Hi @zacharyburnett, WebbPSF's associated data files are stored in this Box folder.

Hi @zacharyburnett, WebbPSF's associated data files are stored in this Box folder.

Thanks! However I can't access the folder, would it be possible to give me access?

Right, the folder is restricted but links to its individual files are public. Do you need more access than that? These are the links to the full and minimal files for the latest WebbPSF release.

Thank you! That's all I need.

the poppy tests pass on my local machine, but fail within the context of the macos-latest image, with the following:
https://github.com/spacetelescope/stenv/actions/runs/3182721627/jobs/5189366880#step:17:115

AssertionError: Pixel values different more than expected; max relative difference is 0.10739394972269882
assert 0.10739394972269882 < 0.001

also, webbpsf has the following test failure:

https://github.com/spacetelescope/stenv/actions/runs/3191216044/jobs/5207680817#step:18:56

>       s = glob.glob(surdir+'/example_alldof_A1-SM_sur.xml')[0]
[56](https://github.com/spacetelescope/stenv/actions/runs/3191216044/jobs/5207680817#step:18:57)
E       IndexError: list index out of range

@ojustino I apologize for the delay; I was working on infrastructure updates to make iterating tests faster for stenv. For this addition, I'm having some test failures with poppy:
https://github.com/spacetelescope/stenv/actions/runs/3235256747/jobs/5299649453#step:16:116

AssertionError: Pixel values different more than expected; max relative difference is 0.1073939497226991

and also an issue with webbpsf tests:
https://github.com/spacetelescope/stenv/actions/runs/3235256747/jobs/5299652960#step:16:57

# Test every DOF on A1-1 and SM and check the OTE state updated accordingly
s = glob.glob(surdir+'/example_alldof_A1-SM_sur.xml')[0]
E       IndexError: list index out of range

(I assume this is an issue with test data discovery, though I am not sure)

Are these failures important? I can ask pytest to skip these specific tests if not.

Hi all - this is NOT the right way to do this. Please do not add all the WebbPSF data files into this repository! That's unnecessary and furthermore would create a lot of hassles downstream with needing to keep in sync with the actual webbpsf data distro.

When astroconda was our supported distribution mechanism, @jhunkeler set up a webbpsf-data package within astroconda contrib which included scripts to retrieve the Zip files and install from there: https://github.com/astroconda/astroconda-contrib/tree/master/webbpsf-data

That was a great way to get the data files installed , without putting a lot of binary files inside the git repo. Maybe we can revive that approach for stenv?

Sure thing! I currently have the data files hosted via Git LFS, but we can set up a Box redirect to directly retrieve the files. How does build.sh download the data?

Oh, I see now that a redirect is set up already at https://stsci.box.com/shared/static/ftj8esrt0apzbnff8j6m5kseii2jzy9e.gz. Thank you for pointing this out, this makes it easier. I will replace the LFS additions with a download

it retrieves the file from the box URL. Not sure exactly what conda itself uses for the download (curl, wget, requests, or something else). I confess I personally never understood the exact details of how this was set up, but you can take a look at the files in the linked repo above.