Open keithralphs opened 4 months ago
Sorry, I did not get to this one in the hackathon.
I'm interested in looking into it but have lot's to do on J20 too. @coretl let me know if you want me to spend some time on this this week.
I'd prefer you look at J20, I'll see if I can get round to it.
When I work out what's wrong I also need to pull in the docs from https://github.com/bluesky/ophyd/pull/1136 and amend
An attempt at fixing it here: https://github.com/dls-controls/aioca/tree/contexts
Using the first seen context in all event loops, but keeping the channel caches separate
When using the python asyncio teminal (python -m asyncio), with the Run Engine, it's possible for timeout errors to occur when running abluesky plan because the ents for the plan get sent to the event loop for the asyncio environment rather than the Run Engine Causing a Tiemout exception and lockup(TomC tells me).
To reproduce this situation using my Ophyd-async Synchrotron device implementation (See i22-source branch, devices folder) :
This has apparently bee addressed in iPython environments using