Deprecate certain "authors" when calculating the selected author for a work

DiskuteraRecommend Site Improvements

Bara medlemmar i LibraryThing kan skriva.

Deprecate certain "authors" when calculating the selected author for a work

1r.orrison
jun 21, 2010, 3:06 pm

This is along the lines of http://www.librarything.com/topic/91861, except that in addition to ignoring blank authors, it would be great if it could also ignore certain other authors -- "various" "no author" "anthology" etc.

2r.orrison
jul 1, 2010, 5:52 pm

It would also be nice to ignore "n/a" in favor of a real name.

3r.orrison
jul 6, 2010, 7:52 am

"Editor"

4staffordcastle
jul 6, 2010, 6:45 pm

"Unknown" and "Anonymous"

5PortiaLong
jul 6, 2010, 9:16 pm

God YES!

6SilentInAWay
Redigerat: jul 7, 2010, 4:00 pm

Great idea, although there should probably be some threshold -- If the author for Beowulf were to be listed as "Anonymous" or "Unknown" 5000 times and (mistakenly) as "Mark Twain" once, I would hate for it to be listed on his author page...

ETA: as in Tim's "fuck you" example.

7jjwilson61
jul 7, 2010, 6:59 pm

I agree with you about Anonymous and Unknown but not {blank}. Anything is better than nothing.

8SilentInAWay
jul 7, 2010, 8:12 pm

agreed

9BTRIPP
jul 8, 2010, 12:15 pm

Re. #6: "as in Tim's "fuck you" example"

Heh ... suddenly want to run off and crank a book out under the nom de plume of "Fuk Yu"!

 

10r.orrison
jul 9, 2010, 5:43 pm

"Anon"

11brightcopy
jul 9, 2010, 6:06 pm

10> You mean even more? ;)

12r.orrison
jul 9, 2010, 6:23 pm

I just tried to get a work off the Anon page... 9 copies by a real person, 2 by Anon, and when I combined them it chose Anon. WTF? It's like the system prefers bogus authors.

13lorax
jul 9, 2010, 6:33 pm

Speaking of Anon, is there any reason not to combine "Anon" and "Anonymous"?

14brightcopy
Redigerat: jul 9, 2010, 6:42 pm

13> For that matter, "Unknown" and "No Author". That last one is more tricky, I suppose. A lot of the books assigned to it are books where it might be better described as "various", while others are ones where it should be "anonymous."

ETA: Actually, I guess "Unknown" suffers the same problem of being various/anonymous. As does "Anon". "Anonymous" seems a bit better, but it has some "various" mixed in.

15jjwilson61
jul 9, 2010, 6:39 pm

Makes sense, except that the edition page will become even more unwieldy.

16jjwilson61
jul 9, 2010, 6:51 pm

On the other hand, Anon could conceivably be someone's last name.

17brightcopy
jul 9, 2010, 7:05 pm

16> In which case, you should do as other authors and list your first name. If you're an author, have no first name and your last name is Anon - well, you're already royally screwed.

18justjim
jul 9, 2010, 7:09 pm

Imagine the royalties you could claim though!

19brightcopy
jul 9, 2010, 10:12 pm

18> Unfortunately, almost all of your work would be in the public domain. Even Primary Colors was written by Anonymous.

20andyray
jul 17, 2010, 2:21 pm

Anon means "i'll be there soon," doesn't it? "We'll be there anon . . ." Today's utes talk alot about things but do not talk about spelling, meaning, or etymology (which is derivation, not meaning.) Oh, and when did alot become one word as most spell it today? And whewn did "bad" come to mean "god," and good come to mean a girl is like Kraft (she spreads for everyone)? God condemn the King and save the language.'Oim a good girl, eye yam!"

21theapparatus
jul 17, 2010, 2:36 pm

Considering there are treatment centers with the name Al Anon out there, I'm sure at least one of them as written up at least something and at least one person has it indexed here.

22r.orrison
aug 19, 2010, 4:50 am

Bump... any possibility of implementing this some day, or is it a reject?

