2019 LT Redesign: Implications of Unchosen/generic editions?

@Conceptdawg and I are working an improved "Add book" button. (Don't get too excited, it's about 1% done.)

As often noted, LibraryThing's Add Book buttons go to the Add Book page, and execute a search. Many users want it to just "add the book" without any sort of search. The core justification behind this is that LibraryThing is designed around books, and "works" are not books. You can't add a work any more than you order "a beverage" at Starbucks—they'll make you pick one.*

Our plan is to allow members to add works by adding books with data from the most common edition, and mark the books as "generic edition" or "unpicked edition." If the user wants to specify the edition later, they can do so. If not, okay.

We're now working out the implications of this. Some of them include:

Basic Data

1. Minimal data.
2. Basic data (title, author, cover) drawn from the work level.
3. Other data drawn from most common edition (e.g., publisher?). This is going to be a calculation, and potentially tricky. By default, we'll use Amazon data there, not library data. I suppose it could be an option.
4. Data left blank, so it can be inherited from the work level (DDC, LCC, etc.)

Special Data

1. Books marked as "edition unpicked" or "generic edition." (Other terms wanted!)
2. When user chooses to pick an edition, all their specific bibliographic data is forfeit, as necessary.
3. When a user edits their data, we give them a notice about it, and internally track that it was from a generic edition. If we find problems, we'll treat this data differently.
4. "Unpicked edition" books do not count toward popular bibliographic data. This would be a sort of self-reinforcing loop—unpicked-edition books would get the most popular data, and make that data yet more popular.
5. I think if users don't change anything, their data doesn't change. So, for example, if users add the Hunger Games and get ISBN and cover X, their data is not changed if the Hunger Games gets a new most-popular ISBN and cover.
6. We show generic/unpicked editions differently in peoples' catalogs. I'm unsure how, but something like italic type, or a special icon.

That's my thinking so far. Thoughts appreciated!

Tim

* A side problem is that, when LibraryThing allowed it, people would use the add-book feature in damaging ways--adding multiple copies of Harry Potter and the Sorcerer's Stone, and then editing the titles to match the books they actually had, but neglecting to change the ISBNs, causing wrong ISBNs to spread across books like the plague.

>1 timspalding: * A side problem is that, when LibraryThing allowed it, people would use the add-book feature in damaging ways--adding multiple copies of Harry Potter and the Sorcerer's Stone, and then editing the titles to match the books they actually had, but neglecting to change the ISBNs, causing wrong ISBNs to spread across books like the plague.

Ok, so waaaay back in the dawn of LT I also did this once or twice with obscure foreign magazines. Later, as I learned more of the site I realized the combination problems and such and tried to fix them, but the impetus surrounding doing it in the first place was that I wanted to make a new blank row directly in my catalog for easy editing, and this was the closest way to do it.

Being completely honest, this is still a problem for me with when adding my books manually. I will open up 4 or 5 blank manual-adds and copy/paste data that is the same, changing only the 1 bit of info that's different, and 99.9% of the time I will mess up and create a mess that needs separating. (I have many books that are part of series.)

I love this generic book idea! Only, if it's really a generic book, it should only have the title and author. The cover should come from the most common cover for the work (the logic for displaying the cover could be that if there is no ISBN and the user hasn't picked a user-defined cover then it chooses the most common Amazon cover).

A lot of people just want to record that they read a book and don't really want or need to fill in the book-specific data. This would meet their needs perfectly and has the added bonus of not filling the system with the false data that results when people like this just randomly pick an edition.

Could I use this as my wishlist without having it show up in my books? :-)

>1 timspalding: I think if the generic edition is just going to pull data from the most popular edition, that defeats the purpose of having a generic edition. A user might as well just search in "add books" and click on the first one.

I guess you are correct. I really, really want to separate my wishlist from books I have read or owned. :-(

I think the generic edition should also have the comments field or a small size notes field so you can mark something down if necessary.

Also fully agree with >5 norabelle414: . Half of my own catalog is suspect because I add by ISBN but don't usually check beyond that, especially when adding in bulk, and I don't like the idea that it makes the potentially false data more prominent. I edit in small batches here and there when I have the time.

>1 timspalding: 1. Books marked as "edition unpicked" or "generic edition." (Other terms wanted!)

If you don't like Platonic Form, I suggest it be called an epitome. (Seriously.)

I agree with those stating the generic edition should be minimal data, perhaps just Author and Title. On the other hand, I see the use of bringing in some canonical data: typical number of pages, for example. Maybe that could be pulled in from some representative edition, but only if it were used, e.g. providing readers with pages read based on collections or read dates.

I suggest having a couple versions of a generic edition could be useful. In addition to the minimal version (Author + Title), versions which specify format, language, perhaps just those two categories. Some users would leave it blank, but others could select in those two categories and still preserve their "generic edition" status. The resulting aggregate data on generic editions would be interesting.

I don't think Cover or ISBN makes sense for a generic edition, unless a specific cover design exclusive to generic editions. What purpose is served by bringing over a popular Cover or ISBN? I'm assuming that the generic edition is combined with these other editions at the work level, so let the user choose their own cover or leave it blank.

>5 norabelle414: Except that the first result in the current "Add book" search might have no connection to the most popular edition. I frequently run a few searches, grab the closest match to the book I'm looking for, then edit it, *and then* combine it with existing entries for that work. Sometimes, the version I grab from the search turns out to be too problematic, so I have to delete it to start over with a manual entry.

We can start with the manual entry. But for me, finding the data can be a headache without online references. I'm having to pull up what I can find on LT, Amazon, and various other sites before entering a work.

Adding a book with minimal data in the current system allows auto-population and missed connections that's misleading. From my understanding of this improved process, we could maintain a cleaner edition so it's easier to check for fields that need attention.

>1 timspalding: "A side problem is that, when LibraryThing allowed it, people would use the add-book feature in damaging ways--adding multiple copies of Harry Potter and the Sorcerer's Stone, and then editing the titles to match the books they actually had, but neglecting to change the ISBNs, causing wrong ISBNs to spread across books like the plague."

That's happening with the current system, too. What surprised me is that more experienced members insisted that ISBNs don't matter as long as the titles are close. IMO, generic copies with minimal data would be far superior to wrong data. Generic editions could help considerably.

Re: "Books marked as 'edition unpicked' or 'generic edition.' (Other terms wanted!)"

Suggestion: Basic edition?

Re: "When user chooses to pick an edition, all their specific bibliographic data is forfeit, as necessary."

I'm feeling greedy now, so I'm going to respond to this line with a request or two.

May we get a Combine option for personal editions when moving from a generic edition to a specific one? An alternative to that could be that when selecting an edition, we could view our current selection side-by-side with the new one while making edits (in case of no exact matches)?

Suggestion: Basic edition?

Nice. More positive. Will consider.

>1 timspalding: timspalding:

Re: "Books marked as 'edition unpicked' or 'generic edition.' (Other terms wanted!)":

Sounds like Edition By Subtraction.

"Any edition"?

(Actually, that came from a mental back-translation from the French, "N'importe quelle édition" -- literally, "doesn't matter which edition.")

I like this.

Would these basic editions be combined automatically at work level? I ask because I have a shed-load of semi-pro RPG stuff to add, most of which requires manual entry because it is OOP and not found anywhere except as a work in LT if I am lucky. It would make a huge amount of sense to search LT for these works and create a copy of an existing record to add to my catalogue.

I'm with those who say this should just be title and author.

Would it be possible to have the Add Book button open a lightbox with the ten most popular covers (or as many as there are, if there are fewer in the system), and the one that the user clicks on will be the cover used for the generic ediition they are adding?

I love the idea, but strongly agree with others that you only include the title and author.

Cover should be the most popular member provided cover, or no cover. If you wanted to give them an Amazon cover you'd need to fill in an ISBN, which contradicts the generic edition idea, and then the cover can change later which always confuses users.

ISBN, Publication information, pages, and all other fields should be left blank. Maybe you could put "generic edition" in the publication info field.

I don't think it's necessary to identify to the user that it's a generic edition - they won't know what that means. Nor do I see why starting from a generic edition needs to affect anything else later on (catalog display or editing). It's just a record in their database that only has the title and author filled in.

"Basic edition" is a good term, if it's necessary to have a term. Perhaps when adding... the work page could have two buttons where the current Add work button is: "Add basic edition" and "Search sources for your book".

