This year's Reith lectures are a balm and a call to action. Highly recommended. www.bbc.co.uk/programmes/a...

Relieved to hear this. Thinking of you and yours this evening.

This was made by the gruniad for teens but honestly it’s the best explainer I’ve seen for anyone on the Aus social media ban www.theguardian.com/australia-ne...

Surely far better to keep the thing on so that your Amazon ring doorbell continues to work while you’re away from home. The odds of your router being hacked only while you’re away are considerably lower than the odds of your home being robbed only while you’re away.

Certainly one way we debug autoactive verification failures is to sprinkle assertions in, until we localise specifically what can’t be proved. So having a debugging methodology based around assertions rather than pre/post annotations seems reasonable

Right in which case having a strongest post condition at the point of the failing assertion is necessary. With wp you could instead tell the student “assertion on line 55 requires precondition to assert Y but instead it only asserts Y’” maybe ?

I think the answer is not independent of the question of whether your program uses pointers. All the symex based verifiers I know are for sep logics. But there is also a vcg based verifier for Viper, so symex is not synonymous with seplogic either

And a pocket sized edition of their meeting proceedings, called log logs

I also liked @fionakatauskas.bsky.social's commentary here: www.theguardian.com/commentisfre...

Does sharing that intuition require in-person contact? Record a video, put it on YouTube. Let it be free