hipcat labs
designing effective personal data ecosystems for people on the Web
The Web and Cloud Computing
The Web has radically changed nearly every thing that we do, from how we work, to the ways that we socialise and manage the trivialities of our lives.
But even as people's lives are becoming enhanced through the convenience of millions of information resources, products and services, people's personal data are being released and spread across dozens, if not hundreds, of different services. These services include the "big" sharing services such as Google, Dropbox and Facebook to activity-specific services such Twitter, Nike+, Mint.com, Amazon and countless others.
Fragmentation, Access, & Control
Three problems arise from this move of personal data into the current set of cloud-based platforms :
Fragmentation
Information gets split up among multiple separate sharing and storage platform(s), making working with and access to the data difficult and inconvenient.
Access
Services dictate ways people access their data, including the protocol(s) and data format(s) they have to use. Also, services may not guarantee retention and access guarantees on their critical data.
Control
Users have limited control over how their data is represented, stored and organized, meaning that people may not be able to store the data they want, in the ways they want on particular services. Moreover, in many cases, users may not have control over how their data is used, handled and to whom it is disclosed.
In some cases (except in the EU where it's required by law), people may not even have access to the information being collected about them or stored on their behalf (e.g., Medical records).
Personal Clouds: Our Approach
hipcat labs is currently leading the design, development and testing of several experimental community-driven open source projects to improve the state of the art of personal cloud computing.
The following projects are being done in conjunction with dancak.es.
Nodebox (was HappyCloud)
Nodebox is an architecture that makes it possible for people to control and maintain full ownership over data they keep in the cloud.
Instead of writing to one particular storage API or platform as a service, Nodebox takes the less is more philosophy. Inspired by jQuery plug-ins, Nodebox is a simple mix-in that lets web pages uniformly query a large variety of different cloud based data stores and services. Nodebox relies on node based proxies to transform proprietary interfaces into a single, uniform, CORS-enabled single-auth REST API with CouchDB-view semantics.
Nodebox can also encrypt all of a user's information before it is sent to ensure that the cloud provider has no access to the content of data being stored.
Nodebox's github repository will be made public soon!
WebBox
WebBox turns the Web inside-out: the web server/database are placed alongside the browser. Multiple Web servers (web boxes) form an overlay network and message each other to keep data up-to-date and allow data to be shared in a decentralized fashion. The data fragmentation problem is eliminated because all a user's data is kept consolidated and replicated across a person's own machine(s).
Find out more about WebBox on Github or from our paper. (WebBox is not yet feature complete, but is an experimental work in progress only for hackers!)
- List-it
- Popular and practical note-taking tool for Firefox developed to study note-taking in situ. Supports taking free-text notes
- Atomate
- Personal automation framework that allows RSS feeds and web-based data feeds to trigger routine behaviours, such as reminders. Presented at WWW2010 and set a precedent for services such as ifft.
- Twiage me
- A Game with a Purpose (GWAP) for allowing players to filter and rank Twitter content to find tweets that offer "good advice"
- Poyozo
- In-browser personal diary/lifelog that tracks all pages you view on the Web and summarises information consumption for later retrieval and re-discovery.
hipcat are looking for collaborators and sponsors! Please get in touch if you would like to join, propose, or fund a project ~ your contributions will be put directly towards these MIT-licensed FOSS projects.
Currently, hipcat labs is based at the the University of Southampton, and are friends with the W3C, Haystack at CSAIL, and Harvard ProjectVRM.
To find out more, please contact eMax at max (at) hip.cat or see previous papers.