Obviously if you add a basic edition it should be combined with the work page that it was added from - bypass the usual auto-combiner and force the correct combination.

>7 macsbrains: I think the generic edition should also have the comments field or a small size notes field so you can mark something down if necessary.

Generic editions should still have all the fields in the catalog and edit books page available for editing, they should just start out blank.

>16 r.orrison: I don't think it's necessary to identify to the user that it's a generic edition - they won't know what that means.

True but others viewing their catalogue may care. Others looking at an author page may care.

Let me explain. There are some authors who have a short story and a novel (or collection) with the same name. If the record is just title + author then there is a small chance that there could be manual separation / combination issues as people combine generic versions of the novel / collection with the short story.

Having a flag, even if it doesn't affect display, could supply interesting metrics.

Tim

Has you and conceptdawg had any thoughts on how to handle the case where someone adds a generic edition and then wants to upgrade it to a real edition? Obviously deleting and then going into add books is a possibility (but that might lose some data important to the user - comments, review, reading dates etc), it is also pretty poor from a UX pov as well.

>17 andyl: True but others viewing their catalogue may care. Others looking at an author page may care.
Fair enough - I hadn't considered that side of it. I'll go farther then and say that the indication that it is a generic edition should appear in work and author combine/separate lists.

I'd stand by my statement that a generic edition should automatically be combined with the work it was generated from, whether that's the short story or collection. Short story vs. collection is a clear case of the auto-combiner possibly getting it wrong.

>12 davidgn:
Also:

"Unchosen edition"/"Edition unchosen"
"Unchoosy edition"? "Unparticular edition"? -- erm... "Indecisive edition?" "Undecided edition"?
"Unspecified edition"/"Edition unspecified"
"Unselected edition"/"Edition unselected"

eta
And yes, I think restricting to title/author might be necessary, as others have suggested.

>17 andyl: Wouldn't the member just add the specific information in the edit book page?

I really don't see any reason to mark it as a generic edition, but if it is decided to do that, adding publishing information or ISBN should cancel that. Perhaps the marker could be 'generic' in the ISBN field. Certainly if there is a marker the user should be able to change it.

But as we already have many books entered with only Title or Title and author, I really don't see why these would need an extra marker added.

"Placeholder edition"
"Ghost edition." :-)

>18 r.orrison:

Yes I agree it should auto-combine. However sometimes manual separation / combination takes place. Some people may use scripts to explode out all editions of works when doing clean up. Some people may incorrectly combine a short story work TitleA / Fred Bloggs with the novel work TitleA / Fred Bloggs. If there is no discriminator it would be impossible to separate correctly.

A generic edition flag is such a discriminator.

>20 MarthaJeanne:

Well maybe they might just add some specific information. Maybe they want the entry for Amazon (which pulls in a field, ASIN, that we can't edit), or maybe they want the data from a library source. I would agree that when they add an ISBN (or upgrade a generic entry into a work from a source) that the generic flag should disappear.

Unspecified edition? That sounds the clearest to me.

I'm also of the opinion this should just include title and author, and either the most popular cover or a choice of several most popular ones. Alternatively, just point the user to 'change covers' to pick one after the book is added.

Not including ISBNs would go a long way towards preventing bad data. I like the idea of putting 'basic edition' or whatever into the ISBN field.

It seems a given books added in this way would autocombine with the work, assuming the user chooses from the work page. Not sure this can work at all if not chosen from the work page.

I don't think unspecified editions should include any edition specific data. Just fill in data available on the work level.

Count me as another who thinks the basic edition should include only title and author.

If its status must be indicated somewhere, I suggest putting that information in the Data source field as something like "LT Generic Edition" where now one might see a library, or Amazon, or "manual entry." After all, it is using LT as the data source for the data added automatically.

I strongly disagree with the idea of using the ISBN field to indicate this, for at least two reasons:

(1) Movies and similar materials would not have ISBNs.

(2) Likewise, and of specific interest to me, older books also do not have ISBNs. The most likely use case for basic editions for me would be adding a book mentioned in Talk or listed as part of a series to a wishlist, and I find that often these books only exist in older editions (pre-ISBN). Thus, I would never have a "real" ISBN to replace text in that field.

>24 harrygbutler: You could just delete it. If the purpose is to indicate to combiners that this is a basic edition, then the ISBN field is the one they would see. Just not having an ISBN doesn't say anything.

My preference would be for this to be truly minimal. Populate title and author and nothing else, as though they'd done a manual addition and done only that. Pick the most common work-level cover, as you would for a similarly-generic manual addition. Then allow them to pick an edition, or to do further edits, as they see fit.

>25 MarthaJeanne: And if I delete it, then the ISBN field would still be blank and unhelpful for combiners, just as it would be for a manually added book with just title and author.

In any case, I think it would be bad design to add non-ISBN data to the ISBN field, which then must be deleted by a record owner — who only wanted title and author —, when it seems to me there is an explicit place for that data, the Data source field. And if that field for some reason isn't appropriate and this must be indicated somehow, then perhaps it belongs in a new dedicated field.

>1 timspalding: Works for me. About the only thing I would add quickly after the "basic" add of author and title might be the published year as that helps me a bit!

>27 harrygbutler: Personally, I don't think it needs marking at all. But if it to be marked to help combiners, there isn't much point to marking it somewhere they won't see.

>28 ulmannc: But published year is supposed to be of your copy, and can vary a lot from first publication which is a CK field.

Just one more click...

Adding a book by title/author/isbn/whatever should in all instances bring up the generic listing. People who don't need or want more data just stop there.

But at that point a question is: "Choose an edition?" and the user is taken to something like the current Add Books page where an 'author,title' search will run on (Amazon? LOC? Oversat? Bowker?), or they can enter their own search terms.

>24 harrygbutler: Data source field is indeed the most logical place to indicate this.

>30 Crypto-Willobie: Maybe wording more like "Search for records with your specific edition"?

>32 davidgn:
Sure, exact details can be adjusted. It's the two-step concept I'm advocating.

>24 harrygbutler:
Agree that data source is a sensible place to indicate that it started as a generic edition.

>30 Crypto-Willobie:
I wasn't expecting to see this as an option on the Add Books page. If you want to add just the title and author (or ISBN or whatever specific details you have), you can add manually from that page and not fill in anything else. I think this would be most useful as a replacement for the Add Book button on the work page (as >1 timspalding: said "working an improved "Add book" button").

Generic editions should still have all the fields in the catalog and edit books page available for editing, they should just start out blank.

I'm pretty set on this model:

1. All bibliographic fields will be there, but uneditable. To edit them, you need to de-select that it's a basic book, and/or pick an edition.
2. All personal fields (tags, collections, reviews, etc.) will be entirely editable.

Would it be possible to have the Add Book button open a lightbox with the ten most popular covers (or as many as there are, if there are fewer in the system), and the one that the user clicks on will be the cover used for the generic ediition they are adding?

I think cover-picking should be involved in specifying an edition, but if you have a basic edition, you get the basic cover. What you're suggesting is a sort of halfway solution, and I'm not sure we get a lot out of it.

Cover should be the most popular member provided cover, or no cover. If you wanted to give them an Amazon cover you'd need to fill in an ISBN, which contradicts the generic edition idea, and then the cover can change later which always confuses users.

We're not giving people basic books which look like nothing and have no data. We're not going to have basic editions you can't search for subjects on, and have no classifications. We're going to pick an edition and then freeze the data. If you want to change it, you have to unselect "Basic edition," or however we're going to handle that.

I think this would be most useful as a replacement for the Add Book button on the work page

Yes, this is where this comes up. We want to allow people to quickly add a book, without specifying an edition.

There are some cases in which title and author aren't sufficient to identify a work. It depends how accurate the 'generic edition' needs to be. Maybe include date as well (if the record includes the original publication date)?

Examples:
Harold Monro, Strange Meetings (1917) https://www.librarything.com/work/11944228
Harold Monro, Strange Meetings (2003) https://www.librarything.com/work/10879504

Arthur Miller, The Crucible (play) https://www.librarything.com/work/22988
Arthur Miller, The Crucible (screenplay) https://www.librarything.com/work/7700477

Sylvia Plath, Ariel https://www.librarything.com/work/20186
Sylvia Plath, Ariel: The Restored Edition https://www.librarything.com/work/7223026

>35 timspalding: Generic editions should still have all the fields in the catalog and edit books page available for editing, they should just start out blank.

