Library

So, I went to the library in my hometown a couple of days ago looking for a book (I. Kant - Critique of Pure Reason) and while I was waiting in line to check that book out I was thinking about how inefficient and low-(|no-)tech the process is:

• go to the index drawers and look for the category where you believe that book may be palced
• look for the book in the subcategories
• find the book, write on a piece of paper the exact author and book name along with the identification number (which is something like 1/K20)
• head to the librarian and ask for the book

This is the case when you know the author and name of the book. Say, if you would want to search for a book on linguistics (language for example) you should go to the librarian and ask her to tell you what books do they have in with that name in the title. She would then search the book database with a tool running in DOS (it’s old and not very bright, but it doesn’t look like a FoxPro application though) and maybe find the book you are looking for. Should you find something that interests you she would then have to go and search the depot for that book, because there is no statistic whether a book is available or not.

There are also a few more possibilities (you going through the shelves looking for the book, using the author index, asking the librarian what books does she know to have regarding a subject, etc), but there really is a hell lot more space for improvement.

Looking through the news, as the National Library is on the verge on collapsing, or hundreds year old manuscripts are being kept in the basements of an insalubre building, I suspect money from the government is short - and they ain’t coming this way too soon.

So, what would contribute to improve the system? I was thinking about an application that would, on one hand, allow the librarians to manage the books in the database (and immediately know if a book is checked out or not) and on the other hand allow the users to search the database from a web interface (which can run both locally on a couple of low end computers — heck, you could slack up a Damn Small Linux on a 486 and you’re on — or on the internet) which should really decrease the time spent by a person in the library. You could know in an instant if it’s worth going all the way accross town by verifying if a certain book is in or not.

While I was at this I also found myself thinking about a novel UI for categorizing the items. We will consider the categories organized in a tree structure. You would have the root (general) node and then all the subcategories spanning out of it, then those subcategories would have another set of subcategories and so on and so forth. Now, given the infinite number of categories that can exist under the root node we can safely assume that traditional (Explorer like structures, if you like) ways of displaying this information might lead to lots of annoying operations, like trying to click the “+” icon next to a category, or having the items span all the way to the other side of the screen, both horizontal and vertical. So how about this one? You have a listing of the categories displayed in a vertical stack with a caption on top denoting the name of the current category. In the left you would have a button allowing you to go up the tree, in the right the subcategories available. Should you click a category or a subcategory it would go right in the caption box and display the appropriate data (beneath it, grayed out, the categories on the same level, next to it, the categories beneath it in the tree). When you would hover the grayed out buttons the UI should display a preview of the subcategories within. When you get to the desired category you would click a suggestive icon next to it. A nice feature would also be a “find as you type” category listing.

I’m still thinking about this one, I’m not really deffinitive about the way I would like it to look but I’m sure it would be pretty usable. Later.