The real problem that the percloud wants to solve, and why it's still necessary

Believe it or not, I only discovered arKos last Friday, through this Slashdot announcement: a project (apparently) very similar to the percloud, which is my own proposal for a Free Software alternative to Facebook, Gmail &C.

Following the links from Slashdot I discovered this interview to the arkOS developer and even more projects in the same space that I didn’t know: buddycloud, Personal Clouds and unhosted.

update 2013/10/08, 10:20am GMT+1: I discovered just now the IndieWebCamp projects. Great project! Practically everything written below about arkOS and Personal Clouds applies to them too.

After a look at those projects and a few email and tweets exchanges, including the explicit question “why not just help the FreedomBox Foundation, instead”, I came to two conclusions.

First, I’m happy that all these projects exist. On one hand, they prove very well my point that now is THE moment for personal clouds. On the other hand, they make my own work much easier and more likely to succeed (if it does start, see below) because they are already doing parts of it.

In the second place, just because of what I read about those projects, I still believe that there is space and need for the percloud as a separate effort. Before explaining why, let’s deal with the FreedomBox question (which was already a FAQ anyway…):

Q: why not just help the FreedomBox Foundation, instead?

  • because @FreedomBoxFndn itself seems uninterested, and that’s PERFECTLY fine, of course!!
  • because it’s a bit like asking “why don’t help Debian to become Ubuntu, instead of forking it?”
  • I have no problem to help whoever is working towards certain goals. That’s why I explicitly said since the beginning that all my work would be “Free as in Freedom” code and documentation. However, at this point my only feasible way to help is to get paid to do the percloud “phase 1” myself

What about those other projects?

arkOS is “a Linux-based operating system… designed to run on a Raspberry Pi – a super-low-cost single board computer – and ultimately will let users, even of the non-technical variety, run from within their homes email, social networking, storage and other services”. It also seems a very flexible, general purpose environment, not a locked-down (=much simpler to use) one. Buddycloud is (emphasis mine) “a set of tools, open source software and protocols to help you build a completely new kind of social network.” Unhosted is about “serverless”, “client-side”, or “static” web apps. “Personal Clouds” is, if I understand correctly, a great, complete ecosystem of web apps, (tools to enforce) user-controlled Terms of Service agreements, network services etc…

The percloud, instead…

The percloud is just another GNU/Linux distribution, and this is a good thing. It will certainly be possible to run it also on a Raspberry Pi, or any other computer hosted at home, but I do not want to tie it to any specific hardware device. I want to build one single blob of software that can run on everything from data centers to home computers, with the smallest possible set of external constraints or dependences.

ICT experts will tell you that only a cloud running on computers you physically control can provide the greatest possible privacy, or that stuff like buddycloud or “Personal Clouds” could become much more complete and scalable than the percloud. Let me say one thing: they would all be right!

In my opinion, however, a really portable, software-only, relatively “quick and dirty” percloud as I am proposing shoud still be done and widely adopted for (at least) these reasons:

  1. If somebody is sleeping in a burning house, you don’t wait until another house, at least as good as that one, is ready: you wake them up and tell them to get out NOW, to take shelter in whatever refuge you can set up in a hurry. PRISM and friends prove that we should really start to deploy realistic countermeasure soon. We can’t wait until greater but complex platforms or, even worse, actual alternative Internets like this are ready. 2 It is certainly not done on purpose, but ANY version of “running your cloud on some hardware you own and keep at home” is limited to minorities. We can celebrate Raspberry Pi for being “so cheap” (terribly relative term, don’t you think?) all day, but the reality is that only people with affordable and reliable electricity and flat rate broadband and reasonably high confidence that their home hardware wouldn’t be stolen or sequestrated could use arkOs as proposed. And that’s not even the biggest issue, because… 3: in my opinion the real, or at least the most urgent problem, is social and psychological, not technical. While the real solutions to PRISM-like issues are not technical, we can’t get there unless a lot of average Internet users are willing/prepared/able to get there. We need awareness and confidence much more than “platforms”.

Today, most average Internet users can’t see at all how replacing with something open the corporate walled gardens in which they currently live could ever be within their reach. Or why they should want it in the first place. I want to prove to how many of those users as possible, as soon as possible, that they can live online outside those walls. Why should they care if their first “refuge” may not be everybody’s ultimate, perfect digital home, since they could leave it whenever they wish for something better, without losing their data?

Let average Internet users get something that is really easy to use and, in many cases, perfectly adequate as-is for all their needs and skills. Do that, and they will all become both able and, often, willing to move to other, more complete solutions when they (both the solutions and the users…) are ready. But tell the same people that they have to either buy/configure/use extra hardware, or that they have to enter more than a handful of configuration parameters and they’ll NEVER get started.

This is why the percloud, by design, won’t be some ultra-flexible environment able to do whatever cloud computing you may want. It will, instead, be the simplest possible replacement for the main cloud activities that the majority of cloud users needs, with as little as possible initial configuration.

Final note for software hackers:

Software hackers who already run their own servers won’t need a percloud, but they should still recommend it to all their non-hackers friends. It would be the most realistic way to make sure that all email and other content they exchange with those friends does not end up on some centralized server that makes centralized, large scale surveillance much easier.

Of course, nothing of this will happen…

(not from me, at least) if you don’t whatever you can to fund phase 1 of the percloud. Thanks for your support!