Charlie Zender

Just add this line to all the function protoypes down to `nco_sph_plg_area()`: ` (rgr_sct * const rgr, /* I/O [sct] Regridding structure */ ` It's the best way to expose...

Thank you for fixing that. Let's finish the printing modifications before addressing subsetting and hyperslabbing, which will require some careful thinking about the front-end treatment of user-defined types. For now,...

XML needs `shape="*"` for VLEN to distinguish it from `shape= ""` for scalar. Same for multi-dimensional VLEN. Please add an asterisk as the final dimension for VLEN, e.g., ``` ```...

Regarding this: `vlen_flt_t vlen_int_scl:vlen_att_flt_1D = {1, 2, 3, 4, 5, -Infinityf, Infinityf}, {8, NaNf, 10} ; ` It is probably a minor error that ncdump CDL includes the trailing "f"...

BTW, above I asked you to use `enum4` instead of `enum8` for XML ENUM printing of `int64` and `uint64` without explaining why. Just so posterity understands, it's because the NcML...

Please elaborate what you mean by "incomplete". The handlers I wrote do not explicitly care about the rank of the VLEN. Comparing ncdump to ncks output it looks like the...

So the "missing" values were due to an ill-formed format statement. Everything printing as expected now. Remaining items include identifying VLEN missing values in time to avoid putting braces around...

FYI ENUMs are CDL-safe in 4.7.5. Removing braces from VLEN missing values is the last task remaining on this.

Correct. I have a proposal pending with NASA to add that support. They rejected the first version last year, let's hope this one is more successful.

FYI the current snapshot of NCO has nascent support for NCZarr. Most operations work for most operators using the` file:///` protocol. S3 support will take longer. Testing would be very...