Most Requested Minor Feature, The
LibraryThing is now sorting books without initial the/an/a, so a book like Shelby Foote's The Civil War goes under C rather than T. It took a while, but it's done now!
I've made a stab at foreign-language support. At present it works with: English, German, French, Spanish, Portuguese, Romanian, Dutch, Swedish, Danish, Norwegian, Icelandic, Welsh, Albanian, Hungarian, Tagalog, Turkish and Malagasy. The most important missing language is Italian.
At present, it's a totally dumb process, removing definite and indefinite articles without attention to the underlying language, which LibraryThing doesn't current track. This leaves it open to interpreting Die Another Day as Another Day, Die (German die), and is why I've absented Italian from the list (the article I would produce such non-books as Was a Teenage Dominatrix, I). This problem will go away when I start tracking languages.
For cataloging geeks: (1) I used the list provided by the Library of Congress. (2) Although MARC records can indicate "the number of character positions associated with an initial definite or indefinite article to be disregarded in sorting and filing processes," I have decided not to use this information. The system needs to adapt when someone changes a record, at which point the MARC record can no longer be a reliable guide. Also, many records don't have MARC records. (3) Unless I'm mistaken, here's an example where language is important, even when the definitive article has no potential ambiguity: Les Bons Mots : How to Amaze Tout Le Monde with Everyday French. Right?
I've made a stab at foreign-language support. At present it works with: English, German, French, Spanish, Portuguese, Romanian, Dutch, Swedish, Danish, Norwegian, Icelandic, Welsh, Albanian, Hungarian, Tagalog, Turkish and Malagasy. The most important missing language is Italian.
At present, it's a totally dumb process, removing definite and indefinite articles without attention to the underlying language, which LibraryThing doesn't current track. This leaves it open to interpreting Die Another Day as Another Day, Die (German die), and is why I've absented Italian from the list (the article I would produce such non-books as Was a Teenage Dominatrix, I). This problem will go away when I start tracking languages.
For cataloging geeks: (1) I used the list provided by the Library of Congress. (2) Although MARC records can indicate "the number of character positions associated with an initial definite or indefinite article to be disregarded in sorting and filing processes," I have decided not to use this information. The system needs to adapt when someone changes a record, at which point the MARC record can no longer be a reliable guide. Also, many records don't have MARC records. (3) Unless I'm mistaken, here's an example where language is important, even when the definitive article has no potential ambiguity: Les Bons Mots : How to Amaze Tout Le Monde with Everyday French. Right?
posted by Tim at 2:53 AM
25 Comments:
Nice addition. But could you *please* work on the FireFox page flicker bug reported here ? It's driving me crazy.
Sorry to nitpick but it's a "definite article". A "definitive article" would just be an authoratative or conclusive article.
... I'll go back into hiding now ...
Any possibility to override that, either for a specific book or for one's own bookshelf?
While in general it is a good thing to ignore the articles, there are certain books whose placement I find jarring when they are sorted without the articles. It seems wrong to me, for example, to find The Paper and the Papers among the P's, when "The Paper" refers to a specific paper, The New York Times, not to newspapers in general or to a writing surface.
But I'm perfectly capable of sucking it up if this isn't possible. ~smile~
By the way, are you sure that book isn't about Lester Bon Mots, the well-known raconteur, known to his friends as "Les"? (I'm also capable of seeing ambiguity in the oddest places!)
I think I don't want to get into the practice of absenting individual books, but a preference feature would work for me. Indeed, I'd like to compile a whole bunch of "preferences" into one page.
(1) Default way of expressing dates (US/Euro)
(2) Default currency (for value calculations)
(3) Sort with The/An/Etc or not
In theory, I could put this all on the Profile/Edit Profile page, but that page is getting a bit crowded...
Well done! I bet that took some effort.
I see it works for the Title field now. This kind of sorting should probably work for the Summary field also, which typically begins with the title information.
Thanks.
The most important missing language is Italian.
I think you mean "Japanese" ;*
Seriously, though, any support for Chinese or Japanese coming?
Prorata: But neither Japanese or Chinese have definite articles!
