An edition group organizes various editions of the same (or very similar) publication, namely by the same author(s) and with ostensibly the same content.
Naming
The name of an edition group should reflect how the editions it contains are best known by readers, in a specific language. Edition group aliases represent known variations of edition names.
Editions
An edition group can include multiple editions of the same published works in different formats. Example:
An edition group can include sequential editions of the same published work. Example:
Some Book
Some Book, 2nd Edition
Some Book, 3rd Edition
An edition group can include editions that share the same intent, but may be composed from different works. For example, the collective works of an author:
Edition groups are a BB concept which I have never fully understood. I know that they try to transfer the concept of MB release groups to BB, but I’m not sure whether they have an exact definition even.
As far as I understand their current usage, for editions which contain a single work, their purpose is basically equivalent to the work entity’s, they group editions of the same work.
So I’m really not sure whether grouping editions which contain different versions of the same work(s) makes sense or not, i.e. whether edition groups should have a wider definition than works (for the one-work-edition case).
An edge case when it comes to translations: Should editions which contain a completely disjoint set of translations to the same language be contained in the same edition group or not?
They contain translations of exactly the same work, their content has the same language, but the translators are different.
P.S. I guess the exact definition of Edition Group is a problem which didn’t require much attention so far, because at the current stage we simply have not enough users to have a non-negligible amount of edition groups which contains multiple or even many editions.
Yes, that’s the way we handled an edition group so far.
I think it does and it should at least contain variants like revised, abridged and expanded versions (as @pbryan has suggested).
This is more difficult. The different translations often have different titles. What title do you chose for the edition group and how do you find it when adding a new translation?
Agree, this is where it gets problematic. In the Aesop fables example I used, each of the editions are translated by a different translator, and each has a slightly different title. I proposed they be consolidated under one edition group.
I see a couple of options for editions translated into the same language but with different titles:
Group them together. Pick the name that is most commonly used, and list the others as aliases. Not great because potentially popular names get shunted to aliases and don’t show well in search results.
Group them separately. This makes finding the edition you’re looking for easy because you’ll see it directly in the search results, and won’t have to sift through aliases to see them.
I would vote for #2. If that’s the consensus, then the Aesop examples above would be invalid. I think I can live with that.
I would vote for #1… Remember it’s not just in translations that can have different titles, the same book in the same language can have different titles in different countries. Here’s a Wikipedia list at least a dozen of examples.
It’s inevitable one title will have to picked, this is not an issue as long and as you remember to add alternative titles as aliases.
OK, let’s try it. Would be great to have an automatic addition of work title variants to the alias field of the edition group they are included (but that’s just a dream at the moment ;
FWIW, MB Release Groups are made to group releases and release variants. For example, a simple release, another with bonus tracks plus a collector edition. MB Works are a completely different, in a way, on a more abstract level. A Work could be in several Release Groups (for example, in compilations or as covers).
In books, the analog of an MB Work would be a novel but also a short story. A short story could be in several completely different Edition Groups. I suggest that an Edition Group could regroup distinct editions from different publishers or in different countries.
OK, where does everyone stand on the issue of translations? The current proposal is to have a separate edition group for each language. Should that stand?
I’m of two minds about this. What makes logical sense to me is that all translations should be in the same ed. group, but the interface should group them by language. As things are now… yes, it would probably be a mess. But, if we are to separate the languages, there should at least be a “translation of” ed. group to ed. group relationship to connect them.
When it comes to dreams, @indy133, yours can easily come true… All editions are connected to ed. groups, so adding an “add title to edition group aliases” checkbox to the edition creation wizard should be relatively easy to implement.
This relationship would be between edition groups, which can have different translations by different translators. But works do already have translator relationships, and when there are multiple translations, the disambiguation indicates the translator. You might want to look into those guidelines.
OK, then your schema is about the same as MB’s. But then I am missing something : if translations are different works, how can different translations be in the same edition group?
Ed. groups contain editions, editions contain works. Using the “translation of” relationship on a work has no effect on the ed. group. Having such a relationship, would allow this Ed. Groups to be recognized as translations of the same Group and be displayed as such in the UI. Just think, one book can be translated into dozens of languages, if there is no way to group them, Ed. Groups can also become very messy (still not as much as ungrouped editions) — exactly what we were trying to avoid to separating languages.
OK, I understand your point. But as a database developer, I would like to be sure that this would not open the way for data inconsistencies. Could a work be a translation of another work and their Ed. Groups not be ? Or the reverse ?
