The more I think about it the more I am convinced that we should use something else than just numbers to distinguish different credits with the same name. In bookogs we not just have names of persons, but also any other imaginable terms and topics.
Idea: Would it be possible to attach a free text field to the credit name, where we could specify the credit name; i.e. US-author, Spanish band, generic term, movie etc.?
Or maybe we could use the method, that is used i.e in wikipedia. If you move the cursor over a link the profile appears.

Other ideas, thoughts?

In this thread:

https://www.bookogs.com/forum/529251-copy-to-draft-profile-name-identifiers-split-info-works-by-and-works-about

The staff stated this:

This has been discussed here, I'm not sure if the community reached a firm consensus on the way to go. I agree that the numbering system isn't ideal, but a good way to distinguish between entities with the same name is important. There are no technical reasons to use the same format on Bookogs.

I've added credits with (Band) behind the name much like Wikipedia pages to differentiate between two different topics.

I think this would work for many types of credits, except ones like "John Smith" where there may be 5+ credits. Then it gets a little more tricky.

Some users have suggested not having suffix at all, but a system, when entering credits, that show details about each credit so it is easier to distinguish between them. Unfortunately, that is not a system in place at this time.

With the current system, I prefer descriptive text to a number.

As it is at this moment we are using the numbering system, all credits should be entered in this way.

With the current system, I prefer descriptive text to a number.

Same here. In addition to not being really helpful, the numbers also create a false impression of hierarchy. There is a long history of users, artists, etc. swapping suffixes for personal preferences in Discogs. It would be nice if we didn't have to deal with that issue at all.

Other system that many databases seem to use is birth/death years, or years of activity. Works could use the publication year as the basic system.

Some users have suggested not having suffix at all

That was probably me. :-) I still think that would be the best option, but I do also acknowledge that it would require other system updates to actually work. With the current system it just complicates things.

As it is at this moment we are using the numbering system, all credits should be entered in this way.

This is not entirely true, as this is a holdover from Discogs. But this is not Discogs proper and the staff has actually stated (see above) that a consensus has not been made.

Here is the direct link where it was stated by staff:
https://www.bookogs.com/forum/529251-copy-to-draft-profile-name-identifiers-split-info-works-by-and-works-about

With the current method of entry on the *ogs sites outside of Discogs, descriptive text is a much better solution than a number. There is no technical reason for only using numbers.

Maybe a vote should be held to determine this. At this time, there are three users on this thread in favor of using descriptive text.

I will concede to the method that is decided, but the decision will need to be confirmed by the staff.

To further the case for descriptive text, I will use Wikipedia as an example.

For multiple credits with the same name, Wikipedia will have a disambiguation page.
We'll use Robert Smith for this example.

https://en.wikipedia.org/wiki/Robert_Smith

A similar page could be created here and at the top would say, "do link link to this page, but to a credit listed below" (or something like that) - I've seen these pages on Discogs too.

This would work better than having to find Robert Smith (29) after opening each page to find the specific one.

In addition to not being really helpful, the numbers also create a false impression of hierarchy.

Also this.

I agree totally - instead of using the discogs numbering system, lets improve on it! Great example from wikipedia by the way.
Thank-you

I don't like the numbering system either, mainly because of the false hierarchy impression. Also, I would like to see some kind of disambiguation function or page.

But, one problem I can see is that a lot of credits exist in name only and there's not enough data available to determine if the persons are the same or different. For instance: Book A shows Robert Smith as Designer. Book B shows Robert Smith as Illustrator. Book C shows Robert Smith as having a photo credit. They could be the same person, but without more information, you can't be sure.

Perhaps a type of hybrid system could be worked out. On the obvious credits where a lot of information is available, a descriptive text system could be used. And, on the low information credits, we could keep the numbers. At some future time, when more information becomes available for a specific credit, it could be changed from a number to a descriptive text.

I don't see the point in replacing one flawed system with the another.
Making credits with someones or somethings name and putting (Band) or (Author) behind it won't work. It will be just as confusing and just as difficult to find what you are looking for as more credits and people with the same name start to get entered to the database another problem would be to change all the credits with a suffix, how many do we have? Hundreds, thousands? That's a huge job to take on. Who's going to do it?

