Hi,
I’m still using Scrivner v. 2, and am starting to aggressively set up Collections vis-a-vis Keyword searches. In general, the approach seems to serve my purpose. Forgive me if I’m asking as remedial or obvious question, but I’m wondering if there’s another way to see the full list of Collections that I’m setting up, apart from the Binder / tab view. For instance, is there a way to see the list of full list of Collections that I’ve created in the same way that one can view folders and documents in horizontal Binder view?
That tab view is the only way to work with the list of collections themselves. It’s one area of the design that I feel could use some improvement, but at the same time it does work for the majority of projects, and probably for most users at that. I really only have a few projects where there are over a dozen collections, where I find myself wishing I didn’t have to choose between The List and a reasonable view size below them.
I tend not to save Keyword searches though, since they are so easy to “create”. You have found the button in the floating Keywords panel for that, I presume?
Thanks for your reply. Well, I thought there might still be a way to create smart folders, or something like that, which would provide me with an alternate view. Any chance that might be possible? Just trying to find any kind of solution – even a hack that might provide another view, esp. as the number of my Collections (for Keyword searches) grows, thereby cluttering the tab view…
To clarify, I’m only saving Keyword searches as part of dynamic Collections…if that makes sense.
Well come to think of it there are the lists of collections in the main Navigate menu, which will switch the sidebar view to the selected collection. You can also right-click on an editor header bar and select a collection from there, which will load the contents into the main viewer.
I do understand what you mean about using search collections for keywords—I’m just saying that personally I would not create a collection specifically just to search for one keyword. It’s so easy to hit the shortcut for Project/Project Keywords, click on the keyword and then the magnifying glass, that the “cost” of having a tab in the list merely to not do that is too high, in my opinion. Especially if I’m working with keywords, because then I leave the panel open and it’s almost like having a “list of collections” in a floating panel, when you think about it that way. There are plenty of things I use search collections for, but keywords is not among them. Your mileage may very well vary.
Thanks for your message, though I’m a bit confused… I don’t see lists of collections in the main Navigate menu – does that exist in v. 2 or is it just available in v. 3?
I hear you about hitting the shortcut for Project/Project Keywords. But what if I want to compile (export) discrete documents that are tagged with Keywords? Is there another, better alternative?
Oops! I forgot you mentioned you were on version 2 still. The menu listing I refer to is in the View menu, instead, and the ability to jump straight to a collection listing from the editor header bar is v3-only.
Here is how I handle most one-off compiles like that:
Run your keyword search (or whatever; this approach is flexible).
Open Compile, and go to the Contents pane. Depending on whether you need a flat list or structured:
Flat list: set the main compile group dropdown to “Search Results”.
Structured list: leave the compile group setting alone, and enable the Filter option. Set the filter to “Include documents in collection: Search Results”.
The nice thing about this particular approach is that your compile settings are generalised. If you’re doing a batch of these, the way I understand you’ve been doing it thus far would be to tweak the compile settings every time, after having set up and created a series of collections. But with this approach you merely search for a keyword and then compile without adjusting any settings—and keep list bloat to a minimum.
Another really handy tool is the “Current Selection” compile pseudo-group (also available as a filter setting). That isn’t quite the right tool for this job, but it’s something to keep in mind when you’re doing stuff like this.
Ok, so that gives me a sense of how to compile these selections – when the time comes.
Forgive me, but I’m still trying to understand why you’re suggesting this approach to organizing / culling selections of text marked with Keywords as opposed to setting up Save Searched with Keywords using Collections… My understanding was that, according to the approach you’ve suggested, one can just summon Keyword-marked text selections using the Search function in the Keywords window, instead of having to set up / mange the clutter of Collections. Have I got that right? Thanks…
Also, one other question that I’ve raised in a previous post (so please forgive the redundancy)… For now, I’ve just been making text Splits and applying Keyword-marked text selections for one interview. But I’d like to import additional interviews with different people. So…how would you suggest that “mark” or “label” (in the broad sense – not necessarily using those tools!) Keyword-marked text selections so that they contain the subject’s name? I just want to ensure that my colleagues and I know who’s speaking in Keyword-organizated text passages. What approach would you suggest?
I mention this here because it also seems to relate to my earlier question about the best set up I’m using both to neatly view Keyword-marked text selections – as well as folders per interview subject, with each folder containing text Splits / Keyword-marked text selections.
Again, I’m totally open to any other suggested approaches! And I thank you again for all of your assistance…
BTW, I’d love to have Scrivner 3.0, but unfortunately I can’t update my Mac OS given the age of my laptop…
Right, and to an extent you can even almost think of the floating Keywords Panel as being a “list of search collections”. It’s not, of course, but since it requires of you less management overhead and roughly the same number of clicks to make regular use of, it is a conceptual substitute, if that makes sense. Furthermore since keywords can be organised into groups, you can effectively manage much larger lists-of-lists than you would be able to with Collections.
