That Neubot update due one year ago (eh, ...)
That feeling when you discover you have not updated the blog at neubot.org for more than one year: you cleary feel very miserable.
Being a professional procrastinator, then, I then spent some more time asking myself why I did not update it in such a long time.
Who said procrastination?
I could now list all the reasons I found, including that I do not like working with Drupal, but I fear this may be a bad use of your and my time, as it would certainly be more interesting to write about what we did.
And so, here we are, Neubot update: October 2015 — January 2017!
During this period of time, and especially since we were awarded an unrestricted research grant from Google Research, I have mainly been working on measurement-kit. With (in alphabetical order) Davide Allavena, Carmine D'Amico, Arturo Filasto', Antonio Langiu, Lorenzo Primiterra, and others, we have been steadily improving the library.
The most exciting features developed have been: the implementation of the network diagnostic tool (NDT) network performance measurement methodology and the work done to support the first implementation of OONI for mobile devices (a fact that is already hitting the news even though we've not yet pushed the publish-the-app-on-all-stores button).
On the Neubot "proper" front (i.e. on the development of the tool called "Neubot") we made less progress. What satisfies me is that we were able (with help from Nathan Kinkade of Measurement Lab) to publish a new stable version of the Neubot server on Measurement Lab servers. We are moving towards a point where the server and the client are two separate pieces of software, with the server written in Go and including a NDT server.
Botticelli is not only the name of our new Go server, but also the surname of Sandro, a quite famous Italian painter during the Renaissance (you may have heard of him...)
Speaking of the client, I am less happy of what we have achieved, but I also reckon that probably these not satisfactory developments have been more caused by lack of resources than any specific wrongdoing on my part. To be honest, there is an ongoing witch hunt inside me, trying to find culpable parts, but I guess I'd better save this discussion for my Emacs psychiatrist.
Talking to the doctor was not useful either. But at least now I have installed Emacs and I know how to tell the macOS terminal that option is meta.
In all seriousness, what was very positive on this front was the robust internal discussion we had a Nexa over Neubot in the last few months. At the end of this discussion, to which I'd like to dedicate a separate blog post, we drafted a plan for moving forward the development of Neubot (indicating with Neubot specifically the client) so to reach a specific milestone (read: a stable release) where the software is cleaned up and workable again.
This is all, for now. I'll blog again soon to comment OONI mobile from my perspective and to explain the Neubot plan.