Some users have suggested not having suffix at all

Well, that's what the idea of a free text field as description field is about. No suffix, but an attached field. I don't know if this can or will be realized.

For multiple credits with the same name, Wikipedia will have a disambiguation page.

2nd best idea imo: suffixes plus disambiguation page

But for sure everything with an additional description is much better than the numbering system.

Maybe the staff could tell us what they are able to implement and what they can't.

We don’t believe the numbering system is the best way to go. I believe its existence on Discogs is a remnant of (old) technological constraints rather than it being a good user interface. It being in the guidelines on Bookogs is similarily a remnant of those being copied from Discogs. If people are in agreement with going ahead without the numbering system we will update the guidelines

We are thinking of good ways of making it easier to distinguish between entities with the same name (a lot of good ideas in this thread, thanks for the feedback). In the meantime perhaps disambiguation parentheses might help?

I think before anything is decided whatever new system is chosen and it's looking like that's what people want we need to look into the future to see what problems it might entail and what do we do with all the suffixes? Do you know how many we have in the database at the moment? How big a task will it be to change all those credits correctly to the new system? Will that task be taken care of by developers or will it be a contributor task?
So far I have not heard any idea that has been proposed that I think will work. I like the idea of the Wikipedia style disambiguation page that someone suggested.
How possible and realistic would that proposal be?
Jay
.

In the meantime perhaps disambiguation parentheses might help?

Sounds good to me at least, especially if it's a temporary solution. I think a description would work better than a number.

But as suggested, I think we can also keep the numbers as a secondary system, that way we're not in a hurry to change them all, and we are not forcing anyone to do extensive research.

Anyone have ideas what to use for Works?

Anyone have ideas what to use for Works?
I think the numbering works well for the work pages. The artists are also seen in the lists and we already agreed to the suffixes for the case of the same name for a collection and a i.e. short story or poem.
We will have some work titles that contain dozens of entries but a search for title + author should work well here.

In the meantime perhaps disambiguation parentheses might help?

Yes, but since we are still in an embryonic stage of the database (this will easily reach the number of - don't laugh - 100.000.000 entries, if it will be successful), we can also just wait a bit until we have a final solution. At the moment it doesn't really matter, which system we use.

I think the numbering works well for the work pages. The artists are also seen in the lists

The site search results are only part of the problem. If we really want these solutions to help with submitting, the descriptions also have to be visible in the auto fill fields in the submission form. Currently the Works field in the submission form is no different from the other credit fields: it shows just the title(s).

Anyway, I don't think the numbers are a good system for either credits or works, and I'm glad the staff is looking for a way to get rid of it.

The site search results are only part of the problem. If we really want these solutions to help with submitting, the descriptions also have to be visible in the auto fill fields in the submission form.

Hmm maybe I'm just used to have a second open window too search the correct credit page ;-)
You're right of course. When you are in the submitting process, the autofill info is no help at all in these cases.

I think before anything is decided whatever new system is chosen and it's looking like that's what people want we need to look into the future to see what problems it might entail and what do we do with all the suffixes?

Wikipedia uses the named suffix system and it seems to be pretty good. It's one of the largest and most visited sources of knowledge on the internet.

How big a task will it be to change all those credits correctly to the new system?

It doesn't have to be a task at all.
A numbered or named suffix has the same task: to differentiate between two or more credits with the same name.
Therefore, numbered and named suffixes can exist in harmony.

Numbered suffixes can stay as is and be changed when and if necessary.
It's already been mentioned in this thread by kitchi-gummi and mirva, and a really great idea.

An imagined guideline:
- Use named suffixes for credits that are established. I'd use Wikipedia as a guide for this.
- Use numbered suffixes for unknown or less known credits. As more books/information is added, these credits may be updated, expanded, and/or changed.

The named method would work perfectly with the issue on the other thread.
https://www.bookogs.com/forum/531126-what-is-the-current-practice-with-credit-hijacking-in-this-database
No guessing which credit is the correct one.
Visions and Visions (2)
vs.
Visions (Magazine) and Visions (Publisher Series)

@Supernaut1970:

Do you know how many we have in the database at the moment?