I'm pretty set on this model:

1. All bibliographic fields will be there, but uneditable. To edit them, you need to de-select that it's a basic book, and/or pick an edition.
2. All personal fields (tags, collections, reviews, etc.) will be entirely editable.

I like this kind of placeholder edition. At the moment I can't tell which books in my library have correct information that I've inputted myself vs books that have possibly incorrect info that I've imported without checking. (Yes, I could have tagged, but I didn't). I have to log books right away when I bring them home or I'll immediately lose them to the stacks which is why the unchecked data has accumulated. If all the bibiographic data is blank and uneditable it will force me to dig out my actual book from whatever mountainous pile it is hiding under to get rid of the blank fields, but it lets me do it at my leisure. I would be motivated to improve my catalog accuracy.

timspalding (#35):

1. All bibliographic fields will be there, but uneditable. To edit them, you need to de-select that it's a basic book, and/or pick an edition.
2. All personal fields (tags, collections, reviews, etc.) will be entirely editable.


I think that it's the fields in the grey area - date, publisher, ISBN - that are most problematic for most people. Those are not "personal", but they're also very much tied to a particular edition. If we want this to be a "generic" edition, rather than an "arbitrary" one, they really shouldn't be populated.

>35 timspalding:


I'm pretty set on this model:

1. All bibliographic fields will be there, but uneditable. To edit them, you need to de-select that it's a basic book, and/or pick an edition.
2. All personal fields (tags, collections, reviews, etc.) will be entirely editable.


This all makes the most sense to me of any options. I really like it, actually. It indicates that people are able to do more if they want to, but also prevents bad data from propagating.


I think cover-picking should be involved in specifying an edition, but if you have a basic edition, you get the basic cover. What you're suggesting is a sort of halfway solution, and I'm not sure we get a lot out of it.


If I'm understanding you correctly, you won't be able to pick your own cover if using the basic edition. I think that will frustrate people. To me, title, author, and cover are the core aspects of a generic record. People won't care if the detailed metadata doesn't match their edition, but they will want their book to look like their book.

>35 timspalding:. Works for me. The Electronic Luddite.

I still think people should be able to pick the cover for a generic edition rather than getting bland default cover. I may not be able to remember the exact edition of a book I read decades ago but perhaps some of the covers already added are vaguely familiar.

>39 lorax: I concur.

For a generic edition, these fields should be blank or — to my mind a poorer choice — have some sort of "None specified" content. Otherwise, you are not adding a generic edition, you are forcing the choice of a specific edition. For a use case of adding books to a wishlist and then sharing the wishlist, for example, this will inevitably give the erroneous impression that that specific edition is desired.

>35 timspalding: Even if you go this route, please do not require "choosing" an edition. If the edition the user eventually has in hand has not been cataloged in a source, what would be the user's recourse if that choice were mandatory?

>40 casvelyn: I agree on this point as well. People looking to simply add the book via the Add Books button are not going to want what they get to have an "incorrect" cover and be unable to change it to the cover that matches their copy or, as mentioned in >42 jjwilson61:, the copy they remembered reading years ago.

>I actually do my own version of "generic work" for my TBR collection. Just a manual entry of title and author, plus a plain black cover from the pre-programmed cover options at the bottom of the choose cover page. Then I fill in the appropriate details, including the real cover after I buy the book (or get it from the library).

But people who want to use the basic version as their finalized catalog entry would probably not be happy with a blank cover.

Adding a generic edition should be the same in effect as going to manual add and entering only the title and author. The only difference would be ensuring that it is combined with the correct work, instead of using the standard auto-combiner; and perhaps choosing the most popular member-uploaded cover. After that, everything should be editable just as if the book was added manually.

>35 timspalding:
All bibliographic fields will be there, but uneditable. To edit them, you need to de-select that it's a basic book, and/or pick an edition.
How do you even pick an edition? You can edit your book now, whatever edition you have searched for or entered manually, without picking a different edition (whatever that means).

I think cover-picking should be involved in specifying an edition, but if you have a basic edition, you get the basic cover
You can change your cover now without changing any other information about your edition, why would you prevent that in the case of a basic edition? And again, having entered your book, how do you specify an edition after the fact?

We're going to pick an edition and then freeze the data. If you want to change it, you have to unselect "Basic edition," or however we're going to handle that.
If you're picking an edition for the user, it's not a generic edition, and it's probably the wrong one.

I was liking where this was going, but it seems to have gone in a very different direction to what I was expecting. If some of these proposals are implemented, I'll just be using Add Manually to add generic editions.

>43 harrygbutler:
"Even if you go this route, please do not require "choosing" an edition. If the edition the user eventually has in hand has not been cataloged in a source, what would be the user's recourse if that choice were mandatory?"

Well, one of the 'choices' would be the manual add where you can do what you want.

==================

It looks like the Generic author/title only level would have to be supplied with a range of cover choices. Otherwise the Generians would rebel..

I am in someone's library. I see that they have the exact same book as i do: same cover, same publisher, same everything. If I press "Add Book" why can you not add the book I am looking at? Why must it be generic?

>45 r.orrison: I agree with these points.

In addition to the wishlist case I mentioned above, I have a use for generic editions to track reading of works in collections (e.g., one of five novels in an omnibus), much as @casvelyn does for TBR books (see >44 casvelyn:). If the generic edition contained incorrect publication data (rather than none), I would have to stick with adding manually (as now) with its increased likelihood of erroneous auto-combination.

>46 Crypto-Willobie: I would not consider editing the existing book details, after "de-selecting that it's a basic book," to reflect a copy in hand to be "choosing" an edition, since I would not be picking from among various editions called forth by a search but simply editing the record. I may have misunderstood what was intended. It also isn't clear whether someone who was largely happy with the information pulled in but didn't like, or didn't want, one element of the "bibliographic data" could get rid of it simply by deselection or would have to search to pull in new data by running a source search.

I am in someone's library. I see that they have the exact same book as i do: same cover, same publisher, same everything. If I press "Add Book" why can you not add the book I am looking at? Why must it be generic?

No, I agree. If you are looking at a specific book in someone's library, it should at least offer that as the primary option. There are some problems here on the data end, but we can almost always provide the same book.

>49 timspalding: If I'm wishlisting I usually wouldn't care about the edition at this point, but I reckon others might. It would be neat if simply clicking "add book" and/or "adf to wishlist" would have a default (ie this edition or unspecified edition) which ideally the user could set, with the other option (or options) available from a little context menu the user could make appear.

>49 timspalding: Now that's exciting!

>35 timspalding: I think cover-picking should be involved in specifying an edition, but if you have a basic edition, you get the basic cover. What you're suggesting is a sort of halfway solution, and I'm not sure we get a lot out of it.

Please don't force me to have covers that don't match what I read (or just like the look of, because I'm shallow enough to pick book editions based on covers). A lot of the books in my catalogue are not books I actually have, but books I've borrowed from the library, but I still want my collection in LT to look nice, not just a jumble of basic covers.

>40 casvelyn: To me, title, author, and cover are the core aspects of a generic record. People won't care if the detailed metadata doesn't match their edition, but they will want their book to look like their book.

Yes. Exactly.

First, my mouth is kind of agape from discovering via this thread all kinds of things that LT can do. As a VERY non-techy user, I've always struggled with the site and its lack of a real Help function. It's the biggest flaw to a site I adore.

Most of this thread has identified a lot of the things I find annoying (e.g., Add Books approach so clunky) so I'm not getting into that here.

I do want to urge LT to step back from the details of what to change and why to think about the most important thing for ANY new product development or redesign: who is using it? That is, for whom are we designing, who is the primary user group, the target, and what other segments do we have to serve? I have been in consumer new products and market research for about 35 years, and these are always the first questions to nail. They drive everything.

The needs of a small library are vastly different from those of someone like me who just wants to make sure I don't buy a book I already own. The site seems to be driven by people who understand and love programming - that makes it tough for the non-techies like me. There are die-hard LT people versus casual users - which group is larger? Which is more important?

