Feature:90155 Add an optional "nature of the relationship" field to the Make Variant screen

Summary: Add a new optional field to the Make Variant screen. That field would be a drop down box with a list of allowed "relationships" including "abridged", "expanded", "restored", and a general purpose "revised". We could then have this field appear next to the Variant Title on the author's Summary Bibliography page. We will also want to modify our Title display logic so that Publication records would be grouped based on their respective Title record.

Original discussions from Kraang's Talk page

I believe variants should only apply if the work is essentially the same, only retitled. That's why I prefer new records for revised works, regardless of whether or not they've been retitled. Most editors here on the ISFDB create variants because it's simpler to associate the two works. I don't know what the "official" stance is, but it may be somewhere in the Help section. Back to the Bradbury: if I were doing it, I'd create new records with the only association to the original work being in the notes of the play versions. Perhaps Ahasuerus has had more experience in this situation and can provide some guidance. Mhhutchins 10:29, 9 Dec 2007 (CST)

This is a very tricky area that, AFAIK, is not well documented in our Help pages at this time. Unfortunately, I don't think there is a perfect solution given the state of our software at this time. Let me use George O. Smith's Highways in Hiding (1956) as an example that illustrates the options that we currently have. An abridged version of the novel was published by Avon as Space Plague in 1957.

Option #1 would be to enter these two Titles separately and document the relationship in the two records' Notes fields as Michael suggests above. Our data would be then complete, accurate and comprehensive, but the only way for our users to find this information would be to review the Notes fields for every Title record in George O. Smith's bibliography. That's quite a bit to ask and I doubt that most casual users will be in a position to do so in Smith's case, let alone Silverberg's case. I wouldn't be surprised if some excited George O. Smith fan eventually ordered Space Plague online, not realizing that he was getting an abridged version of something that he already owned.

Option #2 would be to link the two Title records the way they are currently linked in the database and explain the relationship in the Notes fields. The danger here is that a casual browser may not read the Notes and assume that the two texts are identical. He may then make decisions based on this incorrect assumption, e.g. "I have read Space Plague, so I don't have to bother with Highways in Hiding".

Option #3 would be to add a brief description to the abridged Title, e.g. "(abridged)" or "(revised)". A quick search reveals that we have used this technique over 50 times, but it's not common. The problem here is that we would be separating the title of the Title record from the title of the Publication record, which would take us further away from our "what you see is what you get" ideal.

After making the first ~10,000 database changes in the summer of 2006, I concluded that the best compromise way to handle revisions, abridgments, etc would be to add a new optional field to the Make Variant screen. That field would be a drop down box with a list of allowed "relationships" including "abridged", "expanded", "restored", and a general purpose "revised" (with "butchered" reserved for Keith Laumer's titles). We could then have this field appear next to the Variant Title on the author's Summary Bibliography page, thus addressing the concerns raised by Option 3 above. We will also want to modify our Title display logic so that Publication records would be grouped based on their respective Title record.

Unfortunately, the proposed approach will require software changes and we all know how much free time Al has. If we do decide to implement this feature, we may want to do it in conjunction with Feature:90075 Pseudonym and vt sourcing. Ahasuerus 14:13, 9 Dec 2007 (CST)

This one would be fairly trivial to implement. The only question is once the feature has been put in place, what state should all the current title variants be placed in ("revised?"), whilst waiting for someone to put them into the correct state? Alvonruff 12:45, 10 Dec 2007 (CST)

Well, I assumed that most variant titles don't change the text and only modify the title and/or the pseudonym, so I figured that we would make the new field optional and make the default value "blank" (or NULL or what have you). Do you expect most variant titles to have textual variations?

Also, would it be feasible to make the list of allowed relationships ("abridged", "expanded", "restored", "revised", etc) editable by moderators the way the list bibliographic sources is editable? I am sure we will miss something blindingly obvious on the first pass. Ahasuerus 15:45, 10 Dec 2007 (CST)