There seem to be around 2000 things in the db that end with a parenthesis with a number inside. But some of those seem to be indicating a year, rather than using a Discogs like numbering system.

Will that task be taken care of by developers or will it be a contributor task?

We could write a script to bulk update the titles, but that's a rather blunt instrument. We could also list them for users to edit. It doesn't feel too time sensitive or urgent to me though, we can take our time. Another difference with Bookogs from Discogs is that we aren't restrained from renaming things (again and again if we want to). Older links will still work and redirect to the correctly named page.

Another thing we could do would be to more prominently feature the id of the credit/entity. This is the number before the name in the url. In a way that number already disambiguates between entities (albeit with a longer number)

Hmm maybe I'm just used to have a second open window too search the correct credit page ;-)

I do that too, but it would be nice if I didn't have to. ;-)

There seem to be around 2000 things in the db that end with a parenthesis with a number inside.

I don't think that's too bad, even though we also have to count the first entry too, which is numberless, so the actual amount is a bit more.

Another thing we could do would be to more prominently feature the id of the credit/entity.

Sure, as long as it's not the whole solution. The id is as descriptive as the number in parentheses. :-)

I went ahead and changed the two Visions credits with text descriptors.

https://www.bookogs.com/credit/48705-visions-magazine
https://www.bookogs.com/credit/530531-visions-publisher-series

Hopefully, if there are any disputes, the staff will defend the decision after the discussion on this thread.

😃

best solution for the moment I think, thx ;-)

I know a bit of time has passed since this discussion, but I will just add my thoughts.

I agree the number suffix is unnecessary when it comes to Works which display with the author name below it (assuming a user has added the author when generating the Works), and the same applies to Book Credits.

However, when trying to select a Credit that has multiple entries e.g. John Smith, the only way you can ascertain the correct variation is to keep opening the variations in a new tab, or alternatively opening a new window and searching for the name in the Bookogs database and methodically working through all the variations until you find the correct Credit

If there is no means of differentiating the variations (such as a number suffix), then it is near impossible to know which ones you have already looked at, apart from recording the unique url number.

I don't know how users sort through variations but I start at the first variation and keep searching numerically until I find the right entry.

By all means replace the numbering system, but it would helpful to have a better system in place. Text descriptions may work some of the time, but I would suggest that this would be almost impossible in the example of John Smith, where few or no details are known about the actual person. It could be case of John Smith (WTF).

I posted a mockup a few months ago that could work to alleviate some of the problems with choosing a credit of the same name.

https://books.discogs.com/forum/634916-new-feature-on-credits-events-dates#post-15

I still think sharpmath's suggestion is great, and I hope we get something like that. I think it would be really helpful. :)

I've been only using the numerical suffix if it has been absolutely necessary, i.e. there's no other way of creating a distintion between that and other existing credits. If I know I'm able to provide dates and a picture, I've left the numerical suffix out because the picture and the dates are a lot better identifiers than the suffix.

As the ID number is longer and harder to remember, I do agree that at the moment the suffix makes browsing a little easier when there is no other info entered. But you still need to open the profile, and sometimes even the books, to figure out whether you have the correct credit.

Top idea sharpmath and I hope it is implemented at some stage. The system you are proposing requires at the very least, that every Credit has a basic profile (such as the stated role) for it to function properly, but I am sure with enough willing workers that can be achieved.

But until then I would suggest that the current system is still required for the reasons that I outlined. I just think it is precipitant to dispense with the numerical suffix before having a fully functiong replacement.

On Discogs there might be 50+ variations to search, and no doubt this will occur on Bookogs as the database increases.

I don't know what other users do when searching 50+ variations, but I search for the name and then tick 'Strict', e.g. https://www.discogs.com/search/?q=John+smith&type=artist&strict=true and I scan the text in an attempt to locate a match, but if that doesn't work then I open and close each variation until hopefully I find the correct entry and then I return to my submission and enter that credit. If it wasn't for the number suffix, I would have no idea which credits I had already checked (although one advantage with Discogs is that the visited link does change colour, I just have to remember to clear my browsing history first).