I’m not sure if I would advocate that approach for all things, but I am identifying a point where it would become more efficient than using search collections. If the latter exist only to mark the search of one keyword, and one finds themselves having to not only create and assign a keyword, but then search for and create a collection for that keyword, and is doing this potentially dozens of times—then it makes less and less sense to do all of that, and suffer the clutter of that procedure, than to simply drive one’s workflow through the tools provided by keywords themselves.
I may not entirely understand what you are getting at here, but it seems to me there are two different desires being expressed:
The ability to collate material by some binding concept at a larger scale. For that we have item-level metadata tools, such as the aforementioned Label, Keywords and even Custom Meta-Data to a degree (though that tool is often better for accentuating uniqueness than binding groups of items together, given its plain-text field format).
The ability to indicate text level concepts, where the presumption is that multiple concepts would be indicated within an item level outline node. Scrivener has fewer tools for that particular approach—at least where it comes to gathering. I’d say both inline annotations and comments are great tools for marking up text with additional meaning or notation—but neither of those tools are particularly well-suited toward gathering the associated context around them and gathering them, as “particles” of information, into an aggregated view. Where we can run a simple Keyword search to find all larger item-level relationships, there is nothing that can search for all fragments of prose that have been marked with some kind of token, or other indicator.
So the relative pros and cons of these two methods should, in my opinion, drive how atomic you make your item-level splits. Scrivener’s design is such that the penalty for fine-grained splitting is minimal. There are vast areas of some of my projects where items describe mere sentence or paragraph level amounts of content. I can thus very precisely keyword and embellish with metadata individual pieces of text.
But that is very much a matter of taste, I feel. Some find it difficult or counterproductive to work in item-level chunks any shorter than a thousand words or so. And there they will find conflict in the design of Scrivener, should they wish to design a tagging+aggregating system that works at a level finer than their preferred item length.
Sorry for all of the high level theory here, hopefully it doesn’t obfuscate matters further. The short and simple answer is that you can choose to make your outline extremely detailed, and the more you do so the more articulate you will find the metadata systems in Scrivener—but depending on your taste, the less useful or productive you may find the end result.
Yeah, it’s really unfortunate that Apple has simultaneously driven their ongoing macOS development toolkit to the point where it is extremely difficult to provide modern features that build for older systems—while at the same time cutting off the threshold for what older, perfectly functional (and in my opinion, better designed) hardware, can install.
But Scrivener 2 is still good software! We put just as much love into it, and it should serve you for as long as your equipment continues to run.
I guess what I’m asking is if there’s a way to use the Labels for the names of the interview subjects, and have them appear as headers within the text – just as a way of identifying the source name when reading the transcribed material (which I’m sorting by Keywords, as you’ve suggested).
You would have to compile to get a result like that. Have a look in the Help menu, where the list of placeholder tags is documented. Meta-data like the Label are listed under “Document Variables”. These are codes that can be added to your Formatting compile option pane, for the type of document that would print these sections. The most useful way of doing that is in the “Section Layout…” button. For example you can put " (<$label>)" into the suffix field, to have it printed after the title itself (or even without a title, though in that case you might just want the label but itself). You can style the prefix/suffix fields separately, back in the Formatting mock editor:
Thank you so much for your reply, and I’m sorry for my delayed answer…long weekend!
Yes, others have suggested that I use Compile for those purposes, and I appreciate that. But…what about using Labels (or some other metadata) to form headers in the Scrivner project before using Compile?
After all, right now I’m using it as a working document – with lots of sources and Keywords (topics) – so I’d love to maintain some order while I’m still working on it.
Also, I thought it might be better to use Keywords for topical subjects and Labels for sources, both because there are many more topical subjects than sources – and it seemed like it would be better to keep these things in separate taxonomies (or tools). However, I’m certainly open to going in a different direction. Does my approach seem best – or is there another, better way of organizing sources v. topical subjects ?
Unfortunately you would need v3 for labels to make a presence in the editor itself. We added a feature that tints the background behind Scrivenings titles in the editor, based on assigned label. The closest thing to that in v2 would be the View ▸ Use Label Color In ▸ Icons setting, which does impact the editor header bar.
But do also consider how easy it is to have an outliner/corkboard open along side a Scrivenings view. There is a button in the footer bar with two opposing arrows, when you click that it sends what you select over to the other editor (like how the binder works). So you can keep your organisational data easily accessible in one view and the work in another. That’s how I tend to manage things.
Ok, thanks. Appreciate your advice. I’ll take a look at those visual tools.