Basically, you need to do a quick user segmentation (survey would do it) to identify what user groups exist; how large they are; what their key characteristics, attitudes, needs, and perceptions are, including their degree of involvement/depth of usage of LT; and then decide which group's needs will drive design decisions. I can help you design a survey instrument or you could find someone who is more local -just please don't wing it without a consumer researcher guiding you. Surveys are actually quite tricky to design properly. They're not just a bunch of questions thrown together.

I know designers want to dive right into the process, but it will be a mess if you don't do at least some user segmentation work to guide you.

>53 stringcat3: The site seems to be driven by people who understand and love programming

The site? Nah. The whining? Yeah, that's fair. :-)

>53 stringcat3: That is, for whom are we designing

That's easy...Me.

53. I'll second that.

Since we are talking about who and what the changes are for (>53 stringcat3: and >56 bnielsen:) are we (the poor little users of the code) going to get a crack at a test system or a beta system to see how the changes work so they can be tested by those who do NOT write the code can mess? Developers area the worst testers in my "opinionated opinion". I speak as one who was a developer/coder for many years and then became a hard ball tester since others didn't do it.

I agree with r.orrison (>16 r.orrison:, >45 r.orrison:) about pretty much everything.

Tim, I think you should be more concerned about protecting LT data on the back-end than crippling user experience on the front-end (i.e., the same issue we have with the green plus right now). Basic editions are fundamentally about making the casual user experience easier and more streamlined. The idea that you need to convert your basic book to a non-basic book in order to do anything with it is just going to continue driving people away.

People who don't care about the exact edition of their book should still be able to choose the cover they want.

I'd treat this like empty Dewey numbers: populate most fields with green data from elsewhere, but let the user edit as they wish to create their own black data.

I don't think it's necessary to mark basic editions separately in the catalogue, any more than manually-added books are marked separately in the catalogue. If you need to add this, please make it an optional column that 99% of people don't need to see.

>59 _Zoe_:

The manual entries are identified by the Source being 'manual entry'. So the approach described in >24 harrygbutler: is probably a decent workable one.

The way I'd like to see the Add Book button work is you click on the button and get given 3 choices:

Add a generic copy of this work to your library. This adds the book with title, author and current cover (if any) to your library. No other field is completed, and clicking OK takes you straight to Edit your book to complete the addition process.

Add an exact copy of this work to your library. This copies the other user's data into your book record and drops you into Edit your book as before.

Search for this work. This takes you to the Add books area with the search box pre-populated with title and author, and/or possibly the ISBN. Possibly make this a toggle? Search by title & author, search by ISBN...

Either way, the first two options would auto-combine with the parent work.

I can see most people using the first option to create a basic record and editing it to fit their copy (or not as the case may be). Could this help with throttling issues? If people are adding a book from an internal search, rather than an external data source...

The second option is useful for more obscure or pre-ISBN books or where a member has (possibly manually) added the exact edition you want to add. The third option would be for new releases or when you want a specific edition not recorded in LT.

As with >59 _Zoe_:, I agree you don't need to flag a basic stub entry or indeed, treat it in any way different from any other work in your catalogue.

>60 andyl: That sounds good to me.

>61 Maddz: Search for this work. This takes you to the Add books area with the search box pre-populated with title and author, and/or possibly the ISBN. Possibly make this a toggle? Search by title & author, search by ISBN...

And it uses the source from that record if it's not Manual.

I'm late to the party, but thoughts:

1) Get rid of the danged green plus button all together. "Use the Add Books tab to search for an edition that matches yours." Something.

2) The word "basic" is having a moment these days, but I and others in my age cohort have struggled to understand its current nuances. I think I kinda get it, but I have had to explain what I think "basic" currently means to others, even though they are younger than me. (Looking at you, Basic Witch.) How about "default edition"? "Basic" is gonna sound really dated down the road.

Skeleton (key) edition? Passkey edition? Meh...

Universal editions, prime editions, ...

Quote"Basic" is gonna sound really dated down the road./quote

Isn't that like being concerned that when talking about temperatures, "cool" sounds dated? Or is there a shift in the LT's desired audience that goes away from the dictionary definitions of "basic"?

>61 Maddz:, all of what you wrote is what I'm hoping we get.

Undocumented editions?

I think Bogus Books has a certain ring to it.

>69 anglemark: Good one!

Isn't that like being concerned that when talking about temperatures, "cool" sounds dated?

Yeah, I agree with this. The fact that "basic" has taken on a specific slang usage doesn't mean it can no longer be used in the traditional sense.

Yes like the work you are planning for the *side problem.

Also is it possible to enable selection and search of multiple sources in

Hmmm on generic edition, but really like MADDZ 3 options. search for ISBN to add books really works for my collection, but always happy to add copy data later.

>71 _Zoe_: et al, I confess I might be feeling retro pain from the ascendancy of "basically" which is now so ubiquitous most people basically don't know why it's a problem.

The Urban Dictionary tells me that "basic" these days means trend-following to the point of boring and generally derogatory. So it's possible that younger users could see its use in this context either funny or clueless. Y'all might not hang out in spaces where "basic" is used like that, but LT power users on a thread like this are self-selecting, not broadly representative. There are other demographics besides ourselves.

https://www.urbandictionary.com/define.php?term=Basic

One thing that ought be highlighted is that 'Your book' really equals 'Your specific edition': so at work level the book is 'David Copperfield' by Charles Dickens. Your book is the 1923 Everyman edition with a specific size and binding as distinct from the 2001 Delphi Classics edition as distinct from the 1995 Project Gutenberg ebook. All are the same work, but differ wildly in the book details.

So perhaps making that distinction clearer will help new members understand why the work has less detail than the book. The basic 'Add books' will put a copy of the work in your catalogue without any edition detail - which you are free to add if you desire either directly after adding or at a later date when you actually get your hands on a copy.

Maddz (#74):

One thing that ought be highlighted is that 'Your book' really equals 'Your specific edition':

No, it really doesn't. It equals 'Your specific copy'. Which, yes, it's a particular publisher and year and binding, but it's *also* the specific instance of that edition that you have on your shelf. There is no "edition" layer on LT.

so at work level the book is 'David Copperfield' by Charles Dickens. Your book is the 1923 Everyman edition with a specific size and binding as distinct from the 2001 Delphi Classics edition as distinct from the 1995 Project Gutenberg ebook. All are the same work, but differ wildly in the book details.

Yes, but referring to "Your edition" rather than the current language of "Your book" will make this less clear, and just make people wonder why they can't find *other* people with the 1923 Everyman edition. If they aren't going to have an Edition layer, they need to stay away from the word "edition".

>1 timspalding: I like the idea of generic works very much. In my opinion in these cases title and author as information are sufficient and a most widely used cover. Sometimes, I know that I've read a work as a translation. But if there are different translations available and I don't know which one I've read in the past I would very much like to have the possibility to add a generic version to my library. Later I could change this one to the exact edition when I remember which version I've read, or just leave it as a generic work in my library if I don't remember.

>73 PhaedraB: I'm familiar with the usage (I teach college students), I just disagree that the new meaning precludes continued use of the earlier meaning. The syntax isn't even the same—it's a predicative vs. an attributive adjective.

1. Yay! This is really needed.

2. In theory, I like the idea of only filling in "Author" and "Title". In practice, this becomes a bloody mess with combining (some books have multiple revisions that aren't combined, some authors have more than one book with the same title, sometimes there are multiple anon works with the same title, some series get bad consensus titles that don't specify which volume, etc), so there needs to be some way to make the generic edition of a certain work "stick" to the right work, and plain author+title isn't good enough. Idk what the solution is, depends on if there's going to be some kind of greater rethink of work/edition/book, but just plain author + title will be a bloody mess.

3. It seems like you're implying there will be a way to "convert" a generic book to a specific edition? That would be really great - one reason I don't use the LT wishlist much is that when I do get a book that's on the wishlist, it's a real hassle to change from the wishlist entry to the owned-edition entry.

4. It would really be nice, re: some of the discussion above, to make the workflow for doing multiple manual entries in a row smoother. Even just making it so you didn't have to skip back to "add books" every time would be nice; but so would some kind of automatic clone that carried over certain fields (not title or ISBN!) That would help reduce the temptation to try to clone entries in other ways.

