Google, Microsoft, Facebook and Twitter team up on data portability project

Tech giants collaborate to reduce lock-in - yes you read that right

Four of the world's largest technology companies have come together on an open source project designed to make transferring data between competing services much easier.

The Data Transfer Project (DTP) was started by Google last year. It has now been joined by Microsoft, Facebook and Twitter, who together released a white paper on Friday.

Rather than having to download data from one service provider and upload it into another - lengthy and difficult where there are large volumes, low bandwidths and proprietary formats - the idea is that a user could simply move their photos, videos, favourites lists, blog posts or membership data to a new provider in a couple of clicks with participating providers taking care of the process.

The white paper provides an illustration of what the interface might look like.

In this hypothetical example, a Google Photos user wants to move their photos from Google to Microsoft OneDrive. They go to Google's file transfer interface, choose the destination, and hit ‘send.' They then must authorize the transfer using both services' chosen methods, in this case OAuth. The selected files are automatically copied and routed to the destination, without using either bandwidth or hardware of the user.

The existence of project might come as a surprise given that the organisations involved are better known for trying to lock people in than helping them to leave. Some might suspect an attempt by the big players to carve up the market between themselves at the expense of smaller providers. However, the white paper addresses this saying that the project is open source to encourage participation of as many providers as possible, and that "small providers would reap relatively larger benefits from participating, as the DTP reduces the amount of engineering work".

More likely it is in response to legislation such as GDPR (not mentioned specifically in the paper) which stipulates data portability among its provisions, the recently published interoperability and portability standard for cloud (ISO/IEC 19941:2017), and an acknowledgement that decentralised identities and multiple platforms sharing common protocols are the way things are going.

"The DTP will enhance the data portability ecosystem by reducing the infrastructure burden on both providers and users, which should, in turn, increase the number of services offering portability," the paper says.

"This concept of allowing users to choose products and services based on choice, rather than being locked in, helps drive innovation and facilitates competition."

DTP consists of three components: Data Models, Adapters and Task Management Libraries. The Data Model for transferring files consists of a file type and the additional metadata needed by the receiving provider to accurately import the data. Adapters convert file and authentication formats to make them usable by the system, and the Task Management Libraries handle the background tasks necessary for the transfer to take place.

The system uses as many existing APIs, standards and protocols as possible to reduce the number of core dependencies and avoid creating new proprietary silos.

"We are really excited about [Friday]'s announcement as I think this will generate a critical mass for the project and attract a lot more contributors to help make direct service-to-server transfer easy for users across a wide variety of online services," said Google Engineer Jessie 'Chuy' Chavez on the project's Google Groups page.

While it's still early days, the project bodes well for identity-centric ideas such as the personal information economy, in which personal data remains in the control of the owner not the service provider.