Is Open Execution the new Open Source?

Fifty is the new forty :-). Visiting Lake Como this winter brought back some nice memories of 2006, when, at the second OSS workshop, Giancarlo Succi and myself started the IFIP Working Group WG2.13 on Open Software, under the benevolent eye of IFIP TC2 Chair Robert A. Meersman and the other WG chairs (I believe Erich J. Neuhold was one of them), who had joined us and held their Technical Committee meeting on the lake.  

I believe Tony Wasserman, later to become our thought leader and now Chair of WG 2.13, was not with us on that day in Como; but we could count on top industrial specialists from the start (to name but two, Massimo Banzi and Ari Jaaksi). For many of us, the whole point of OSS was not "using software for free": it was the collective control over source code that open development communities could guarantee.

To this day, I still like to think that only crowd-sourcing code walk-through can guarantee software that holds some non-functional properties, including the key one of being harmless to its users. At the time, I was so sure of this that I even wrote a book (with Claudio Ardagna and Nabil El Ioini) on certifying open source software security properties. 

We were not alone: large companies quickly saw the interest of certified, industrial-strength open source platforms, and after the certification of Red Hat Linux, consortia like OW/2 (https://www.ow2.org/) were started to deliver them.

Some skeptics objected that establishing consensus on software non-functional properties via the **indirect** path of open source code development, inspection and testing (or even by analyzing formal models derived from code) was too burdensome and error prone.

In the end, I think the skeptics had a point: the discussion led nowhere, and with the exception of Linux and a few others, I would say that open source platforms are not much more trustworthy than proprietary ones whose code-base is inaccessible.

But there is still hope: we can dream of a **direct** mechanism to trust open software to deliver the non-functional properties we want. Can Blockchain-based execution deliver it? More and more people think so (read Khaled Salah's - and my own - take on this: https://blockchain.ieee.org/newsletter/december-2018/open-execution-the-blockchain-model). Still, there is still a long way to go. I would love to hear from the open source champions on this one..