A first look at Diaspora's private alpha test

[crustyasp46]

Hands-on: a first look at Diaspora's private alpha test ( another social network )

The Diaspora project has launched a private alpha test of its open source social network. It is opening up its own hosted instance of Diaspora to a select group of testers, starting with people who contributed financial support when Diaspora was first getting off the ground. The initial group of participants can invite other people, and the developers will be opening up the service to more users each week.

Diaspora emerged as a response to the privacy concerns raised by mainstream social networking services. The aim of Diaspora is to create an open source social network alternative with decentralized architecture, giving end users more control over their private information and how it is shared with other people. The project was financed by individual contributions through the Kickstarter crowd-funding service.

When the first Diaspora code was first published in September, independent reviewers found some serious technical defects, including a number of security weaknesses. The developers say that the issues that were identified have been addressed and security has continued to be a major focus for the project.

Although the discovery of basic vulnerabilities obviously doesn't instill a lot of confidence, it's clear that the availability of source code and the transparency of development have made it easier for third parties to help find and fix those kinds of issues. In a blog post about the alpha launch, the Diaspora developers cite extensibility and code cleanliness as other high priorities.

Starting Cautiously

I got an invitation to test the hosted service (special thanks to Ryan Singer, who kindly gave me an invite) and was able to register an account. The Diaspora test server has already become a bit sluggish due to the increase in traffic caused by the private alpha launch. I initially thought that a private beta seemed like an odd approach for launching an open source application, but it makes sense in light of the early performance issues. The developers have opted to open up the service to the public gradually so that they can address the inevitable scalability challenges incrementally rather than having the server crushed on the first day.

At the current stage of development, the service is still extremely limited. You categorize your contacts into different groupings, which are called "aspects" in Diaspora terminology. The purpose of the aspect feature is to help you manage which of your contacts can see your messages and other private information. Diaspora aims to eventually have more granular and transparent privacy settings than Facebook.

When a user requests friendship in Diaspora, you will see a thumbnail of their profile icon in a requests sidebar on the aspect management screen. You can accept their request by dragging and dropping their icon into one of the aspect boxes. Because individual contacts can be assigned to multiple aspects, it's generally easier to understand if you think of aspects as behaving like contact tags.

The core functionality of Diaspora right now revolves around posting short text messages and photos. You can "reshare" and comment on individual messages. You can select an aspect from the tab bar at the top of the site in order to post a message that will be visible to only the people in that aspect. You can also post from the "Home" tab to send a message to all of your aspects. When you send a global message, you can also optionally choose to make it public, which will cause Diaspora to make it accessible through a public RSS feed and cross-post it to your external social networking accounts on Twitter and Facebook.

Unsurprisingly, Diaspora feels a lot like a less polished version of the Facebook activity stream. It offers a similar user experience, but with a stronger emphasis on contact groups. It's important to remember that it's at an early stage of development and is still maturing. It's hard to get a clear impression of how well the service accommodates discussion at this stage because there aren't very many users yet. The problem is exacerbated by the lack of contact discoverability. Although there is a built-in people search tool, it's currently quite hard to find your friends.

Unlike other popular social networking services, Diaspora doesn't seem to let you dig through the people who your friends are friending in order to find additional people who you might want to add to your contact roster. This characteristic of Diaspora's design is possibly an intentional privacy feature, but it diminishes the immediate usefulness of the service. It raises questions about how much of the efficacy of social networking tools is predicated on the willingness of users to compromise their privacy.

One of the most compelling aspects of the service is its built-in data export tools. From your account settings page, you can download an XML file with all of your accumulated Diaspora data. It also lets you download an archive with all of the images that you have posted to the service. These export features reflect Diaspora's commitment to ensuring that users aren't locked in or deprived of ownership of their own data. The account settings dialog also lets a user completely close their account.

Can Diaspora succeed?

It's still too early to make a judgment about the project's chances for success. There are a lot of rough edges, but it also has a lot of potential. If the developers continue to move it forward and manage to overcome some of the early performance challenges and add more functionality, I think it's likely that it will attract a niche audience of open source software enthusiasts, privacy advocates, and Facebook refuseniks. As we have seen before with the modest successes of the similarly focused StatusNet, it's not hard to get open source software developers to join an open source social network.

It seems unlikely that a project as small as Diaspora will ever have the visibility or appeal to truly compete with giants like Facebook, but it could still have an important positive impact on the industry by raising awareness of privacy issues and encouraging better interoperability among established social networks. It may seem far-fetched, but Diaspora (or something like it) could someday help to inspire change in the social network arena in much the same way that Firefox has helped to reinvigorate the browser market and accelerate conformance with open Web standards.

