>
You own your data.I was super excited when I came across that first sentence in bold (italics here) on one of the points under "Here’s what makes Talk different".
Wow, I thought to myself, does this mark yet another step in the revolution of the re-decentralization of the internet? Is Talk a product sharing the same ideological foundations as awesome projects like Sandstorm (https://sandstorm.io/), Matrix (https://matrix.org/), remoteStorage (https://remotestorage.io/), Solid (https://solid.mit.edu/), and many others that I'm doing a grave injustice to by not mentioning here (please do remind/enlighten me in the comments if you're aware of others!), that strive to give back to users the ability to control their own data to whatever degree technically feasible: where it lives, who has access to which pieces of it, how they may access/manipulate it, under what circumstances, and for how long, etc, etc.
The concept of apps built on user-provided and user-controlled data-sources, envisioned by projects like these, has always been immensely appealing to me. If users truly controlled their data, and only granted apps access to the data they need to function, instead of depending on each individual app to host user data in their own locked-off silos, then switching to a different app would be a simple matter of granting another app access to the same pieces of data! Lock-in would completely stop being an issue!
Imagine that! We could have a healthy and highly competitive app ecosystem where users choose apps by their own merit instead of by the size of their moat built on nothing but network effects. Newcomers could unseat incumbents by simply providing a better product that users want to use! Like a true free-market meritocracy!
And then I read on:
> Unlike the most popular systems, every organization using Talk runs its own version of the software, and keeps its own data. Talk doesn’t contain any tracking, or digital surveillance. This is great for journalistic integrity, good for privacy, and important for the internet.
And I realized, No, they were not talking to me, a lowly user hoping against hope to one day live in a world where we can truly own and control access to data that rightfully belongs to us. Rather, they were talking to platform owners like Washington Post, loudly proclaiming that with Talk, that they get to own my comments, my profile, my data, and that somehow it is "good for privacy, and important for the internet".
Sadly, this is just yet another data point in the unrelenting trend in silo-ization of user data, and stopping that trend is becoming more and more of a distant dream because both newcomers and incumbents today realize the massive competitive advantage lock-in and network effects afford them. Incumbents will never give up their moat and allow the possibility of interop without a fight, and newcomers all end up racing to build up their own walled-off data silos because they have ambitions to become an incumbent enjoying a moat of their own one day. Even products that are built on top of open protocols and allow non-trivial interop tend to eventually go down the path of embrace, extend, extinguish, once they reach any significant scale.
Regardless, for this particular instance, my rational mind realizes that platform owners are the group that Mozilla must appease and market to, in order for Talk to stand any chance of success, and that Talk as a product is definitely the lesser of many evils in the same space, and that fundamental change usually manifests itself as a series of small, incremental improvements, much like in this case where an open-source, privacy-respecting product manages to defeat numerous much more deeply pocketed and deeply entrenched rivals to secure a foothold in a competitive market. But as a longtime Mozilla Fanboy, however irrational it may be, I can't help but feel a little bit betrayed when even they are forced to send messages like this in a painful compromise that confirms my lingering suspicion that we're still a long way off from winning the battle for a user-centric internet.