>78 melannen: Regarding No2, I believe adding the generic/basic edition will only be possible from the work page, in which case it obviously belongs there and should auto-combine. Perhaps I'm wrong, but I don't see how it's possible to add a basic edition unless you're on a work page, from which it then pulls the title and author

>79 SandraArdnas: In theory this should be true, in theory there should be some kind of persistent "Work ID" that the basic edition can be permanently attached to, but afaik it's not how it actually works behind the scenes - autocombine can be really unreliable if there's no ISBN to act as glue. So I'm just saying I hope they fix that someone before they roll this out, or using only title + author will be a giant mess.

Even if the original autocombine works correctly, title + author is NOT a unique identifier (I collect King Arthurs, I have spent way too long separating out books with the same author and the same title) (Also old scientific books - so many butterfly experts writing three completely different books just titled "butterflies" over the course of their careers), so unless there's some extra way to make it "stick" to the work, later combiners would have no way of knowing which work was a generic work that belonged there as opposed to bad data that might or might not belong there (or a generic book that had previously been attached to the wrong work.)

It wouldn't need to be any other "parsable" book data like date or ISBN or publisher or whatever, it could be an internal code only visible on combining pages, or it could even be come kind of mechanism to make it so the generic edition could never be separated from the work, but it would need something more than we currently have to make it clear what work it was in cases where title + author isn't going to be unique.

Right now it's a relatively small problem that mostly effects small numbers of copies, but if generic editions take off, we could end up in cases where the majority of books with a title were attached to an ambiguous edition.

The way it works, as I understand it, is that an edition is any unique combination of Title, Author, and ISBN. So any book with the same title and author will be in the same edition, and since it's editions that are actually connected to works any two books with the same title and author and blank ISBN have to be combined into the same work.

However, this isn't any worse than if the used manual entry to add the book with just a title and author.

ETA: Maybe Tim could change the algorithm to use the publication date of the book to determine the edition for it when there is a blank ISBN to better handle this case.

Yeah, it certainly wouldn't be worse in quality, but it would probably multiply the problem a lot if it was easier to do (and baked into the system). It's especially a problem in cases where the majority of editions predate ISBNs.

You can manually separate books with the same title and author and no ISBN - as long as the book entries aren't identical down to the keystroke, anyway - although it's a ton of work to figure out what's what, and then you can get creative with canonical titles and combination notices, but it doesn't stick long-term, partly because I think it confuses the autocombiner. (And last time I tackled the King Arthurs I ended up having to create some "generic" works anyway because there was no way to tell which work they went with, even though that made me feel yucky.)

So yeah, it's not a new problem, but it's one that generic editions defined *only* by title/author would bake in, worse.

Adding original publication year would help a lot, except of course there are a huge number of older books where that's not defined on LT or is imprecise (or, in some cases, is completely unknown - a lot of the anon or fuzzy-authorship works where it's a problem don't have definite original publication dates either.) Original publisher would help too, at least for books that have a record of their original publication. Other Authors helps - often when it's a case of abridgements, translations, etc. having the illustrator/adapter/editor etc. is what makes the difference . (Those are all things I've used when trying to do combining on these, can you tell?)

But none of those would work for all works; I really think the only way to handle it well even the majority of the time would be some kind of database code that just said "this book was generated from this work". Or, keeping generic books out of the combine/separate system entirely somehow.

But since the concept of the "work" itself isn't static or clearly separate from the editions layer, that might be tough to do, too.

Maybe define a set of non-edition-dependent information (some of the other authors fields, most of the CK data, etc.) and have all of that on the generic editions? But I totally see the point of people who want minimal data on these, so IDK.

The original proposal implies adding the most-popular ISBN, which would solve a lot of the combing, but wouldn't fix it for works with no ISBNs, and it would also be kind of misleading because the ISBN *is* extremely edition-specific. OCLC CN or LCCN? New LTCN that will (of course) quickly supersede them both, worldwide? ISBN-like number, only for generic works?

You are never going to get a set up that is 100% at getting books properly sorted into works, no matter what you set up as your criteria. The data simply isn't good enough. Lousy data can be member caused, it can be caused by using Amazon for a source for older books. Basically, there are plenty of members who don't care about having accurate data, and as long as they have some 'King Arthur' in their catalogue, that's all they care about.

A generic edition should start off in the work it was created from. What happens later isn't any different from a manually entered copy that has only title and author, or even only title listed. Some of these will end up eventually in the undefined works that most of the more problematic groups of works have.

I have several books that I can't find a publication date on. Insisting on some particular piece of data that may not even be discoverable even if members care enough to try to enter it is a good way to make the available data worse with deliberately wrong data entered. Yes, it is frustrating, but there is no way to change it.

Yeah, I was thinking of the book level publication date, but that doesn't make sense for generic editions.

What if Tim added a Signifier field to the book record (put it next to Title on the edit page) which could be used to distinguish your book from others with the same title and author. It would be a text field that the user could use to add something like "Short Story" or "Anthology" or "includes x". Then the definition of an LT edition would be changed to add the Signifier to the Title, Author, and ISBN (and the Signifier could be seen on the Editions page).

This is basically the same as the user appending the signifier to the title but a lot of people don't like to change the title.

jjwilson (#61):

What if Tim added a Signifier field to the book record (put it next to Title on the edit page) which could be used to distinguish your book from others with the same title and author. It would be a text field that the user could use to add something like "Short Story" or "Anthology" or "includes x". Then the definition of an LT edition would be changed to add the Signifier to the Title, Author, and ISBN (and the Signifier could be seen on the Editions page).

I think this would cause more harm than good. Those things can already go in the disambiguation notice, and people would use the "Signifier" to say things like "Paperback" or "Signed First Edition" or "Audiobook" that don't matter for combination.

The problem I'm trying to solve is when some different books are stuck in the same edition because the have the same title and author and no ISBN. The problem of not being able to ever separate those books isn't solved by the disambiguation notice.

I like the idea. Disambiguation notices won't separate two books with the same title/author/lack of ISBN (e.g. a short story vs anthology), but a signifier could. It could also show up in the combine/separate list, so you'd know ahead of time what you were doing instead of starting to do the wrong thing and only finding out on the next screen when the disambiguation notices appeared.

Things like "Paperback" or "Signed First Edition" are useful to distinguish a book from a movie of the same name. Only audiobook isn't, and even then only if there exist both abridged and unabridged audiobooks.

It seems so useful that I'm pretty sure it won't be implemented.

Would it be a good idea to add some form of threshold for the combining and separating to be able to go through? I understand that it might be annoying for those who do a lot of combining and separating to have to wait for others to agree with them (but a post in the Combiners group should make that somewhat quicker), but it would probably stop some of the too-hasty combining and separating.

>87 jjwilson61: ff.

Perhaps if instead of a free-text signifier space there was a small number of check boxes to choose from?
# short story
# abridged
# etc...

jjwilson (#89):

The problem I'm trying to solve is when some different books are stuck in the same edition because the have the same title and author and no ISBN. The problem of not being able to ever separate those books isn't solved by the disambiguation notice.


I understand the problem you're trying to solve; what I'm saying is that I don't think this is the right way to solve it. "Signifier" would end up being used much more widely than just in these cases, and it would mean combiners would need to go in and combine thousands upon thousands of books that are incorrectly separated because of the signifier field being used to distinguish paperback from hardcover, or Kindle edition, or first edition, etc. Yes, it would enable a separation that isn't currently possible - but I also think that if someone cares so little about their data that they would use a generic edition in the case where it's ambiguous, it's not worth creating massive disruption to solve their particular issue.

Edited to add:

Crypto-Willobie (#92):

Perhaps instead of a free-text signifier space there was a small number of check boxes to choose from?
# short story
# abridged
# etc...

Now *that* I can get behind.

Why not have the signifier field be a drop-down list of selections? That would avoid people putting junk in there, but I suppose there would be the usual arguments about what should be on the list.

Let's begin the usual arguments!

# short story
# abridged
... what else?

Anthology? Omnibus?

Enhanced? (For editions with extra material)
Illustrated?

>95 Crypto-Willobie: Let's not? Keep that for a fresh and clear RSI thread. Granted, out of all the redesign threads, this is the one concerning a new feature. But I don't see the link to generic editions. This proposal came out of a request to reduce ambiguity when combining works (?), but in this discussion it has become off-topic.

>98 ianreads:
Well, the signifier is something that would be added to the generic title to make it more identifiable when combining etc. And we've already been discussing the signifier on and off since at least post #61, so it seems to me that it's too late to go back. And I certainly don't want to start my own Redesign thread lest I get in trubble. And, and, I didn't start this sub-subject anyway, was just pouring some petrol on the fire...

