search

NSLS-II / nslsii

NSLS-II related devices
BSD 3-Clause "New" or "Revised" License
11 stars 22 forks source link

Add config from CSX-1 #8

Closed stuwilkins closed 6 years ago

stuwilkins commented 6 years ago

Attn @ambarb @wen-hu @cmazzoli

This is the initial cut from the profile directory into this structure. I would like to review all that is in here and see what can be replaced / refactored and what is obsolete.

ambarb commented 6 years ago

@stuwilkins I am a little confused. Why are the files in ../startup no longer numbered? Are we going to hope alphabetical order is enough to always keep things running in the correct order?

ambarb commented 6 years ago

I also cannot find the tardis setup. Did you leave that out on purpose or can I simply not find it? I've looked in ../ophyd/devices.py and /startup/endstation.py.

stuwilkins commented 6 years ago

@ambarb

Thanks! I was thinking that we will define in __init__.py

from .detectors import *
from .accelerator import *

so that we have better control over the imports. This way, our profile would be:

from nsls2tools.csx1.startup import *

Only!

I didn't add the tardis yet as it was all commented out.

ambarb commented 6 years ago

Nice idea to import it.

Are you looking at 98-tardis-mu.py? This is the one we use and these things are in sd.baseline so they have to be defined.

stuwilkins commented 6 years ago

Thanks, just added. I was confused by 98 and 99 ... anyway it is in now.

stuwilkins commented 6 years ago

@danielballan @tacaswell @stuartcampbell

Is this what you were thinking when we discussed this with @ambarb a while ago in the conference room?

stuwilkins commented 6 years ago

@ambarb if you are in, lets try and get this done Tuesday AM and see if it works, if it does then this is a model for SIX......

ambarb commented 6 years ago

Sounds good. I was able to get fix some of our custom plans so that the header information is populated. Just need to fix the plot scaling. I will need some help from DAMA with that because I don't know how to pass that variable to bec. @stuartcampbell can someone help me with this tomorrow?

stuartcampbell commented 6 years ago

Someone should be able to helpout... @danielballan @licode is one of you available ?

danielballan commented 6 years ago

I really like how this is looking. I'm wondering whether the subpackages bluesky, ophyd, and analysis, are the organization we want to go with long term. I can imagine some cases where things might not fit neatly into just one. Would it be better to move all those modules in the top level (csx) package? No strong feelings here -- just brainstorming.

I think it does make sense to have a startup subpackage so that we can from nsls2utils.csx1.startup import *.

stuwilkins commented 6 years ago

Thanks @danielballan

I was thinking that perhaps the distinction should be devices and plans rather than the actual package names? What do you think.

I like the idea of separating the locations of these things

danielballan commented 6 years ago

I almost made that exact suggestion myself. Let's go with that.

Get Outlook for iOShttps://aka.ms/o0ukef

On Wed, Sep 27, 2017 at 6:02 PM -0400, "Stuart Wilkins" notifications@github.com<mailto:notifications@github.com> wrote:

Thanks @danielballanhttps://github.com/danielballan

I was thinking that perhaps the distinction should be devices and plans rather than the actual package names? What do you think.

I like the idea of separating the locations of these things

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/NSLS-II/nsls-ii-tools/pull/8#issuecomment-332668621, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACLIrndJ4IO_euWQ7eTW_dPu6j4G2XD2ks5smsXwgaJpZM4Pi99d.

stuwilkins commented 6 years ago

@stuartcampbell @danielballan

This is now a fairly stable config for CSX. I think it is ready for review comment. Could someone take a look. I am keen for this not to be some monster PR as well!

danielballan commented 6 years ago

Haha, too late! But sometimes monster PRs are necessary -- the goal now should be to merge this promptly. I'll take a look.

ambarb commented 6 years ago

Add a logic test to the fccd configuration attrs to determine if the actuator for dark images is in or out.

danielballan commented 6 years ago

This content will be moved into csxtools and/or back to the IPython profile.

@stuwilkins Do you think a *tools repo for your program would be useful, or is it better to put things either in csxtools, sixtools, or scikit-beam?

tacaswell commented 6 years ago

This has moved back to the CSX program for now.