At the risk of stating the obvious, not all Credits are going to have an image or Events.

Discogs also has a spell checker in the Forum, which would be handy if we can't edit comments!

The database can be used in various ways. I just open every credit I want to check out in a new browser tab, and then go through them. It doesn't matter if they have the suffix or not, and I'm used to using the URL ID number if needed.

The previous discussions have shown that a lot of users want to get rid of the suffixes, and they are not really that intuitive for users who are not familiar with Discogs.

The system already allows us to create identically titled entries. The search displays image and dates, plus by hovering over the link one can also see the unique ID number. Why do we still need an additional, manually added numerical suffix, and how its purpose is different from the ID number?

At the moment, I see the suffix as nothing but a minor convenience at most. Whether that is a good enough reason to keep it still around is a matter of opinion. We can keep them around for a bit longer for the minor convenience; or equally we could get rid of them now, possibly causing minor inconvenience until the system is improved.

Another option, and the most flexible option, is to let the two systems co-exist for the time being. We have abandoned the suffixes on Works - except where needed. That could work with credits, too. It would also lessen the work in the future when the system is improved, allowing us to gradually leave the suffixes behind.

As far as I can see, none of the options hinder the database use significantly. Obviously I believe we should go boldly forwards, but I don't mind the current co-existence either. I do believe though that we should move away from the suffixes - they are an unintuitive, outdated way of creating a distinction between credits, and create a false impression of rank.

I need the suffix to keep track of the ones I have already looked at, i.e. I start at variation 1 and then I look at variation 2 and so on. If they were all listed as John Smith then I wouldn't have a clue where I was on the list. Maybe I am slightly stupid, but I don't have the ability to remember 50+ ID numbers and I certainly don't have the inclination to write all of this information on a piece of paper as a checklist.

I have no qualms about dispensing with the numerical suffixes, but I would like to see an effective system in place before it is discontinued.

The system that sharpmath has proposed is a step in the right direction, but it remains to be seen if the staff will implement it anytime in the near future.

I and, someone else, have with the first name credited added a 'disambiguation' (like Wikipedia or a few Discogs profiles) on a few on them, eg.:

John Smith (2) - editor
John Smith (3) - painter

etc.

Only problem is its tiresome searching for info on 'backroom' credits from the past when more often than not nothing is available or its confusing eg. book published 1994 but copyright 1964, so is design credit from 1994 or 1964?

I don't think there is an easy answer.

This is a problem on Goodreads, and Wikipedia at least has the advantage that most of their listings are not highly specialized such as the Credits on Discogs, where you might have 30 variations that perform exactly the same role.

One problem that I see with the solution suggested by sharpmath is the temptation for users to lump all credits that appear superficially the same (e.g. John Smith - editor) together, but I readily admit the suffix system has a similar problem where the first variation becomes the lazy option.

There is no easy answer, or a perfect answer. Users' willingess/ability to match the credits to the ones in the database varies, and there will always be misplaced credits. It's a really common occurance in Discogs too, and not just with the first entries. All we can do is to ease the process: the entries need to be easily recognizable.

Even though the numerical suffix is better displayed than the ID number, in the end it's nothing more than a second ID number. It doesn't do anything that the other disambiguation methods already do, especially when it comes to identifying the entity.

sharpmath's suggestion is a step forward in the way that in some cases you wouldn't need to open the profile to check who it exactly is. For example, there are a couple of writers that I run into on a semi-regular basis even though I'm not too familiar with them. Both of their names are not unique to the database, and even though I've updated their profiles to have a picture and dates, I always have to double check and open the profiles.

Discogs uses the "display first x characters" but obviously this doesn't always work as the first characters can be anything - it has proved difficult to guide users to think about the search results when they write profiles.

I have nothing against switching to the Wiki type of suffixes either for the time being (assuming they'd be temporary), especially since sharpmath's suggestion is kinda the next step: instead of including it in the title, it has its own field, and can be a bit more elaborate than just "painter".

I've used them occasionally, mostly to separate subjects from the rest of the credits. It might be just me, but it feels weird to use the numerical suffixes for subjects.

Login or Register to post a reply to this topic.