Jun 11, 2021 by Joanna Rutkowska
On behalf of the whole team, I’m proud to announce the first public release of Wildland client! This is a reference implementation of the Wildland protocol, which we have described in our “Why, What and How” (aka W2H) paper.
Leaving aside all the usual disclaimers about how early-beta and for-power-users-only this version really is, I’d like to focus on what’s already working and possible :)
Firstly, you can connect multiple storage backends (S3 buckets, WebDAV servers, Dropbox accounts, etc) and expose them all as one unified file system as shown on the screenshot below. We believe that this is how data management should really be done. Users should not be concerned with which service provider their data is currently being held on. Instead, we should be able to address our data using infrastructure-agnostic addressing. Whether our files live on my Dropbox, local NAS, some p2p-synced storage, or on my local disk, should be of secondary importance, and should not be included in addressing of the data.
Secondly, Wildland natively implements what we call multi-categorization. This means that every data container (the smallest unit of information on Wildland) can be assigned more than one address/path and can be accessed using any of them. Functionality-wise this is similar to having a directory on a traditional filesystem with multiple links, except that on Wildland every path is a first-class citizen. We believe that this feature is vital in making filesystems powerful enough to allow for their use as a primary tool for information and knowledge management. This is a broad and exciting topic and we will be talking more about it in future posts.