@joelbrownstein Will do. This made me realize I need to update some of the docs for tree and sdss_access for DR19.

Also, now that Riley has updated the files for the dashboard for DR19, we need to find a final location for these on the SAS, and create a new envvar and/or path. Do you have any suggestions? We'll need to serve everything in this directory, https://data.sdss5.org/sas/sdsswork/users/u6054929/.

I moved Riley's sdsswork/users/u6054929 -> https://data.sdss.org/sas/dr19/mos/explorer/
Except I skipped sdsswork/users/u6054929/ipl3, since this appears to be replaced by sdsswork/users/u6054929/dr19; and "may" included leaked data? Let me know if you need that or if its ok to skip it.

Actually, I realized that we only serve these files via the docker, with Riley's directory mounted inside as a fixed top level explorer datapath envvar. So we don't technically need to release these files. If we release them, then they need datamodels, right? For time's sake, I would say we not release them and leave things as is, but what do you think?

A minimal alternative would be to only release the dr19 files, as you've done, but we don't document them, and we continue to point the dashboard to Riley's user directory. The dashboard expects all releases to be available from the top level datapath.

And yeah, I think the IPL3 files were never updated and do contain the extra data. I think we can safely replace those with the dr19 versions.

I suggest we separate the discussion of Riley's users directory from this PR.

So if you don't have any questions on the allspec file's paths or the other changes committed, then would you do the merge and tag tree and sdss_access please?

Sure but I don't need to tag for the readthedocs to update. The latest docs always point to the latest main. https://sdss-access.readthedocs.io/en/latest/paths.html is already updated. When I tag I want to make sure it has everything we need for DR19.