If anyone knows of specific hacks, you can list them here. For instance, we can remove anywhere that uses NUMPY_LT_1P6.

In any case anyone is working on a patch for this, I think all of the hacks starting on line 56 can be removed from here:

https://github.com/astropy/astropy/blob/master/astropy/io/fits/py3compat.py#L56

But I need to double check that (against Numpy 1.6 at a minimum).

Also there is an instance of _NUMPY_GTR_15 being used, so that code can be removed.

I just grepped on NUMPY in the astropy source tree and there are a number of relevant hits in table and io.ascii where numpy 1.5 (or < 1.6.x) is being special-cased. I think most of these are straightforward to take out.

Note that #2992 requires numpy > 1.6, so it needs to at the least remove the test in 1.5 before it can be merged. before this issue was created I pushed up a couple of commits that update the minimum numpy version, but I could roll that back to just removing the 1.5 test from travis (and then this issue would actually bump up the officially-listed minimum).

I don't really mind either way, but I do really want to see #2992 merged ASAP, as it's blocking other critical work.

Just FYI: I'm working on this (I've assigned myself, but just in case there's confusion).

PR attached.

There is one extra place to remove, but I didn't feel like I could quite navigate it: io/fits/column.py:_scalar_to_format says it only exists for Numpy 1.5 support. Cc: @embray

I don't know if we've started a "What's New" page for 1.0, but this should definitely be in there too. Do we at least want to open an issue for that and start a checklist?

@mdboom I'm not sure if anything really needs to be done for astropy.io.fits.column but I'll take a look.

@embray - see #2935

👍 from my perspective, nice!

👍 from me too - I can rebase #2992 after this is merged - @mdboom addressed my concern about this delaying #2992 by just doing it all right now!

Also @taldcroft - I'm not surprised you're happy with this: it's roughly a 7:1 ratio of lines removed to lines added. 😉

There have been no objections so I think this is fine to merge 👍

I've deleted the unused imports mentioned by @embray.

There have been zero objections to this, so I will go ahead and merge so that @eteq can continue the ERFA work. Thanks @mdboom!

Thanks @mdboom !