dybr

Esteban helped me today to install all kind of unusual (company-specific) shit in docker. Then I got to te code. Well, so, three tests are still failing. Esteban is trying to convince me that they crash for everyone. But Jeremey says that they do pass for him. Also, Jeremey has taught me to run integration tests via `sdc ic-test` - never heard!

I just wanted to intercept logs in the tests and view them; ok, I do intercept them, and there's nothing there, since all this is done in a wrapper, and our tests are unittests, and don't log anything. Kind of reasonable, but I have to figure out.

This company's code... it has so much "functional programming", totally inadequate, in the style of this idiot Travis Brown (who manages to create a left adjoint functor to the sum functor - Tony Morris was right saying that Travis never understood anything in ScalaZ).

The thing is, there's a certain number of companies that got hooked on Scala when it was a fad. Now they'll have to stay with Scala for good. Meaning, I'll always have a job. (Ha-ha-ha, that was stupid. Good that it ended, though. - 09/21/2025) But in general, I don't advise it. They are a little bit out of mind too much. Today one really smart person showed to me on twitter how one can, via λP2, rewrite parameterized types in the form of a type with an internal type. And now Scala 3, and Martin, its Father, plans to move Scala in this direction. Why not, we have dependent types, therefore we can.