I'm curious... how would you imagine this might break things downstream?

FWIW I love this kind of stuff and if properly documented would rather work around it than remove it, but nobody relies on my code so what do I know :).

Expect: a tool (from tcl/tk) used to write tests, whether we like it or not.

Expect that "gimme gimme gimme" is a cheap way to determine if the tests are running after midnight, nor not. If so, do more extensive tests; if not, do less burdensome tests.

I know - its lame. But, you cannot predict these things.

If you don't want people to use your Easter egg for something, don't include it in the release.. ;)

As described it only worked if you run it at exactly half-past midnight.

Because dozens of junior engineers have encountered this error before. However, instead of posting on SE about it they added a branch to the test case if it was running after midnight, and at least one of them added an assert to verify the output was "gimme gimme gimme".

I think if you're relying on the presence of an easter-egg in something you're using in production, you probably have bigger problems.

Don't forget the first rule of SW development. If a change results in user programs breaking, it's a bug. We never EVER blame the user programs.

/s

https://lkml.org/lkml/2012/12/23/75

I might suggest that man and the kernel have a different relationship to user programs.

There is no reason the philosophy can't be the same.

No, but there also isn't any particular reason to think it should be.

Chesterton's Fence: https://en.wikipedia.org/wiki/G._K._Chesterton#Chesterton.27...