Are you complaining about general lack of support for them. I do admit there are problems. If so, are you complaining about Romanization problems with library records, or problems showing non-Romanized characters.
Are you complaining about general lack of support for them.
Yes, especially with regard to searching for Japanese or Chinese authors or titles in the actual languages. Would be very useful (not to mention there are a lot of readers in that part of the world - think of all those potential Library Thingers).
As regards searching you are certainly right. I need to get the database to think and act in UTF-8, and it just aint doing it.
For Chinese, there's Douban.com. I can't remember the Japanese one, but I think there is one. Of course, I'd like to make LibraryThing be the choice for people who also read English...
Hello to all great minds.
I think a most requested major feature should have been how to bring a uniformity in subject headings, I mean the tags.
While I like free-floating tags and idea of open source, but at the bottom there must be some way to use what is already there.
Let me give you one instance. Using the InMagic cataloging tool, it allows accepting the existing keywords, or allows you to add to the database. The priority is to stck to some standard. This is the best practice - rather than having a freedom to use whatever you like-it. My current concern is how do we uniformly use one or two terms for all that is synonymous, for e.g., religion, religions, religious, faith, faiths, creed, creeds, belief, dogma, cult, etc., etc.
If we forget the foundations of an authority list, then, we may be responsible for this mishap. We will also be a facilitator in multiplying the current chaos, whether one is searching or browsing. Internet is already a mess with its vertical, horrizontal and all-directional scatter.
We as library builders, must adhere to what we have been professing: organizing our own home as a model for others to look at.
Any comments?
Hi Mohamed,
The question of "standard" tags is a matter of ongoing debate, but I've been gradually swayed by the arguments against standardising on tags.
Firstly, read the article Ontology is Overrated, which is linked from the "About" section of this site. I found that really illuminating.
Secondly, If you look at any tag on this site you'll notice the tag pages includes "related" tags, and allows you to combine tags which are synonyms. I can see merit in combining the tags religion, religions and religious (in fact I did it just now), but the others you mention are not (in my opinion) synonymous enough to be combined.
If you look at the tag combining interface you'll see some more discussion about this subject.
Embrace the chaos, and look at the more magnificent order that coalesces from it when we set our boundaries a little wider ;)
I think the del.icio.us booklets work well. They offer you some choices from you and other users but you are free to use what ever you like.
Re: tags debate above.
The problem I have with standardization, or even combination, of tags is that it results in groupings of similar words rather than similar concepts. And concepts can be tricky to systematize. There is some overlap among religion, religions, religious, etc., but not enough that would warrant their combination or standardization. I use 'religion' to tag books about that social phenomena - Daniel Dennett's Breaking the Spell, for example. 'Religions,' on the other hand, refers to the institutions, and would be handy for a book on comparative studies. 'Religious' is best suited for texts within a particular institution, such as The Book of Common Prayer or Daily Missal. My interests lie in 'religion' as a type of social phenomenon, and not in any particular instance of such, so grouping all those types of books would be a great disservice. I don't think anyone takes issue with combining '20th century' with 'twentieth cent.' - those are merely superficial differences.
Poke around a bit and you will discover that professional librarians have been and are much involved in these topics—publishing articles, teaching classes, delivering talks, etc. And some libraries (eg., those hirsute people a UPenn) have already added user tagging to their catalogs, and the major OPAC providers are mulling it.
Your assertion that "we librarians call [tags] subject headings," is just wrong. Professional librarians call them tags, and understand they are not the same thing.
Don't get me wrong. There is a discussion about how useful tags can be, and about the level of control that libraries should give to users, but these conversations are conducted without abuse, between clean-shaven people with library science degrees.
You might start by looking at the blogs of some of the librarians that are thinking about these issues:
http://www.librarystuff.net/
http://www.theshiftedlibrarian.com/
http://www.blyberg.net/
http://www.familymanlibrarian.com/
http://www.librarycrunch.com/
http://stephenslighthouse.sirsi.com/
http://www.librarian.net/
http://librarianinblack.typepad.com/librarianinblack/
http://catalogablog.blogspot.com/
http://www.daveyp.com/blog
You'll find a lot more by looking at their blogrolls. (And apologies for the selective list, guys.) Many of these blogs have discussed LibraryThing, not in your language. I also recommend Clay Shirky's "Ontology is Overrated," a good, if polemical, introduction to tagging and a response to some of your concerns ( http://www.shirky.com/writings/ontology_overrated.html ). Personally, I think he overplays the usefulness of tags.
