Lars Bärring

This issue will be relevant to #3388 if a data producer adds the same non-standard attribute (such a `comment`) to multiple coordinate system variables.

Sorry for crashing into this conversation without having digested the technical discussion in any detail. But here is my two pennies worth of user perspective: 1p: The most important thing...

Following up on what @pp-mo just wrote > ... having the same attribute recorded as both local and global, can be useful, and is seen in practice -- despite possibly...

As one "user in the room" I venture some some thoughts: 1. It seems useful to keep a "legacy mode" as an alternative to a "new mode" (or maybe even...

I think that this would still be a useful extension of Iris, despite that the more general question dealt with in #3388 was closed.

Does this PR also take care of what is discussed in [CF/cf-conventions issue #223](https://github.com/cf-convention/cf-conventions/issues/223), i.e. to explicitly state the coordinate axis ordering in the content of the ``grid_mapping`` attribute of...

Yes, I think that this is still relevant.

I may be out of my depth here, but is a [direct] hash of floating point values really useful? I imagine that some rounding, possibly user-adjustable, would prevent unwarranted down-to-last-bit-agreement...

Yes, I am afraid that is the case (without knowing exactly how it is done), which has bitten us several times (e.g. [here](https://github.com/SciTools/iris/issues/3559#issuecomment-1403658618)). Why enforce that level of precision when...

Aah -- this sounds promising :-)) pinging @ljoakim for info