents in a visually logical fashion the nested structure of the Binder items.

So basically I would like to be able to apply two toggling commands ‘Expand’ and ‘Collapse’ upon the index card titled ‘Introduction’, and doing so would not affect the underlying relationships (nested and sequential) of the cards, nor the Binder docs.

Having played around with the system a bit more now, I realise that my wishes are more complicated than I had originally thought. My planned use of this feature would also require that ‘Commit Freeform Order’ would not only revise the sequence of the Binder items (as it already does) but also their nesting relations (which it does not).

So its been good fun to think about, but don’t know if it’s achievable!




Yes, this would be a very complicated thing to achieve, and would really require a completely different corkboard. I don’t think it would work too well either. Although it looks okay-ish in your mock-up, imagine how much scrolling around you’d have to do if you had ten or twenty cards inside any particular folder. You are - if I understand you correctly - essentially asking for a corkboard mode that acts like the outliner. But that, of course, is the whole point of the outliner. :slight_smile:

All the best,

1 Like

Hi Keith,

Yes that’s exactly what I was thinking about, but a more visual version of an outline, i.e. a map, like, say, the hierarchical maps used in Freemind. Such a map and an outline have identical logical structures, but of course, many people find maps more intuitive to work with. Their interpretative ease does come, as you point out, with the cost of using up more space, but that’s a compromise that is often worth it.

But of course Scrivener is not meant to be a mapping program. Currrently I’m having a lot of fun talking to the Freeplane people (Freeplane has sort of taken over from Freemind), trying to encourage a few Scrivenerish features in that project, and meanwhile Im also seeing what mapping possibilities exist in Scrivener. The latest development version of Freeplane is pretty neat actually, but don’t worry --its never going to be a threat to Scriv! :slight_smile:



I’d also like to point out that, while it isn’t automatic, you can set up corkboards from items of many different layers in the outline, if you need it. A collection is always a flat list, and a collection can be viewed as a freeform corkboard—thus you can accomplish exactly what your above mock-up proposes (sans the auto-colour), and a achieve a form of “mapping” power.

Hi Ioa

Many thanks, that’s brilliant. :smiley: Using a collection’s flat view does solve the basic problem since the arrangement of cards is retained even if you shift focus in and out of the collection. Below I describe the way I see using this now, as the argument mapping component of a larger template for research paper writing. I would greatly appreciate if you could check it over to see if there is any flaw (or improvements):

  1. The Research Paper template would have a collection - “Argument Map Template” - that would have five differently coloured and titled cards [Claim, Reason, Objection, Co-premise (reason), Co-premise (objection)]. These cards would sit at the top of their corkboard.

  2. When one decides to map an argument one first duplicates that collection (ideally that collection would be uneditable, but not so for any duplicate of it).

  3. In creating a new map within the freeform corkboard of the newly created collection, one would build it stepwise by duplicating the appropriate type of the five existing cards, to add in the required type of argumentative element. Then one adds in the text of the claim, objection etc to the main body of the card. The pre-existing set of five cards would thus form an “infinite” pool of argumentative elements to draw upon.

  4. When one want to write up that argument, one just uses the Split Window feature to have a view of the map and an adjoining text-editing window.

Having played around with the system this would all seem possible.

Many thanks for your time (and genius)


Actually having played around with this a bit more, I think I can get all the functionality I described above with an ordinary folder. As long as there is no nesting of the items in the binder view, the cards remain where they are as I switch focus. And even if I turn freeform corkboard off for another folder, this does not affect the corkboard setting of my folder of interest.

So am I missing something that only the Collection approach provides? Otherwise it adds a bit of complexity that would probably be good to avoid (especially since my approach involves creating new items that are not actually in the Binder).



Yes, as I was reading through your first message, my first thought was: why not just use folders? So it seems we are on the same track with that.

Collections are good when a group of items have a secondary+ logical grouping, they are not quite so well suited for a group of items that have a primary logical grouping. This isn’t to say the feature just doesn’t work, but that there is already a fantastic system in place for primary logical groupings: containers in the binder. Containers will be in general superior to collections because they can be more articulate about themselves—they have meta-data and hierarchy—and their constructs can be easily reproduced en masse, turning them into replicators as well with the use of Document Templates (more on that below). Meanwhile, collections cannot be duplicated. Sure, you can create a second collection from a first collection by selecting its contents and pressing the [b]+[/b] button, but any mapping will be lost because map data is stored in the container, not in each of the individual items (which given how Scrivener works, makes sense, otherwise you would have items declaring themselves to be at such and such X and Y coordinates no matter if they were in their primary folder corkboard or a collection corkboard). Containers also have an advantage over collections in this case, in that they themselves are entities which can exist in other containers and collections, and thus indirectly can express their children in these scenarios in a navigable fashion.

Collections become powerful when your cards have secondary simple-logical groupings which (potentially, but not necessarily) transcend their primary groupings. If you want to gather all “Claim” labelled cards into a single pool, to analyse a sequence of chained arguments, a collection would be great for this, you could even drag in supporting material and arrange it on the map as well and find elusive connections between the arguments that you might otherwise have missed. Collections are potent when it comes to this sort of associative usage. It’s as primary organisers where they are a bit more limited.

All of the items you listed together would, to my mind at least, be defined together in the same primary logical grouping (the argument), and would thus be good together in a single container at the same level (and this argument container would itself live well as something that can exist on other, higher examination corkboards). They can and should most certainly benefit from a greater articulation of their information with children, but these children are subordinate and not entirely relevant to the top-level. For example, a set of treatise beneath “Reason”, further adding supporting data to that particular node. That “Reason” can itself be a map of its own supporting data is quite powerful, and this type of integration between granularity in thought would be lost in a collection were it created there instead of in an outline.

So, tip one: Create your template argument map as a container with all of the necessary skeletal structure beneath it; perhaps with prompts and so on in the Notes field, etc. Then place this container within a template folder (you’ll need to create a new folder and designate it as a template folder in the Project menu). Now you get this entire structure in the Project menu, and you can easily create new copies of it wherever you need it. Selecting a container from the template menu will generate not only that container, but all items beneath it as well.

Tip two: reducing the navigation expense of hierarchal corkboards. Two shortcuts you’ll want to familiarise yourself with are [b]Ctrl-Cmd-R[/b] and [b]Opt-Cmd-O[/b]. The former will traverse up the hierarchal structure and the latter will traverse down, based on the current card selection. Meanwhile [b]Shift-Cmd-O[/b] will open that card in your split window (potentially as another corkboard if it has children and you have “Treat all documents with subdocuments as folders” enabled in Navigation preferences). With the mouse, you can set up the corkboard to react to a background double-click in the same manner as pressing [b]Ctrl-Cmd-R[/b]; you’ll find this option in the Corkboard preference pane (and also note in that section, you can enable drop ons, which could be useful for your workflow), and you can double-click on card icons to descend into them, or drag them to the other split’s header to open them there. So either preference, keyboard or mouse, is well supported.

Tip three: freeform/standard is indeed container based, as you note, and you can also flip between modes on the same container without destroying any mapping.

Dear Ioa

Many thanks for all those great suggestions, lots of possibilities to explore. The reason I overlooked containers originally was that I was trying them out with nested documents (which didn’t work of course) and only later realised that this was not critical for my purposes.