Martin K. Scherer

Isn't the expression string cast into a Sympy expression before serialization? If so, it would already be checked to be valid, right?

The traverse method needs to be implemented for the implementation classes, e.g. CSM and so forth, right? Or is there a way to hide this asdf related detail somewhere in...

It should be renamed due to political reasons, right? Honestly I think it would make a difference for repositories like NumPy or so, which is used by thousands of users....

This is for no reason a technical debate. For new repos it is now the default name. But default branches can be named arbitrarily. The default branch is being checked...

Sounds like a feasible solution to me.

Honestly I think it is dangerous the encode secrets for PyPI access in the cloud (owned and controlled by Microsoft). There are too many reports where patterns like this have...

Just my fifty cents: 1. Agreed 2. A diff tool for nested structures seems like good idea, we could use it all over the place. However it could be come...

This seems like a lot of effort for a use case which doesn't occur that often, right? I mean how often would you have exactly the same signal transformation setup?...

as a first workaround you could try to monkey patch internal ASDF structures. I haven't looked into the details, but this can be feasible until we have a working upstream...

In asdf-2.8 they introduced the new config class, which is also used in this PR. An excerpt from the asdf documentation: ``` Special note to library maintainers ----------------------------------- Libraries that...