As the abuses and technical gaffes of the mainstream social networking operators contribute to growing concerns about privacy and autonomy in the cloud, it's possible that users who are sensitive to such issues will begin to appreciate the availability of more open alternatives. Even if the open source options never gain serious mainstream momentum, they have the potential to draw some attention to the underlying issues that they are trying to solve. Diaspora doesn't have to topple the entrenched giants in order to inspire positive changes in the industry; it just has to get a critical mass of people to start thinking more seriously about privacy issues and the right kind of interoperability.

If you want to test Diaspora yourself, you can request an invitation at the project's website or you can download the code from the public GitHub repository and install it on your own server.

[Hot Trout]

Interesting. I wonder if it would hold any merit for this site?

[Guest]

Everyone has to start somewhere. But as of right now I think Facebook still has a good set of years before its time is up and something new and truly innovative (again) is released to replace it. Until then, everything else is just dust in the wind.

[crustyasp46]

Source Open, What About Data?


The issue of open data is on the minds of a lot of technology users and watchers of late, thanks to the media blitzkrieg surrounding the WikiLeaks Web site and the legal battles facing WikiLeaks founder Julian Assange. In the face of such a controversial issue, it's only natural to ask questions about who owns data and what rights do they have to use and own it?

A more pedestrian — and far less controversial — application of the open data issue is this: who owns the data you put on more everyday Web services, such as Facebook, Yahoo!, or one of the many Google Web applications? The assumption "why me, of course" may actually not be true.

The Problem

This is because data on the various Web services is often locked into the service itself. If Gmail were to go off the air tomorrow, how many millions of users would completely lose not only the ability to communicate presently, but also lose any archived messages? Are messages regularly backed up? In fact, how does one back up Gmail? (Hint: use a standalone POP or IMAP client and pull down the messages to your local machine... and hope you have enough storage capacity.)

And that's just Gmail: what about all the other Google services? Google's own solution, the Data Libration Front is a good start, but who has an open source stack equivalent to Google's to which to move the data? Without such a stack, what good is exporting a Google user's data out?

Then there is the issue of control, another hallmark of ownership. Who sees the data on Gmail? What processes are looking at those messages in the Inbox, seeing a few occurrences of the word "habit," and concluding an ad for a twelve-step program would be appropriate to display?

For all the emphasis the open source community has on the freedom surrounding software used, that same community is often willing to sacrifice the same measure of freedom for their data — a far more personal aspect of their online lives.

Open data is a solution to the problem that many users may not recognize they have: how to keep their data accessible and controlled in an environment where that data is increasing online and typically out of control and (at times) inaccessible.

Creating an Open Web

One strong proponent for open data on the Web is Stormy Peters, the former director of the GNOME Foundation and currently head of Mozilla's developer engagement program. Peters often speaks about the importance of open data, addressing the topic at this year's OSCON and at the Ohio LinuxFest, where this reporter attended her presentation. Peters cited several examples of services where data can get out of control. Facebook, for example, notoriously holds onto user data rather tightly, to the point where it's difficult for the users themselves to completely export the data to another service, should they choose. Not to mention Facebook's ongoing privacy travails.

For the average user, Facebook's privacy policy has expanded significantly since its beginnings and opting out of individual privacy changes is a byzantine process that is very time consuming. There are some third-party solutions: Matt Pizzimenti founded ReclaimPrivacy.org this Spring as one way to specifically deal with data and privacy in Facebook.

Using ReclaimPrivacy.org is simple; just drag the bookmarklet up to your bookmarks, log into Facebook, and surf to your privacy settings. Once on the page, click the bookmarklet link and off the script will scan for slack privacy settings.

Facebook is just one example in a very wide world of data and privacy problems. Many Web-enabled services used today have a less-than-open data policy in place.

Businesses have a vested interest in open data, as well. Cloud-based services are hugely popular, and vow to protect customer data. But terms of use policies must be closely examined to determine exactly what happens to customer data if the terms of service are violated. What recourse does the customer have? Are their appeal processes in place for dispute? And--most critically--what happens to the customer's data if the service is cut off? Entire businesses could collapse within days to get data access sorted out during a terms of service dispute.

Peters notes some positive exceptions in her presentations: Identi.ca, the open source microblogging service, not only has open code, but open data policies as well. Similar openness is expected to be found on Diaspora, an open source version of Facebook's social services now in alpha development.

But will "open" be enough of a draw? Web service users might not want to go through the hassle of moving to more open systems just for privacy's sake, which will certainly lower customer pressure for open data. And, with so much money involved with mining and keeping customer data, don't expect openness to magically spread through service providers' policies without significant customer pressure.