User talk:Mshonle/Around 2000 Words

Why the number of books doesn't matter
Admit it- you knew I'd bite on this :)

I don't think the number of books really matters that much to the average quality of books, or to the usefulness of the sight. Let me explain why.

First off- the average quality of books. The average contributor doesn't want to come and add content to random_book_0001. He wants to add content to a specific subject. This is especially true with new and starting books- he might make a quick edit of an existing book to add or fix something, but writing significant text requires a passion for the subject. If he can't write about his preferred subject he is unlikely to just switch subjects, he's more likely not to significantly contribute. This is a loss for Wikibooks. Instead of some content, we have none. Reducing the number of allowed subjects only works if the people that would have contributed still do. I don't think that would be the case. (Note: that doesn't mean we shouldn't clarify the policy about what is and isn't a wikibook, it means that we shouldn't discourage any topic IMHO).

And yes, the above is describing how I feel personally as well as what I feel as well as how I think the average contributor would act. If I was told I could only write on certain narrow topics, or had to get approval for some other topic I just wouldn't bother. I enjoy teaching, but there's other things just as valuable I could be doing.

In addition, the place I expect this will hurt most is versus our best assets. You mentioned yourself- the best books are those written in the majority by 1 man or a small team, and then tweaked by others. The same situation goes on in open source- the biggest projects, including Linux, tend to have a core of 1-2 dozen people who do 90%+ of the work. Why do they put in the effort? They're passionate about the subject. If those people aren't allowed to work on the subjects they're passionate about, they won't work as hard if at all.

Secondly- usefulness. People don't generally come to read a random book. They come to read up on a specific topic (say, differential equations) or a general topic (math). These topics will have a large amount of overlap- people tend to want to research the same thing. If people are frequently using a resource, it will eventually become good through user additions. So all the important books become good very quickly.

What about those who look for infrequently used books? One of two things- we have something or we don't. If we don't have it, we get a reputation for being useless. If we do have something thats at least decent, we win. Who knows, they may even know enough to add to it.

The only thing that really hurts us is if a very bad book exists for a topic. This gets taken care of by vfd. But this can and will happen even with a very restricitve policy. Instead what we need is some type of alpha phase so books aren't advertised til they're at least usable (which may be an interesting idea- an in development page for books being actively written?). Personally, I'm not adding my book math book to the math bookshelf until its at least half done- that way even if I never finish it will at least be usable, and in a good position to be improved. --Gabe Sechan 23:33, 20 September 2005 (UTC)


 * Just look at the vast majority of the books on our site. Most of them don't come anywhere close to the narrow gray areas excluded by AdSense. We would mostly just be turning away books that we delete anyway, just faster. If a person could only contribute to one of these undesirable areas (doubtful, if they are a motivated and intelligent individual) I don't think it's a great loss if they need to find some other resource on the web instead.


 * Also, when you count contributors turned away that otherwise would have contributed you must also consider new folks who will be attracted to contribute when they might have been turned off before.


 * But I should drive home the point that the specific exclusions we are talking about are a minority of contributions, and there are many advantages to consider by eliminating those forms of contributions. (Afterall, the "real knowledge" can be communicated in a book that is more suited to be instructional.)


 * In regards to your alpha phase comment, have you seen what I have to say about charters? The idea would be that the main page would show mostly chartered books, but might have an unchartered books section much like our "new books" sections. A book can start to be written before it has a charter, but once the charter is in place it will be easier for others to both contribute and make the book more valuable (because it increases the chances of the contributions remaining in the book and being appropriate for its scope). It would also lend towards better intrabook communication and coordination, so contribution efforts are even better allocated. --MShonle 00:12, 21 September 2005 (UTC)


