We propose per-user cross-origin cloud storage, much in the sense described in http://www.w3.org/DesignIssues/CloudStorage.html We are a non-profit project and have so far defined a first draft of our standard for this: http://unhosted.org/spec/dav/0.1 We have researched a lot of aspects in the last few months, and are about move to version 0.2 of our standard. People are starting to implement this with significant user base sizes, and other people are starting to develop apps that rely on it, which is now would be a good time to make this into a w3c cg.
Note: Community Groups are proposed and run by the community. Although W3C hosts these
conversations, the groups do not necessarily represent the views of the W3C Membership or staff.
Chairs, when logged in, may publish draft and final reports. Please see report requirements.
Some people will arrive earlier, but we officially start on 6 September with a touristic walk through Prague.
7 and 8 September are Unconference: anyone can prepare a talk, and we will record all talks and publish the videos afterwards. As a venue we want to use Square, the bar on the village square, which has great burgers, cheap beer, and a good space for us with a projector.
9 and 10 September will be hackathon: last year, we noticed so many good ideas came up during the Unconference that it was a pity we couldn’t stay another couple of days to implement them. That’s why we added two days of hackathon this year.
We will be organising at ESWC a workshop entitled “Downscaling the Semantic Web”. The call can be found at :
http://worldwidesemanticweb.wordpress.com/downscale2012/
The topics to be discussed, and the larger goal this initiative takes place in, are related to the work done on unhosted in the sense that we aim at enriching a set of small devices (XOs, Plug PCs, …) with data sharing capabilities. Eventually turning all them into many small data servers – running owncloud, for instance.
We would be very pleased to discuss ideas and ongoing projects of this community group during the workshop. Please don’t hesitate to contact us if you have any question related to this event 🙂
Our current mind set as Semantic Web engineers is to focus on centralized and very powerful solutions centered (and depending on …) around Web hosted servers and (mobile) clients accessing it. As a direct consequence of that, the usage of Linked Data depends on the availability of a Web infrastructure compassing data-centers, high speed reliable Internet connection and modern client devices. If any of this is missing, our community is not able, yet, to provide any Linked Data enabled data management solution. Still, the digital divide that is currently widely recognized separates the world into those who have access to Web-based platforms and those who don’t.
When designing Linked Data platforms we tend to forget those 4 Billion persons who don’t have access to Internet but would benefit from being able to use Linked Data. We should keep everyone in mind when we design Linked Data platforms and aim at helping to reduce this digital divide.
This web site is a collaborative effort to document steps made into that direction. Please feel free to ask for an editor account if you want to contribute.
OpenLink Data Spaces (ODS) [1] is another item for the list of open source platforms aimed at unhosted (de-centralized) “points of web presence”, also known as data spaces.
Note, WebDAV is integral to ODS, which is basically a platform developed using OpenLink Virtuoso [2][3]. Thus, you are able to perform tasks such as:
Use WebID ACLs to protect resources and collections.
Associate collections (folders) with Virtuoso back-end Quad Store so they act as a simple import mechanism.
Associate the folders above with Sponger Cartridges, which means Linked Data transformation of non-RDF resources precedes final persistence in the Quad Store.
Associate virtual resources with DAV collections, e.g., results of SPARQL queries.
Here’s how I put the above into use re. G+.
Go to Google Takeout to get a dump archive (ZIP).
Mount Virtuoso or ODS Briefcase DAV folder to my desktop, phone, or tablet.
Extract the ZIP archive.
Drag the entire expanded directory tree to my mounted folder.
I have Linked Data in my target Virtuoso instance at this point; thus all the data is accessible by a plethora of Linked Data or basic HTML browsers.
I have made a public demo since I want step 3 to be optional — i.e., I should just drag and drop a ZIP or GZIP archive to the special DAV folder, leaving Virtuoso to handle the rest of the effort to import to the Quad Store 🙂
remotestorage/spec
