Dan Allan

Results 756 comments of Dan Allan

Would it make sense to require the _type_ at initialization but allow the value to be this proposed `UNINITIALIZED` sentinel?

Rebased on master. #442 may be the key fix.

This might sway me to finally add a required dependency, one on https://entrypoints.readthedocs.io/en/latest/.

Lewis expert @MikeHart85 could tell us for sure. (He explained this to me in person during his job interview but I have forgotten the details since.)

Yes. I think that question is important for #453. For this, I think clients are always allowed to initiate the TCP phase with a server at some address they happen...

Specifically this break sub-array slicing zero elements: ``` ___________________________________________________________________________________ test_subarray_zerolen ____________________________________________________________________________________ pvnames = def test_subarray_zerolen(pvnames): subarr1 = PV(pvnames.zero_len_subarr1) subarr1.wait_for_connection() val = subarr1.get(use_monitor=True, as_numpy=True) assert isinstance(val, numpy.ndarray), 'using monitor' > assert...

In fact they are ASCII.

On Slack, @ronpandolfi weighed in with the slightly wider filled arrow: `▶`

Hello @pgierz! I'm happy to see someone pushing on this, and help if I can. I would suggest starting a tiled server in a separate process, from the CLI: ```py...