 * I can see a couple of issues with the idea of charters
 * People in general hate administrative duties. Writing the charters would be a pain in the ass
 * It requires shelf owners to be and remain active members of the community. Or perhaps shelves make a decision as a vote of all owners of books on the shelf?
 * It requires active and swift decisions on charters. If we take a week to approve or disapprove, we'll lose contributors.  And it requires the right people to be making those decisions.  Same problem with the right people for shelf owners.
 * It might require software rights not in MediaWiki.  We'd need to give a small set of people not beauracrats or admins the rights to alter certain pages (shelves) but not others.  There's some code that can be leveraged to do so with a bit of code, but I'm not sure if full support is in version 1.4.9.  Although if it is I'll find it soon-  I'm actually setting up an internal wiki for my employer.
 * Here's the big one you'll be raked over the coals for- its not in the spirit of wiki. It might end up with better end results, depending on what your goals are- benevolent dictatorships tend to be more efficient at reaching singular goals than democracies and anarchies.  But the spirit of wiki is free submissions without a hierarchy making decisions.  A lot of people would see this as going against that.--Gabe Sechan 17:47, 21 September 2005 (UTC)


 * It's true that there would need to be an active shelf community. But this would not be a special right or require any software changes. You wouldn't be elected to it or anything: It's just people who are involved in that area. I would assume some people who like the higher-level ideas would be more about starting charters than perhaps the contributors. For example, it's similar to designing a whole curriculum. Someone might not want to author an entire curriculum, but would be very instrumental in designing what is said when and how.
 * I think largely we are already doing charters to some extent. Mostly it's by setting up the table of contents before hand, for those books that aren't mostly written by one person, that is. You can't get around the need to coordinate efforts (who wants to see insertion sort get covered in three places in the same book if it's as if the other explainations didn't exist?)
 * I agree that people would have a reaction against it (yet it amazes me people don't have a knee-jerk reaction against the NPOV requirement-- I think it's a wise requirement, but, man, that limits what and how you can write more than the AdSense limitations do, far more). But it's not against the spirit of the wiki. If someone starts to write MFC content for the C++ book, someone could just say "hey, that's not in the charter," and boom, they just create a new book instead.
 * I don't think the process would have to be swift. A charter is just saying what the book will talk about. Those involved in a shelf could whip out several charters in a go, essentially the same as making stub books. --MShonle 00:23, 22 September 2005 (UTC)

Ratings
WHile I don't oppose a rating system, I'm not sure what use it would be. Generally, you want a rating system to tell which is better between 2 like things- two books on C++, or two on Calculus. You don't compare Calculus and C++ books. Right now, we actively merge different books on the same subject. So there's no reason to compare.

In addition, its quite possible a ratings system could hurt books. If people don't read it because its rating is low, they won't contribute, and it won't improve. A low rating could act as a barrier to improvement, rather than a way to improve the average quality.--Gabe Sechan 23:33, 20 September 2005 (UTC)


 * Excellent point about like-things. To me I see the rating system as a way to avoid unsavory content, much like Slashdot's rating system. But it does only make sense in the context of letting any old troll or spammer write a book (in other words "all legal content is accepted"). But it seems we've already agreed we want more control than that, particularly in the cases of stopping forks from dividing contributions. --MShonle 23:59, 20 September 2005 (UTC)


 * Rating system- it wold be nice in having a rate system but it seems to me an impossibility (who will rate the books?). Since most books are unfinished works (I gess no book can or will be able to claim it's finished), and all are always mutating, how can any rating be valid over even a small period of time ?
 * Does anyone realy gives any attention on book ratings? (I don't, I go by references and quotes), another option would be creating in every book talk page a comment area, but that is almost there on the discussion, if anyone would comment a book why don't they edit, discuss or get involved on the writing ?!?
 * Forking does not divide contributions, fork wars does (like stamping a note for stoping other to contribute to one of the versions), preventing forks on site does provide a dificulty on moving contents, as all books are GFDL fork may even increase contribution and again, I point to you that every derivated work of any given book is a fork.--Panic 02:21, 21 September 2005 (UTC)