Christoph Deil

Yes, maybe that would be the way to go here. I think the hard part will be to have a sane way to handle frame for FITS I/O for HEALPix...

@astrofrog - How about simply adding a `healpix/bench.py` file that can be run via `python -m healpix.bench` that runs some benchmarks with healpix and healpy and prints a report to...

Speed will always depend on machine, C / C++ compiler, Cython, Numpy, maybe even Astropy version. As a user / developer I want to know how fast this package is...

> it'd be nice to see at a glance how our speed compares to healpy for the current dev version Just put `python -m healpix.bench` at the end of the...

> Do you want to get this in before 0.1 is released? Yes, I'd like to have something minimal. I'll start a PR now.

I see you've added a test that runs the benchmark: ``` astropy_healpix/tests/test_bench.py::test_bench PASSED ``` It's not clear to me if this is useful. Benchmark will grow and become slow. FYI:...

I think we should only offer one API, the functions or the class. If we choose the class, then I think there's a few options to do it. One could...

> Maybe we shouldn't be comparing phi values for theta=0 or theta=pi? Yes, phi is meaningless at the poles.

@pmelchior - One thing I don't understand is why you in skymapper and the healpix guys are coding up projections at all. Why not just use `astropy.wcs.WCS` which is the...

Yesterday at the [Python-HEALPix-workshop](https://github.com/astropy/astropy-healpix/wiki/Python-HEALPix-workshop) @fxpineau mentioned that he has developed a high-performance HEALPix Java lib that includes disk and polygon query. It might become available later this year, maybe under...