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.
Ok, thanks. So…it sort of feels like I do, in fact, need to create Collections after all – that is, if I want to create an output in which I have documents organized by Keyword sections, correct?
Actually, I really want to save these documents to a sharable file (e.g., Word / .docx), not print them.
Putting that aside, can you walk me through how to use the keyword placeholder with the aforementioned set up? Sorry – I might b missing something in the 2.0 set up.
Originally, I believe you were creating one collection for each and every keyword and running into issues with having so many of them. That is why I suggested that you could simply use the Keywords panel to call up a list whenever you wanted, rather than storing lists of lists.
That is different from having one big collection list organised by whatever axis you prefer—be it keywords or modification date, and then using that to compile with.
No, that’s why you would have one big list of items sorted by whatever you want.
It’s pretty much the same thing you did before. The only thing different from the previous setup is that you would disable the Title checkbox (I think?), and then edit the prefix/suffix to use the <$keyword> placeholder instead of the label.
One other thing of note, since you’ll be using a collection to compile, everything will be on “level 1”.
Something else to consider, if you don’t need to print any actual content and are just wanting to print metadata, is §25.4, Exporting Meta-Data to a Spreadsheet, pg. 426 in the user manual. It might be easier to do what you want in Excel, I mean to say.
Yes, that’s true. But I also want to share a document in which keywords and sources could be seamlessly compelled and shared with my colleagues.
I guess I’m not following that point; I don’t quite understand why it’s not possible to compile all of my Collections into one document. There’s really no way to do that?
And so, does that mean that I’m going to have to run individual Compile sessions to produce all of my keyword collection sets??
Ok, thanks. I’ll take a closer look at that – esp. after I get an answer to the last question I raised here. Thanks again for your help!
It is not a use case that has been anticipated by the way the feature is designed. The ability to compile by collection was added to make it possible to work with alternate ordering, or to create filters against the main draft. They were never meant to be used like folders, where one can just compile all of their collections one after the other.
That’s one way you could do it, but like I said before, I think using one larger sorted collection is going to be the most efficient approach. Does sorting by keyword in the outliner get you pretty close to what you want? If so, you can drag that sorted view into a new collection to add the items in that order. I’d do it like this:
Use Project Search to search for “*” (which returns everything).
Set the search settings to only look in Keywords (which changes “everything” mean only things with keywords assigned).
Click on the “Search Results” header bar to load the search results into the main editor, and switch to Outliner mode.
Add the Keywords column if necessary, and then click on the column header to sort by it.
Thanks so much for this! Ok, I’ve been trying this, but I’m not quite there. Quick questions…
-Do you literally mean just inserting a single asterisk in a Project Search?
-I ask because I did include a single asterisk in a Project Search, then selected Keywords (though the list seems a little short, but maybe it’s fine), and then I’ve switch to Outliner mode, but then…there’s nothing in the main window. What am I overlooking?
-I did add the Keywords column, but that didn’t seem to affect anything. What else would you suggest that I do?
No worries, my over-explaining is an attempt to help with that larger understanding, rather than a specific yes/no type understanding.
So yes, just an asterisk. When project search results seem incomplete, check the other settings and make sure there are no undesired constraints (like only searching excluded documents from the draft folder). But if everything looks fine, try setting the search type back to “All” and spot check items you feel should have keywords; make sure they really are assigned.
As for the rest, it’s hard to say, maybe a screenshot would help, but there is one step that is very important in there worth clarifying: you clicked on the “Search Results” header bar above the binder sidebar, right? That is what loads the sidebar list into the main editor. Otherwise you may just be turning outline mode on for whatever file you had selected incidentally—and its outliner is going to be empty since it has no child items.
Thanks so much - again! I genuinely appreciate your “over-explaining”! Again, I just want to be careful as not to ask for too much or something unreasonable. That’s all.
Ok, the good news I think I got this to work. I’m honestly not sure what settings I used to achieve the desired result, but now I see the text appearing in Outline mode! Well, there’s one small problem, which is that this Project Search looks like it has turned up selections of text that do and do not have Keywords. Super quick questions…
-Is there some way to set up the “Project Search” so that it only includes Keywords?
-With the “Project Search” I’ve got the following selections: “Keywords” (under Search In); “Exact Phrase” (under Operator); “Search ‘Included’ Documents” and “Search ‘Excluded’ Documents” (under Options). That’s it. Should I include / exclude anything else within the the Project Search?
-Also, am I saving this as a Collection?
-If not, just so I’m clear, how do I go from this set up to Compiling this? When I hit Compile, it looks like it reverts back to the raw draft.
-I feel like I need to answer that question first, before I press on with understanding the exact steps you were suggesting in the Formatting section within Compile…
Thanks so much again for your help – and patience! Feels like I’m getting there, thanks to your help!
The project search settings you describe will only return documents that have a keyword assigned, yes.
My suggestion was to use one larger collection rather than a bunch of small ones, so mechanically there is no difference between these two things, except you’re done compiling after the first collection of course. But it sounds to me as though you haven’t created a collection from your search results yet, as described earlier. So there might be a little cart before horse problem here.
Right. I agree. So, why do you think my Project Search is also including text that does not have Keywords? What should I do to fix so that it only captures split text that has Keywords?
Maybe you could provide us with a little example—maybe three binder items or so, showing a case with all three returned in project search but only one or two of them actually have keywords assigned.
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.