>97 Maddz:
Not 'illustrated', I think. Illustrations do not (normally) make a book a separate work.

The 136 million books already in here don't have the signifier. And most new books won't have it either. The only people who might use it are those who already try to keep their books in the right works. The books entered from Amazon market data won't have it, so you still won't know where they belong.

My offhand guess is that most people adding generic editions wouldn't have any idea what the signifier choses are about, or which one matches the book they are thinking about, and they won't care.

>100 MarthaJeanne:
Good points...

However someone in this thread objected to the idea of generic books because it would make combining harder. This is a suggestion to get around that objection which makes it relevant to this thread.

>102 jjwilson61: Yeah, I'm the one who made that objection.

It wasn't specifically that it would make combining harder (I just brought up manual combining examples to point out the kind of works I already know it would be a problem with), it's that it would make combining worse - that we would potentially end up with thousands and thousands of automatically-added author-and-title-only books that did not have enough information to match them to a specific work, and that would be getting inextricably automatically combined with other generic books that were meant to be a completely different work.

This would be a problem even if they were never touched by a manual combiner. And it would be a problem not just for people who want a clean database, but also for people trying to use the feature, absent of all social aspects, as they wouldn't necessarily be able to go back in their own libraries and figure out which book was meant, either.

(If the user added a generic work based off of an already ambiguous work, fine, but if they've made the effort to find the exact work they need based on date/publisher/lccn/etc and then adding the generic work loses that specificity, that's a bad implementation of the feature.)

The extra signifier as proposed could be cool for many things, but I don't think it would solve the problem I was trying to point out, especially as it would almost require someone to manually add it before any new book record was made, and a lot of works never get touched by manual combiners, and most of the people adding generic editions certainly won't be. (Also, you could probably already do basically the same thing by abusing canonical titles. As I may have occasionally done when there was no alternative...)

The real fix to the sitewide problem with bad autocombines is probably a true edition layer (and combining tools and changes to the autocombiner go with it).

But in terms of "don't let generic editions make a huge mess in the meantime", I think some way or another to automatically "sticky" them to the work they were spawned from is still the only good solution.

>95 Crypto-Willobie:

This would be excellent for music.

At the moment we have to edit the title to distinguish:

Libretto
Full score
Vocal score
etc.

Most people don't bother to change the title, and as many of these have no ISBN they are wrongly combined.
And the best that can be done retrospectively is to dump unidentifiable editions in catch-all works.

I think one of the worst mistakes in any collection of data is false precision.

Anything other than author and title is going to show something that is not necessarily true. Why pull anything from the most popular? That may change over time, and may very well not match most of the editions.

I think you should just call it something intuitively obvious, like "generic".

As a library user it is VERY important that when adding books we must be able to select the correct edition of that title - not a generic edition. Filling in additional data later would just make more work for us. For those users that do not care what edition is entered into their system why does it matter to them if more information is added all at once. Not sure why you would want to do this.

Nobody's suggesting taking away the ability to select a specific edition.

>106 HandelmanLibraryTINR: Because if you're tracking what you're reading you may not remember the edition or if your maintaining a wishlist you may not care about the edition. In either case it's not just more information, it's inaccurate information.

HandelmanLibraryTINR (#106):

For those users that do not care what edition is entered into their system why does it matter to them if more information is added all at once.

Because there's a difference between "unspecified" and "wrong". They may not care about getting the RIGHT information, but some would rather have NO information than WRONG information.

Or, the difference between something which is very precise but not very accurate and something which is accurate, but imprecise (to use the terms used to describe measurements). [1] I know what I'd choose!

[1] https://en.m.wikipedia.org/wiki/Accuracy_and_precision

>106 HandelmanLibraryTINR: The primary purpose I would have in choosing a generic edition would be in cases where I do not own the book in question, either because I borrowed it from someone, or because the book is on my wishlist. In either case, I would much rather fill in missing data on a record than have to fix incorrect data.

Coming in late to this discussion. My problem with adding books seems to be the opposite of a generic edition. I agree with >111 shadrach_anki: that a generic edition is fine for a wish list or for entering books that I only read from the library decades ago and had only noted a title/author in my reading records. I look up books I want to add, usually old, often with no ISBN, and find exactly the edition that matches what I have, but when I add it there's no way to get that edition. Even the cover picture doesn't come up as a option when I've seen it used on LibraryThing elsewhere. This is where not being able to copy an entry is so frustrating.

>112 SF_fan_mae: Where are you searching for the book, with site search or from the Add Books page?

Searching on LT doesn't help you to add it to your library, it just takes the bare bones, title & author, and searches whatever source you have previously indicated. So not terribly helpful. You can try (from Add Books search) looking for "Title, Author, Date" or "Title, Author, Publisher" (without the "" of course).

Also, see what you have marked as the default source for Add Books. If it's Amazon, good luck. Try Overcat (previously successful library searches) or Library of Congress. Finally, if you find the edition you are looking for, use the Members with this Book to find someone who has the edition you are looking for. Look at the Book (not Work) page for their book. Down at the bottom it says "data source". If it's blank, they entered it manually, so that is the only option. Otherwise, set your Add Books search for that source.

I don't understand what you mean by you can't find the cover. Where are you looking? Go to the page (not Catalog view) for your book by clicking the title in your library. On the left-hand navigation bar, you will see an option for Change Cover. Click that to display all covers for that work found on LT.

FYI, the only source that will deliver a cover with the entry is Amazon, but there's all sorts of problems there. We can fill you in if you wonder.

>112 SF_fan_mae:
>113 PhaedraB:

Totally agree that one of the most frustrating aspects of LT is the inabiliy to copy an existing edition.

Inputting using the same data source as the original user generally is generally a waste of time, as the entry may subsequently have been heavily edited to correct it by the original user.

Also I often want to copy and then further edit an edition I have entered manually myself, where many of the fields are the same.
For example, if a long list of multiple author fields is repeated, but then has to be re-typed from scratch.

In the LT social groups (i.e. where users discuss what they are reading and so forth) the book bullet (BB) is a major thing. If I see that another reader liked a book, and I know her taste is similar to mine, I want to make a note that I want to read that book--so I take a book bullet. Supporting this activity, and ideally using this terminology, would be a major UX improvement for LT. I don't give a darn what edition I'm going to read--I just want to be able to add this book to my reading list without a lot of fuss.

>115 NinieB:

the book bullet (BB) is a major thing.

Explain? Never heard of that. What is it?

To quote @helenliz, "A book bullet is a book that you want to read because someone else recommended it to you. That can be a recomendation from someone you know in real life . . . or from a thread on LibraryThing where there's been a good review by someone."

See it in action: https://www.librarything.com/search.php?search=%22book+bullet%22&searchtype=...

>117 NinieB:

So, basically, it means "Wish List"?

>118 lilithcat: Yes, with the added component that you've learned about it socially. That added component is a big part of community-building in groups like the Category Challenge.

I'm not suggesting that the wishlist functionality isn't there. The "Add to Wishlist" button is how I keep track now. But none of the names I saw in this discussion were resonating with me, and I suspect that talking about works and books and editions won't resonate with a lot of users that LT wants to keep. So I wanted to make sure that the UX designers at LT are also aware of this usage, which will resonate with some users.

Just for myself, it irks me to have to go through the hassle of Adding a Book when I'm not actually adding a book and have no real book (virtual or print) to catalog.

>119 NinieB:

So I wanted to make sure that the UX designers at LT are also aware of this usage, which will resonate with some users.

Not unreasonable, but I have to say that when I do a google search on "Book Bullet", this usage does not appear until the second page of results, and guess where it's from? https://www.librarything.com/topic/307050 Indeed, that's the only instance of that usage in the results.

So while it might "resonate with some users", it's doubtful that it would mean anything to the vast majority who aren't members of that particular LT group.

>120 lilithcat: So while it might "resonate with some users", it's doubtful that it would mean anything to the vast majority who aren't members of that particular LT group.

True, for now. It's a nifty neologism, and perhaps this is simply where it originates? In any case, the concept is familiar and I know of no competing term that should be adopted instead. This thread is the first time I've come across the term "book bullet" but it immediately resonates with a relationship I recognise as having had on LT and offline for decades.

