Hello Everyone!
Stylemonkey here!
I want to ask the community what do you think about how to express that the plot of multiple books (works) takes place in the same universe?
I had a think and I think the Series entity type might be a good fit for this.
A series can have several works by different authors attached.
Would love input from the community about this - would it work? Do you have any edge cases where this would not work?
Any other ideas/comments?
Etcetera!
Links:
Ancient pre-Series entity type ticket (closed) [BB-95]
New ticket now that we have series (newly opened) [BB-884]
Disclaimer: Iāve never used BookBrainz, but I use Wikidata for this sort of thing a bunch.
Does BBz support series of series? Iāve seen some nested universes on Wikidata, like Multiverse - Wikidata which contains Earth-41633 - Wikidata and a bunch more. That example is for movies, not books, but I wouldnāt be surprised if some books had something similar. The closest I can think of for books is Chronos and Kairos - Wikidata which could arguably be a multiverse of two universes or a universe with two book series.
I think if there is an official list, a series would be appropriate. If not, there can be some disagreement about what works belong or not, so a collection (which I think BB doesnāt yet support) would be better.
So, I guess what Iām saying is we need collections on BB to support any kind of looser groupings.
-EDIT-
Immediately after typing this I opened BB and saw Collections right on the home page. I guess I just never noticed it, and assumed it would be on the Add drop-down menu.
The general idea still stands, series for official lists; collections for other cases.
I think using the āSeriesā entity type for Universes is a really good idea. It saves us from having to create a whole new entity type just for Universes, since Series generally already do everything we need (grouping multiple works, different authors, etc).
I was also wondering, do you think my open pull request (PR #1225 for ticket BB-634) which adds Series to Series relationships would be useful for this?
Also let me know I can do any heavy lifting regarding this?
coming from a MusicBrainz perspective, I think series could be a good way to handle it, or there has been a ticket in Jira for a new Franchise entity for MusicBrainz [MBS-8273], tho we might not need something quite so complicated for just books
Also just coming from the the MusicBrainz perspective, I think expanding series into non-series āgroupingsā will get confusing.
I think a less specific grouping, like tags or collections, would be better suited.
Otherwise you might be opening the floodgates to a lot of different groupings, that each make sense in their own way but are also not series. Maybe thatās fine, but my instinct is that series should just represent series and that some programs etc that use BB data eventually will benefit from knowing exactly what theyāre pulling when they grab āseriesā data.
I think itās reasonable to be wary of using Series as generic grouping.
I think using it for āin the same universeā would require the addition of a ātypeā of Series. Currently you can only select the entity type (series of authors, series of editions, etc.) but nothing more precise.
Do you think this addition would be enough to keep things separate when required? So for third parties, you could filter by series type, for example?
I will add that it seems the boundary is sometimes not super clear, and some series are described as āXYZ Universeā from the get-go by the author or publisher.
I think in most cases we can use āseriesā for these types of works. I canāt find a series that takes place in the same universe, where the term āseriesā might cause problems. The only series in bb so far that comprises two different series is Asimovs āFoundation Expanded Universeā. Here we have two series that are not linked to another, except for the āloreāā¦the same universe. But I donāt think itās a problem to use āseriesā here.
Does anyone have an example where the use of the term āseriesā presents a problem?
Does āseriesā imply an order? There are universes that donāt have a coherent order of works within them because the works were not coordinated with each other at all. E.g., Strowlers ā SCULA 1.0 & 1.1 seems to let anyone add to the universe.
Yesā¦Star Wars and Star Trek came to my mind first also. Different authorsā¦different sub series with different protagonistsā¦held together only by a series name and the same universe, but where is the problem? Itās just a matter or organizing the series. We will have sub series or/and āis part of āseries abcā relations. Even if you create various types of āseriesā or if you use a different category. You canāt deny, that there is a relation and we want to / or have to document this relation. I canāt see the benefit of making it more complicated than necessaryā¦
The real potential messes are outlined in the second and third bulletpoints. These might be addressed by specifying in the guidelines what should/can be included (e.g. āno fanctionāā¦ though I imagine lots of people would love to be able to group universe fanfiction well so it might not be that simple)
But as mentioned earlier, if an application expects to display a āseriesā to the user, then adding things that are not āseriesā confuses the schema and the output. We canāt predict how the data is used so itās important to keep the schema clean and predictable. However, the addition of a series ātypeā, as suggested by monkey, would solve this particular problem.
Sounds like creating a āBad Bankā for messy sub series I understand the reasonā¦but you canāt erase the messiness by moving it to a differrent categoryā¦it will haunt us all nonethelessā¦It will not disperse to the winds like āAerozolsā.
If thereās a series-series āpart ofā relationship, then it could be handled with multiple series. āStar Wars canonā is part of āStar Wars universeā. Authorized works go under the former, fanfic under the latter.
I donāt think thereās anything inherently messy in grouping same-universe things. What I would like to avoid is something that has a specific and predictable meaning regressing to no longer being specific or predictable - for humans and/or machines.
But I also donāt think itās an insurmountable challenge.
P.S. in MusicBrainz I think tags do a good job of grouping things like this - tags are indeed considered one of the few āmessyā.imprecise elements in MB.
