Uploaded image for project: 'Request For Comments'
  1. Request For Comments
  2. RFC-688

Remove SuprimeCam support from obs_subaru

    XMLWordPrintable

    Details

    • Type: RFC
    • Status: Implemented
    • Resolution: Done
    • Component/s: DM
    • Labels:
      None

      Description

      obs_subaru's support for processing data from SuprimeCam has not been exercised since its successor HSC went on the sky, and at this point it's so bitrotted that reviving it would probably involve copying all of HSC's configuration and mapper files and modifying them again after just throwing away the relic SuprimeCam ones (or at least a line-by-line diff of the two). The SuprimeCam camera geometry may still be relevant - though it's unlikely to still be compatible with the afw cameraGeom interfaces.

      Regardless, any work we would do to revive SuprimeCam support would work as well if the bitrotted files we have on master today were accessible only through git history. We have nothing to gain by keeping them on master.

      We don't have a lot to lose by keeping them on master, but there is a cost, and I believe it exceeds the cost of removing them. We have a lot of configuration split between the config and config/hsc directories, making it harder to read and maintain, even though the distinction is entirely hypothetical. And just having dead files in the package makes it more work to browse and grep the live ones (which I was trying to do recently on DM-24485). It also contributes to obs_subaru's idiosyncratic organization (which we have already approved fixing on RFC-508 but never implemented).

      This RFC proposes that we simply remove (with no deprecation period; that's just extra work here, because we know no one is using this stuff) all SuprimeCam-only files from obs_subaru, and begin to consolidate - when convenient and backwards compatible, not on a triggered ticket for this RFC - any config files or classes split up only so they could be used by SuprimeCam as well as HSC.

      The reorganization proposed on RFC-508 would be amended to just put HSC files directly under lsst.obs.subaru instead of lsst.obs.subaru.hsc (as many are already there, this makes it less disruptive than the original RFC-508 proposal). As in that RFC, that reorganization would require a deprecation period, and I'd like to continue to consider it the implementation of RFC-508, not this one.

      This change would make it logical to rename obs_subaru to obs_hsc, but I am definitely not proposing that: the entire premise of this proposal is that removing SuprimeCam files won't break anything or otherwise be disruptive, and renaming the package absolutely would.

        Attachments

          Issue Links

            Activity

            Hide
            jbosch Jim Bosch added a comment -

            Done. I'll give that a bit of time to stew, and then adopt. I don't think this needs to block adding anything to release notes, but I also don't know anything about adding things to release notes.

            Show
            jbosch Jim Bosch added a comment - Done. I'll give that a bit of time to stew, and then adopt. I don't think this needs to block adding anything to release notes, but I also don't know anything about adding things to release notes.
            Hide
            tjenness Tim Jenness added a comment -

            It seems to me we can adopt this RFC. John Swinbank would you be the one to add a note to the release notes announcing this or would it be Gabriele Comoretto [X]?

            Show
            tjenness Tim Jenness added a comment - It seems to me we can adopt this RFC. John Swinbank would you be the one to add a note to the release notes announcing this or would it be Gabriele Comoretto [X] ?
            Hide
            jbosch Jim Bosch added a comment -

            No replies to the community post. Adopting.

            Show
            jbosch Jim Bosch added a comment - No replies to the community post. Adopting.
            Hide
            swinbank John Swinbank added a comment -

            I'm happy to do that.

            Show
            swinbank John Swinbank added a comment - I'm happy to do that.
            Hide
            jbosch Jim Bosch added a comment -

            I wasn't following the implementation ticket closely enough, and I didn't notice until after it had merged that we hadn't included merging config files on that ticket. But after thinking about it a bit more, while that's very useful work to get done, it's just as fair to say that it was blocked by this RFC as triggered by it, and that's the link I've used for DM-27255, which captures that work.

            Show
            jbosch Jim Bosch added a comment - I wasn't following the implementation ticket closely enough, and I didn't notice until after it had merged that we hadn't included merging config files on that ticket. But after thinking about it a bit more, while that's very useful work to get done, it's just as fair to say that it was blocked by this RFC as triggered by it, and that's the link I've used for DM-27255 , which captures that work.

              People

              Assignee:
              jbosch Jim Bosch
              Reporter:
              jbosch Jim Bosch
              Watchers:
              Jim Bosch, John Parejko, John Swinbank, Paul Price, Tim Jenness
              Votes:
              0 Vote for this issue
              Watchers:
              5 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:
                Planned End:

                  Jenkins

                  No builds found.