So, if I understand correctly, your "Book Bullets" would be a collection of works? That's how I've wished Wish List worked for years.

This talk of book bullets (which by the way is a term I'm familiar with from several groups, including the Category Challenge, 75 books and Club Read) and wish lists has reminded me of somethign I'd really like to see, and that I don't think exists now - but if it does, PLEASE tell me.
I'd like to be able to "mark" an author as someone I'm interested in. Obviously, with specific works I can just label those as "to read" or something, but that only works if I have a particular title in mind. If it's just a matter of "I want to read something by Thomas Hardy" then I have, to my knowledge no way of doing that. For me, this is most needed when it comes to poets, where I'd like to read not just a specific poem or collection but "something" (or everything) by that poet.

>123 PawsforThought: In LT terms that sounds like an Interesting Author checkbox.

A "Book Bullet" is specifically a book that Person A wants to read because Person B read it, liked it, and reviewed it (or at least talked about it) on LibraryThing. I am NOT a fan of the terminology*, but it is common in the most popular social groups on LT (75ers, category challenge, club read, green dragon, etc.). It's an interesting circumstance because it's not exactly a personal recommendation (Person B doesn't say "hey Person A, I think you would really like this book!") but it is a recommendation borne out of the relationship between Person A and Person B. So Person A wants a way to record that they want to read, or did end up reading, a book because it was mentioned by Person B. I think something as simple as a "recommended by" field could fulfill that.

I can understand the reluctance to embrace a term or concept that is not familiar to everyone but I think we should be celebrating things that are very specifically LibraryThing.

____________________

*Maybe we could change the name to Book Wizard**?

____________________

**This is a sports joke.

NinieB (#115):

In the LT social groups (i.e. where users discuss what they are reading and so forth) the book bullet (BB) is a major thing.

Yikes. I tried participating once and gave up just because it was too chatty - I had no idea that violent imagery was commonplace as well!

norabelle414 (#125):

I can understand the reluctance to embrace a term or concept that is not familiar to everyone but I think we should be celebrating things that are very specifically LibraryThing.


I very much do *not* think we should be "celebrating" terminology that many people will associate with violence and death. If they don't happen to participate in the social groups, that terminology could be tremendously off-putting.

>125 norabelle414:

I can understand the reluctance to embrace a term or concept that is not familiar to everyone but I think we should be celebrating things that are very specifically LibraryThing.

Well, it's specific to a small subset of LT users, and it has a narrow meaning (referring only to those books one a) wants to read, because b) someone in a group recommended it).

But we are talking about a name for a generic edition that would be used by everyone, regardless of whether they wanted to read it (they might have read it already, might want to acquire it to give to someone else, etc.), and regardless of how they learned about it.

So "Book Bullet" simply doesn't apply.

>126 lorax: It's about as violent as getting hit by an arrow from Cupid.

>126 lorax: As I clearly said, I don't like the terminology, only the concept.

>127 lilithcat: What is a small subset of LT users? The term "book bullet" is used thousands of times in Talk, not including common abbreviations (BB) nor people like myself who use the concept but not the terminology. A small subset of LT users use the term "Melville Decimal System" but that doesn't mean it shouldn't be used at all.
There are already entire sections of LT devoted to recommendations, including those from members; this is a just a variation on that.

A combination of
1) generic editions for wishlists/TBR lists and
2) a field for "recommended by" a particular user
would fulfill the specifications for "book bullets" without actually using that term at all. (or come up with a different term and then maybe the people who are using the old term will start using the new one)

>126 lorax:

I very much do *not* think we should be "celebrating" terminology that many people will associate with violence and death.

Particularly as they seem often to phrase it as "taking a book bullet", in other words, being shot.

jjwilson (#128):

I doubt many LT members know people who have been killed by an arrow, or live in a country where mass arrow-shootings are a nearly daily occurrence. That metaphor is fossilized in a way gun violence is not for many people.

I thought book bullet refers to bullets as in formatting text.

On the more general topic of the thread, I'm not really the target audience for generic editions, but I do wish there was some way to link to some works without actually adding them to my catalogue because they are not and it irks me.

The most obvious is the wishlist, which for me would work best as some sort of list linking to work pages, but not adding either specific or generic edition to my catalogue. As it is, I've added 2 or 3 and given up on maintaining a wishlist. Perhaps lists as a feature could be used to maintain a wishlist in this way, but I haven't explored it yet. If it can, than 'add to list' button on the work page would be a huge help.

I’d love an add to list button too. Just the work would be fine.

I understand the benefits of putting "wishlist" titles on a list rather than a collection, and the shortcomings of using a collection for wishlist titles. That said, I want to point out that one benefit of the collection approach is that titles added there push LT reviews into my Reviews module. This is a key benefit for me and why I use the Wish List collection rather than a list or another account.

There are other ways to push LT reviews of books into my feed, but none as direct. (Friends or Interesting Library reviews also go there, but of course for every title, not just those I've selected for my Wish List.)

>123 PawsforThought:, you could create a Talk post with author touchstones and edit or repost with changes as needed. A star on the post or a link in homepage Notes would make the list more accessible.

>135 aspirit: Sure, but that would mean a whole lot of work. Editing touchstones is annoying as all that. Not to mention that it'd be clogging up Talk. If everyone did a talk post it'd be even more difficult to find people's actual talk threads.
The equivalent of a "like"-button (but in this case an "I'm interested in what this person has written"-button or "remember this author"-button) would be much simpler, and easier to access.

>134 elenchus: By all means, people could still use the existing form to maintain a wishlist. But for wishlist in particular, and some other cases less prominently, it would be useful to be able to maintain some sort of record for oneself without actually adding it to my library.

I really only barely touched on lists, so I have no idea what and how I can do there, but I believe the feature would in general find more use if works could be added to lists from a work page, both for works in my library and those not.

>136 PawsforThought: So, are you envisioning something along the lines of Favorite Authors (which I see you are already using, or I would have suggested that as a workaround), but called something different (Authors of Interest or similar)?

>138 rosalita: Yes, something similar to that. Preferably a button on an author's page you'd click and they'd be saved as an author of interest (which is a good name, btw) - the same way favourite authors are - although the button for both authors of interest and favourite authors needs to be much easier to find (higher on the page) than it is now.
The problem with using favourite authors as a work-around (even if I didn't already use it for its intended purpose) is just that - it's a work-around. People aren't going to use a work-around in the same way they would a purpose-made function, because it's not obvious. And I really believe this kind of function would help members expand their reading (I know it'd help me expand mine) - because it'd be an easy click to "remember" the author and the information would be right there when you wanted it, instead of having fo forage around all the nooks and corners of LT.

>139 PawsforThought: I agree that Favorite Authors and Authors of Interest are really two entirely different things. Sometimes people will recommend an author to me, and not a specific book of theirs but just in general. It would be nice to be able to have a list of those instead of having to pick a specific book at random to add to a wishlist.

I also agree that the interface to add an author to your Favorite list could be a lot easier to find!

Can we use "book recommendation" instead of bullet?
That is what it is, so calling it what it is seems like a good way to satisfy everyone. It has clarity (yay!), popularity and no connotations of violence.

>140 rosalita: Exactly.

Why can't you simply create a collection with whatever name works for you and put the recommender in the comments? Seems obvious to me.

Personally, I find "book bullet" both incomprehensible and egregious. It's tone deaf at the very least. Until the poster explained I could make no sense of it, and I've been on LT since 2007.

Maybe, I thought, it was a book I vigorously wished to deaccession. Perhaps it meant a volume I'd hurl with great force off the balcony with the hope I wouldn't clobber anyone over in the storage facility next door. Or maybe I could get the proprietor to put a target on some part of the roof.

But cheez louise, in these times talking about being hit by a bullet is simply not something I can imagine a website promoting. Imagine the impression it would make on newcomers!

I've been telling myself that "book bullet" for recommendations from a social network was a play on the phrase "bite the bullet" in getting around to reading instead of "to take a bullet" for a group. Either way, it is an unpleasant phrase.

>136 PawsforThought: I was responding to your urgent tone by pointing out a way to mark authors with the features we have now. An Interesting Authors button that adds an author up a list would be handy. I've used Connections for some, but not all authors are or can be LT members.

