Peter Lindstrom

Thanks for letting us know. Glad to hear it's been fixed.

@subhasis-shell Thanks for this contribution. We've been working on a complete overhaul of the CUDA backend on the `staging` branch, so it will be difficult to pick up this PR...

I see four options: 1. Use the existing support for 32-bit integers with the `zfp_promote_int16_to_int32()` and `zfp_demote_int32_to_int16()` [utility functions](https://zfp.readthedocs.io/en/release1.0.0/low-level-api.html#utility-functions). This, however, requires making copies of the data. If you're OK...

See the third bullet above. Currently, FP16 and bfloat16 can be handled similarly to `zfp_promote_int16_to_int32`, but with the user performing the conversions (e.g., to/from `float`). We do eventually want to...

Thanks for the suggestion. This might be a good idea, but it would have to be done portably. I believe this attribute (and `-Wimplicit-fallthrough`) is not supported prior to gcc...

@r-barnes Take a look at 6814a4e and see if it meets your needs.

@markcmiller86 This looks interesting but I'm not sure how to best expose zfp through such an API since zfp fundamentally operates on multidimensional arrays, whereas the API seems geared toward...

Thanks for bringing up this issue--we should be more careful about how we detect data models. I am a little puzzled about your solution, however. In the LP32 data model,...

I think we need to do a better job detecting the data model, and in a portable manner. I've begun some work on this and would appreciate some feedback. I've...

Good point, but I don't know of anything in zfp that depends on the storage of pointers, i.e., `sizeof(void*)`. Can you please elaborate?