Who owns your data?
A large part of the hype about Web 2.0 is about the ‘cloud’ i.e. allowing a 3rd party company to look after your data for you (e.g. Hotmail, Google Documents) and potentially add value by allowing you to connect socially with others (e.g. photos stored in Facebook or Flikr). Your data is always available to you, from any HaCk3D PC (or mobile device) no matter where you are in the world and you’re protected against losing your data if your PC has problems.
This cloud is turning out to have a dark lining, brought to attention by Yahoo’s leaked slide indicating that social bookmarking site ‘delicious’ would be closed as part of Yahoo’s streamlining of its services. This sent a wave of fear across the web, as people began to contemplate that the thousands of bookmarks that they’d built up over years of use could be summarily deleted as part of a corporate restructuring. It seems that our cloud based data may not be as permanent as the Hello hype suggests. A similar situation occured with Google Wave, which has ceased active development: although to their credit Google have introduced a feature to export waves (albeit 10 at a time), and is looking for alternative hosting for existing waves. And back at troubled Yahoo, yet another similar situation is developing in Yahoo Video, which is set for closure. All user uploaded content is due to be deleted on 15th March 2011. Again to their credit Yahoo have introduced download functionality as a result, but users must retrieve their content by the expiry date to avoid deletion.
Facebook is currently the corporate poster-child, but we cannot anticipate what will happen in the future: MySpace was sold for a huge sum and is struggling to provide a return on the investment (yesterday announcing the cut of half their workforce). If Facebook floats and is then overtaken by a younger nimbler rival, we cannot know what could happen to all the data we’ve uploaded to facebook over the past few years. In both the hypothetical Facebook example and the real-life delicious scenario, the problem is aggravated by the lack of good tools provided for exporting your data from their service. It’s easy enough to bulk upload content over time, but the only option for extracting photos from facebook is viewing and saving them one by one (unless users find a 3rd party application). Given that many people upload photos directly from their phones to facebook, the web-based version may be the only copy they have. In this circumstance people would expect to have greater ability to access and extract their data, rather than being beholden to the provider or forced to turn to 3rd party tools.
An of ideal here would be for providers to provide better tools and APIs for users to withdraw their data wholesale. The concern from the provider’s perspective is that users can then withdraw their data and move to other providers – hence the lack of these data backup / export tools in the first place.
The alternative to this Catch 22 situation is that mooted (separately) by Tantek and Jeremy Keith of hosting your own content and then pushing or re-posting it to other cloud services (such as delicious). Another example of this is a personal cloud such as Tonido as described at CodeLathe. wholesale mlb jerseys Self-hosting the on master copy avoids the total loss of data if a cloud service closes, but has been held back by the current poor tooling available for working in this manner [note I haven’t looked into Tonido in detail yet, so this isn’t a comment on their product!]. Also as Jeffrey Zeldman observes, we can retain *our* content if a cloud service closes, but we cannot preserve the social relationships connected to our data. In some cases (Google Documents / Flickr photos) this may be no great loss, but the content of more socially active sites can be rendered pointless by the lack of context: in the case of Twitter you’re left with a one-sided conversation…
Presently neither of these options are particularly satisfactory. A good tool to master content from one place seems to have a niche, which is presumably some of the intention of Google’s ill-fated Buzz service. Any rise in this approach is likely to come from highly technical early adopters before cheap jerseys spreading to the mainstream and in the case of Google Buzz it seems that no-one cheap jerseys China really understood what it was for. The likes of client apps like TweetDeck may be able to expand to claim this ground if they develop their range of functions to include:
- publishing of multiple content types: not just textual status updates / tweets, but photos and video to appropriate services
- methods for storage of all created content either locally or perhaps to multiple cloud storage services (e.g. dropbox) for redundancy.
In parallel with this, better support for exporting data from cloud services is in the interests of users to be able to move cloud service provider in a similar manner that they might switch their bank or energy provider. Considering this analogy, maybe legislative intervention would be required to enforce switching in this manner, as it is in the provider’s interests to lock users into using their service and to create barriers to exiting. And who would be in a position to legislate across the internet in this manner (but that’s a future blog post perhaps)?
Either of the considered approaches is likely to need significant user demand before the appropriate tooling, legislation or provider policy changes will develop. Whilst Yahoo’s handling of delicious has highlighted the issue of data ownership and access, delicious is a service generally used by tech savvy / web 2.0 pioneer users and so hasn’t provoked significant comment in the mainstream media. Data loss or closure of a cloud service used by a significant mainstream userbase will probably be needed to bring this kind of pressure to bear.
It difficult to envisage which service is likely to close that could bring this kind of mainstream public attention. Obviously if Mark Zuckerberg decided to close facebook that would have the right level of impact (as illustrated by the response to the recent facebook closure hoax by Weekly World News). Potentially MySpace closing would garner this kind of publicity (although most users will probably have moved on already and one would hope(!) that musicians have master copies of their songs external to MySpace). Alternately perhaps a major webmail provider being closed due to commercial pressures, especially if its a non-core activity for the owning business (the likes of a Hotmail perhaps?). In the case of webmail there are tools for exporting mail (POP3 / IMAP), but these are possibly beyond the knowledge of the average ‘Mom and Pop’ user that is likely to attract the level of public outcry required to change practices in this area.