>144 aspirit: I didn't realise I had an urgent tone, and certainly didn't mean that it was something I needed right this second. The reason I put please in capital letter was just for emphasis, because it's a feature I've really missed the past couple of days as I've been surfing around LT and going from one author to another (via recommendations).

>113 PhaedraB: "Where are you searching for the book, with site search or from the Add Books page?"
I generally search for a book first in "my library" to make sure I haven't already entered it, then in site search, and if I find it there, I click on Add Book and search there, usually with OverCat first, by title and author. If it finds the edition I'm looking for, great, but usually I just have to go with what looks like something close. I can then go to "change cover" and look for a match. The problem is with some of my really obscure old SF, that only a few people have entered. I can see that someone else has put in the cover, but it doesn't come up on the covers page. Sometimes, none at all come up. I tried to come up with an example by looking at the books without covers in my library, but from there I can no longer find the covers I know I saw originally.

>146 SF_fan_mae: Most likely, those were Amazon covers if they disappeared. User-uploaded are normally always available to choose, even those by private users are seen by others and available.

For obscure SF, check ISFDB. They have covers for most editions in the database and you can upload them yourself. If you've checked there already, never mind, but it's useful info for anyone with older SFF titles

On the topic of Book Bullets:

Apparently the name would have to be changed, but the concept is actually different than "recommendation". A "recommendation" is a book that someone specifically suggested you should read, but says nothing about whether you agree and have any interest in reading it. A "book bullet" is a book that you want to read because someone else talked about it, and can apply even if the other person's comments weren't very positive. So by all means suggest a different term, but don't just try to merge the concept into something else.

Here's a relevant suggestion for a field where we could specify how we found out about a book. Besides being helpful for individuals* to be able to track, LT could do a lot of interesting stuff with this information.

*It should go without saying that I don't mean *every single LT user* would find it helpful. Sadly that sort of thing doesn't go without saying here, so consider it said.

I doubt many LT members know people who have been killed by an arrow, or live in a country where mass arrow-shootings are a nearly daily occurrence. That metaphor is fossilized in a way gun violence is not for many people.

It may still be a rare occurrence, and necessarily with fewer casualties, but it should be noted that crossbow attacks aren't unheard of in countries with reasonable gun control—it's common enough that I wouldn't say the phrase is entirely fossilized. Several in the Toronto area in the past decade:

https://toronto.citynews.ca/2010/12/03/son-accused-of-killing-father-in-crossbow...
https://www.cbc.ca/news/canada/toronto/crossbow-victims-mother-brother-1.3740496 (3 victims)
https://www.cbc.ca/news/canada/toronto/crossbow-shooting-mississauga-1.5098342

On the topic of generic editions for wishlists:

I want to emphasize that this is only half of what's needed for a satisfactory wishlist, at least from my perspective; there also needs to be a way to keep the wishlist separate from the rest of our books. If someone else mentions a book that seems interesting, I want to track that without adding it to my library.

On the topic of generic editions in general:

I still hate Tim's proposed implementation, where generic editions are distinguished from other books (and uneditable) until you specifically roll them over. This adds an unwelcome complication to a feature whose main purpose is to simplify.

>146 SF_fan_mae: As was said by >147 SandraArdnas: the covers that don't come up are almost certain to be from Amazon. The Amazon covers are tied to ISBN, which means that if the publisher changes the cover on that ISBN, the Amazon cover that is displayed will also change.

There is a trick to harvesting covers from Amazon without bringing along their various add-ons like the "look inside" sticker.

Right-click on the cover and choose "Copy Link Location" or "Copy Image Address". The exact wording depends on your browser. Past that link into the browser and see if it's an acceptable image.

For example, I randomly chose Addictive Organization. I right-clicked and got this URL:

https://images-na.ssl-images-amazon.com/images/I/51QpGYK-q4L._SX326_BO1,204,203,...

This one looks good. I could then go to the Change Cover page and paste the URL into the "Grab one from the web" field. Bingo, you've got the cover, it won't change, and other people will be able to use it, too.

It seems to be less of a problem than it used to be, but if you find extraneous stuff attached to the image, check the URL. If there is anything after ".jpg", delete and try the URL again. Once you've got a clean image of the cover only, copy the URL and proceed with "Grab from the web" as above.

In addition, if you are using Chrome, you can right-click on the cover image and choose "Search Google for this image." Sometimes you can find something with higher resolution that way. And then do the same right-clicking for the image location as above.

>148 _Zoe_: Book of interest?

BookBurger?

>151 Crypto-Willobie: BookBurglar? Cool typo.

Book bullet has the connotation of being struck with a sudden, passionate, and not necessarily welcome, need to read that book. Do any of the rival suggestions imply that?

BookBeMine!

Tome-torpedo

>153 jjwilson61: book whim.

Is there any real push to add the label (with the ammosexual name or any new term) to a generic/basic edition field?

We'll still be able to use tags, collections, comments, etc. after the redesign. I don't remember Tim saying the basic edition would be uneditable.

I don't see the need as well. I've read books based on LT talk posts as well, but also through comments on a host of other sites and places. Comments are so much richer than a simple field. Why bother recording a book came into your attention through a post by _Zoe_ when you can just link to the post itself (in the comments)?

I use 2 sets of tags for every book on my wishlist. The first notes how I came across it or why I want it (completion of a series or author, a connection with another of my books, serendipity, directly recommended to me, etc.). A second set of 3 tags denotes how much I want them: nice to have, want or need. In the comment field are all other details I like to record, like the URL of the exact comment or review that caused me to put the book on my wishlist. In some cases my wishlist entry even becomes a book entry (through editing) with a tag denoting it came from my wishlist and keeping all historical information.

And nothing's stopping anyone from calling something a book bullet in any of the groups where the usage is prevalent, regardless of what a potential feature or field would be called.

>157 ianreads: Well, why bother with any dedicated fields rather than just one big comment box per book? You could equally put title, author, tags, reviews, etc. in a comment box. But giving structure to the data lets you do a lot more with it.

>158 _Zoe_: Nice slippery slope you have there ;-) I do agree that more structure is (almost) always better. However, so far the powers that be haven't even done the most basic and obvious useful things with the date fields (so far). So I don't share your confidence that they would do more with this potential field. I could probably easily derail this thread by setting up the straw man that people want to use the field to power a Litsy-style LibraryThingfluence score: e.g. posts of mine have inspired so many people to read a book, which is reflected in my score, which is bigger than yours.

If there were to be a generic "inspired by" free-form text field (emphasis on the free-form), I don't see how anything useful could be done with that. A LibraryThingfluence-style feature could probably only work if it was built into the Talk functionality itself.

PS: I wouldn't even be that horrified by an influence score as I imagine many others would be. Especially if it were tied in to new and improved green plusses and the work info hover feature on touchstones. Meaning that you would be able to add a book from any talk post mentioning a book and that the system would record and remember where you added it from.

>160 ianreads: I don't think that's a slippery slope. I'm not suggesting that the addition of an unstructured field will lead to the elimination of other structure. I'm just saying that the reason for structured data in this case is the same as the reason for structured data in any other case.

I don't think the field(s) should be freeform. I'd like to enter data of a specific type, most notably usernames (similar to the tags field).

Tim has a particular dislike for reading dates and reading data in general, so there's some hope that other data might be treated differently. And even minimal features would be better than nothing: for example, being able to see all of the books that I added to my list thanks to Nora.

>161 _Zoe_: And I need to make sure I never forget that time I read Rewired: Understanding the iGeneration and the Way They Learn because of you :-P

>162 norabelle414: This is why it's important to distinguish a book bullet from a recommendation ;)

>162 norabelle414: & >163 _Zoe_: That right there is a book bullet for me. Thanks!

>164 lesmel: No no no no no no! Don't do it! It's very bad.

Dodging a bullet here, and coming back to generic editions. Could we have access to adding generic editions as a way to add a work reference to LT but not to a personal library? Suppose you don't possess a work as a stand-alone entity, but do have it as part of an omnibus/combination work, and you want to add the work-to-work relationship to CK, but that work hasn't been cataloged separately on LT yet. You can't add that now. No, I'm not talking short stories...I remember how that played out.

I'd also love my contained works (generic or otherwise) to trigger the green checkboxes, but I'm guessing that's a Clydesdale pony.

I absolutely would like to be able to use LT as a source as I have a lot of obscure books which I often can't find in any of the sources and have to add them manually, only to then find other members have already added them.