txtr (1.22.2) release date: 09:57 02.02.2012
The Reaktor stores and processes documents. A document has a unique document id, abbreviated udid (or sometimes uid). A document always has metadata (title, author, etc), organized into sets of metadata fields, called metadata categories. Most documents have binary data associated with them as well: document versions contain the actual binary document data. A version looks exactly like a file (say, a PDF file) you'd find in your file system.
Documents in the Reaktor exist independently of txtr.com and are not necessarily visible on txtr.com. To make them visible on the web, you'd have to add them to a view. Views are lists of documents that show up on txtr.com as collections. Views are contained in view sets that show up on txtr.com as tabs.
Most of the API requires authentication. You need a txtr user ID and a password to authenticate.
You can get an account by signing up on txtr.com.
You will, upon successful authentication through a call to WSAuth.authenticateUserByName,
receive a token that allows you to access the API. Tokens should never be published. They will time out
30 minutes after the last call to the API. (Sticky tokens will never time out, so use them with care.)
Normally, you should authenticate, use the token for as long as you need, and then de-authenticate in an orderly manner.
Don't worry if you don't feel confident about all this by now. We have code samples ready to guide you through all of this, and there is a starting points page that has some howtos.
(We had a tough choice to make when deciding on the approach for the Reaktor API: Should we give developers full access to the same stuff we were using, or should we do a stripped-down, simplified version to hide complexity from you? We decided to give you the whole thing and try to do a good documentation job, so let's hope we got it right.)