Coursera à la Google: Best practices for ML engineering, found via O’Reilly Data Newletter (see the associated matrerial on ML).

Rethinking Twitter. Fair points, although I don’t share his views on the “Like” feature. I’ve used it a lot in the past (and still today) to bookmark links, instead of storing them on yet another social sharing website.

Another nice post by John D. Cook on differential privacy: Comparing Truncation to Differential Privacy.

Falsehoods programmers believe about Unix time. Or why you need to worry about leap seconds.

It’s been 10 years that I subscribed to GitHub (and a little less to Bitbucket), and I’m slowly rewriting (yet another time) and uploading my past lecture notes to public or private external repositories (--theirs in git parlance). That’s quite an exhausting fucking life, isn’t it? I hope I will done for good with my past (life) after all that…

  New Order, Movement.

Still the same routine after 3 months, I’m afraid it doesn’t get any better: when I get home, I’m useless and all I have to do is rest for at least 1h30.

Another nice post on diffusion models: Going Critical. See also this older post of mine. #dataviz

Things you’re probably not using in Python 3 – but should (via @newycombinator). #python

  Micah P. Hinson, Micah P. Hinson and the Gospel of Progress.