23r.orrison
sep 8, 2010, 5:20 am

And another bump... This would have been nice here: http://www.librarything.com/topic/98263

Can we at least get a "nay" or "yea, someday, maybe"?

24r.orrison
Redigerat: okt 8, 2010, 4:31 am

Prolific Anon
Often true name is known
Again I must bump

25justjim
Redigerat: okt 8, 2010, 4:43 am

There once was a user called rorrison,
Who changed his name to r.orrison,
And with a standard bump,
Proved he wasn't a chump,
But in fact was a good LT citizen.

ps. Me too!

26SimoneA
okt 8, 2010, 6:55 am

I can't rhyme, but I definitely would be happy with this!

27r.orrison
okt 14, 2010, 4:22 am

This would also be useful for surname-only author entries.

28r.orrison
okt 15, 2010, 1:46 am

Copied from another thread:
jjmcgaffey: Yeah...but what about one-name authors? From the classical (Homer) to current children's (Avi), LT would have trouble if it assumed a single name was a surname and the same as the current if it didn't.

What I'm proposing wouldn't cause a problem with that. First, I'm not saying that all single-name authors should be marked as deprecated, clearly those examples shouldn't be. Second, the deprecation would only apply if there was a better choice for the system to use. If everyone entered Wilson as the author for a book, then it would have to be listed under Wilson.

I'm not sure how an entry would get marked as deprecated -- a staff only function, a CK entry, a poll attached to every author page?

29jjwilson61
Redigerat: okt 15, 2010, 9:38 am

We don't want to make this request too complicated though. It seems like it should be pretty easy to automatically deprecate blank authors and a handful of others like various. Adding a function for staff or others to mark certain names as deprecated adds a whole new layer of implementation.

30r.orrison
okt 15, 2010, 9:50 am

Blank is actually already deprecated, somewhat. Most of the time, if you combine a work with a blank author with a work that has an author name listed, the author name will appear, even if the majority of editions have no author name. (See http://www.librarything.com/topic/91861)

For others, there will have to be some way of marking the author entry as deprecated. If it's just a finite list -- say, Various, Anonymous, and Unknown -- then it's easy enough to have a flag that only staff can set.

However, there will always be additions to the list ("n/a"), and it's not unreasonable to consider some crowdsourced way of setting the flag so that we don't have to ask staff to do it. But yes, that could be excessively complicated and possibly dangerous.

Honestly, based on previous experience, if Tim likes this idea, I would expect the first iteration to be something that only staff could control, and then plan to add a user-controlled interface in a couple weeks. LibraryThing has been known to do just enough of a project for it to be useful, and leave the complicated parts for later.

31r.orrison
nov 14, 2012, 11:31 am

Bump, after a somewhat related mention in http://www.librarything.com/topic/144485

32omargosh
feb 8, 2014, 8:15 am

Bump. I combined these two editions and thought it was ridiculous that "Stated, Not" won out over "The Brooklyn Centre Historical Society". Yes, I can override it with Other-Authors, but I prefer to avoid hard-coding the primary author whenever possible. The system should just ignore ratty authors when non-ratties are in the same work.

33jjwilson61
feb 10, 2014, 11:15 am

So how is the system supposed to know that Not Stated is a ratty author name? For all we know there might just be a Mr. Not Stated out there somewhere writing books.

34brightcopy
Redigerat: feb 10, 2014, 11:35 am

#33 by jjwilson61> With a user-applied ratty author flag, which is what this RSI is for.

35r.orrison
dec 1, 2016, 5:26 am

Bump. Here's a perfect example:
https://www.librarything.com/work/8316304/editions
2 editions, one by "Various" and one by "Anton de Wit", WorldCat lists Anton de Wit as the author, LibraryThing chose "Various".

With the recent re-work of author recalculation, it would be nice if this could be included.

36r.orrison
Redigerat: dec 5, 2020, 3:00 pm

Bump.
This feels relevant again in light of recent discussions. If it was possible to flag certain author names as ratty data so they couldn't possibly win as the system selected author, that would get rid of many cases of using Other Author on the primary author.