But I’m wondering what you thought of my organizational approach, namely: using Keywords for topical subjects and Labels for sources, both because there are many more topical subjects than sources.
I’m trying to figure out how best to set up these taxonomies, both for visual and organizational purposes. If Labels doesn’t seem like a good tool / you feel like I should use something else – and/or feel there’s generally a better approach to what I’m seeking to do – please let me know. Thanks!
It might seem like a cop-out, but it’s hard for me to tell you which metadata fields will work best for you. I can say in general the main difference between the two is the most obvious: that a thing can only have one Label, so that tool is best for exclusive assignments. The nice thing about Keywords is that they are freeform—they don’t have to be dedicated to one thing. You could use them for both the source and the topic if that suited your material best (maybe longer chunks of text where there will be more than one source assigned, for example). In cases like that I sometimes use one colour for each major category of keyword. Orange will be sources, red for one major topical grouping, blue for another, etc.
Consider that textual prefixes can broaden labels as well. You can have “src:Rosy” and “topic:Rivets”. Again labels are exclusive, but that approach can work for cases where you have different types of content in the binder, where some documents may only ever need to be exclusively assigned one topic or one source.
Maybe you see why it is hard for me to say. There is enough flexibility that it is difficult to get things wrong, I think. And Scrivener’s metadata management tools are good enough that you can get things “wrong” and correct them later easily. If your exclusive source Label becomes constrictive, you can run a few searches by each label, and do bulk assignments to keywords to migrate over—or vice versa.
Thanks so much for your input! It’s very helpful… I’m still trying to understand which metadata approach would best serve my purposes – and how they generally work as different organizational systems. I haven’t quite grasped which tool is better suited to do what…
Also, I’ve been mulling over this paragraph from your last post…
The thing is, one of the reasons why I thought labels might be better for sources is that they seem better suited for a string function (i.e., one person – or source – one label). In that sense, they seem to match the “exclusive” pattern that you’ve referenced, no? Or have I misunderstood your point?
I think maybe you’re thinking of exclusivity in a one-track sense, but it doesn’t have to be that way. I.e. just because a Label assignment is an exclusive statement made about the item it is assigned to, does not mean that the list of labels itself need be dedicated to solely one function.
You might have a batch of labels you use for Research and a batch you use for the Draft. The research batch has a more topical focus, but the draft batch is more status oriented. Both groups of binder items benefit from the elevated visibility of the Label, and as well that exclusive assignment. But you use a label to mark one item “Apples” and another “Rewrite”.
So long as you never have to rewrite your research files, or topically mark your draft times, it’s more like having two lists of labels than one.
Whether you need that, I do not know, the example was provided more as food for thought than a suggestion.
Thank you again for your help with all of this. I’m still trying to grasp the various different organizational ways that the metadata can be used… Is there any kind of comparative chart? Can’t remember seeing on in the manual…
Just out of curiosity… If you were me, and you had around a dozen sources and 20 - 30 Keywords, how would you put together an organizational taxonomy with Scrivner’s metadata tools?
There isn’t a chart per se, but §10.4.1, Metadata Types, in the user manual, goes over the different available types, their distinguishing characteristics and individual pros and cons.
If all I had to track were the two things you mention, and working under the presumption that all snippets only have one source but may have many topics, I’d use the label for sources and keywords for topics.
Hi,
How can one set up Compile so that the Scrivner document will organize text sections by keywords? And is there a way for one to set up the Compile settings so that the keywords become the titles for those sections?
Compile does not ever modify the contents of the binder, or their ordering. You would need to arrange things that way yourself, probably with a Collection and then compiling using that as your source (instead of Draft). I can say that sorting by keywords is a bit touch and go though, unless every item only has one keyword assigned. It gets more complicated when some have two, others have ten, etc.
As for printing the keyword as the title, the post you quoted is how you would do that—it uses labels for the example, but you could use the keyword placeholder instead.
I forgot you mentioned you were on version 2 still. The menu listing I refer to is in the View menu, instead, and the ability to jump straight to a collection listing from the editor header bar is v3-only.
The short and simple answer is that you can choose to make your outline extremely detailed, and the more you do so the more articulate you will find the metadata systems in Scrivener—but depending on your taste, the less useful or productive you may find the end result.
[size=80][/size]
I can say in general the main difference between the two is the most obvious: that a thing can only have one Label, so that tool is best for exclusive assignments. The nice thing about Keywords is that they are freeform—they don’t have to be dedicated to one thing. You could use them for both the source and the topic if that suited your material best (maybe longer chunks of text where there will be more than one source assigned, for example). In cases like that I sometimes use one colour for each major category of keyword. Orange will be sources, red for one major topical grouping, blue for another, etc.