Lastly, you make a point of repeating that LibraryThing is not a library, but a "pile of clutter." Of course, my assertions about library size are half in jest; LibraryThing is not a library in the traditional sense.
It needs stating, however, that a library is not "a thing presided over by a librarian." It is a collection of books. If there is a second element necessary to the library, it is the reader, not the librarian. If there is a third element, it is some sort of organization. Besides being cluttered up with Dewey Decimals, LC Call Numbers and Marc records—your assertion that LibraryThing is just titles is misinformed—LibraryThing is also organized by tags. So, using the term literally, was the Library of Alexandria. And who organized that? A poet!
If LibraryThing is not a library, it is a collection of libraries, mostly (but not exclusively) personal ones. This fact doesn't deprive them of their "library" status. There are far more books in personal libraries than in institutional ones, many of which started out as personal libraries anyway. I also suspect there are more books in smaller institutional libraries—classrooms, churches, Masonic temples, etc.—which also lack professional catalogers.
Congrats, Tim... two million books and you got your first blog troll. ;)
ExVivre's point is fair - just because I see no distinction between religion, religions and religious doesn't mean that there aren't people who do make that distinction. And if that's the case, the tags can just as easily be uncombined.
I'll now return to my work as an unshaven geek ;)
Well, what gets me in this case is not the opinion, but the authority claim. "CalculatorThingies? We professional accountants use slide rules! Get back in your hole!"
The opinion has value, overblown as it is. "Family Man Librarian" (Steve Oberg) wrote an interesting post "Are folksonomies really the better way?" to which I responded at length. I do not think tags will replace professional classification, but it will be in the mix. The goal is findability. There may be only one road to truth, but not to findability!
Yes, someone should mention Abby is a bona fide librarian and works on LibraryThing. I didn't because it felt like bringing out one voodoo doll to cancel out another—as if having someone to say "we librarians" made the argument right.
Speaking of "Library 2.0," Edward Vielmetti ("Superpatron"), John Blyberg and Eli Neuberger are hosting a "Library 2.0 unconference" in Ann Arbor on Feb. 14. I wish I could go. See http://wiki.library2.net/index.php/Library_Camp_signup
Anyone else going?
Tim
Incidentally, the unconference is on April 14, not February.
{Uninteresting comments re Mcquaid's post deleted}
This is a great place to list the books you own. It has other uses, as well. I like it.
Only people who depilate with face-scraping implements are worthy of respect. Just ask any religous prophet or deity.
My most requested minor feature would be the search for books on amazon.co.uk and amazon.de - both providing similar data as amazon.com in a similar way, but opening up to lots of non-US books. Even with the correct titles and author names (umlauts!).
My major wish would be an import option of my considerably sized BookCrossing bookshelf.
One of the two would make me consider using LibraryThing for real - both of them would earn you not lots of money, but another 'lifetime' user (from Germany, who would have guessed). Probably lots of lifetime-users, actually.
As for the tag debate: The reasonable way seems to be indeed to offer a dropdown list of tags using the same letters (as far as the new tag is typed), but allowing for entering another tag - hassle-free - when none fits perfectly.
About that Firefox flicker.....I only get it on my home computer...and I use Firefox both at work and at home. (I think same version on both PCs). There must be another element to it. I also notice that the display does not completely fill my screen at home, whereas at work it does. Related?
I concur with the comment about a drop-down tag list displayed as one starts entering a tag. But what about records with multiple tags as mine have? Will each successive tag (after the comma) show a drop-down list? Since a librarian recommended this tool, I'm going to add the Dewey Decimal number for our library cataloging. Is there any way to speed this DD# entry up?
Hi
When I do a search and sort the result by title (like this), the articles still are sorted as part of the title. Just thought I'd mention it ;-)
Post a Comment
<< Home