Uploaded image for project: 'Request For Comments'
  1. Request For Comments
  2. RFC-551

Rename MakeCoaddTempExpTask to MakeWarpTask

    Details

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

      Description

      RFC-283 never quite finalized a name for the Task that generates the objects it renamed to Warp. Given the exact same arguments supplied in that RFC, and the discussion in DM-4634 (which was sadly closed as won't fix), we should rename the task to better match its outputs (and be less of a mouthful).

      MakeWarpTask seems perfect: it makes warps.

      It looks like this task is being reworked for butler gen3: this rename could happen as part of that, or just before it. There are only 90 references to makeCoaddTempExp in the stack (pipe_tasks, pipe_drivers, obs_decam/cfht/subaru, plus a couple lines in ci_hsc), so the change itself should be easy to make.

        Attachments

          Issue Links

            Activity

            Hide
            swinbank John Swinbank added a comment -

            Is this RFC now implemented?

            The answer to that seems rather nuanced.

            Most folks running a pipeline today will still be running MakeCoaddTempExpTask. However, MakeWarpTask exists, and is functional (modulo things like DM-17062) in Gen 3-land. I guess that means this is only really implemented when we make the switch to Gen 3 middleware for regular processing and deprecate the old task.

            However, for now we have no timeline or tickets for that switch. I'm not sure that holding this RFC open until those tickets exist and are marked as done is actually useful. So maybe we just accept this as implemented now? Thoughts?

            Show
            swinbank John Swinbank added a comment - Is this RFC now implemented? The answer to that seems rather nuanced. Most folks running a pipeline today will still be running MakeCoaddTempExpTask . However, MakeWarpTask exists, and is functional (modulo things like DM-17062 ) in Gen 3-land. I guess that means this is only really implemented when we make the switch to Gen 3 middleware for regular processing and deprecate the old task. However, for now we have no timeline or tickets for that switch. I'm not sure that holding this RFC open until those tickets exist and are marked as done is actually useful. So maybe we just accept this as implemented now? Thoughts?
            Hide
            yusra Yusra AlSayyad added a comment -

            I agree that the status is ambiguous (and in the same state as RFC-285). We have a partial replacement for makeCoaddTempExp. You could argue that this is fully implemented when we retire the Gen 2 version, tutorials are updated to the Gen3 API, and the word 'coaddTempExp' expunged from the code base and docs.  The high-level plan and timing for retiring Gen 2 will be discussed on Friday.  Or you could argue it's fully implemented when MakeWarpTask reaches feature-parity with makeCoaddTempExp (after DM-17062). (I don't care either way; the spirit of this RFC
            was already in motion and will become reality whether we close this RFC now or not.)

             

            Show
            yusra Yusra AlSayyad added a comment - I agree that the status is ambiguous (and in the same state as  RFC-285 ). We have a partial replacement for makeCoaddTempExp . You could argue that this is fully implemented when we retire the Gen 2 version, tutorials are updated to the Gen3 API, and the word 'coaddTempExp' expunged from the code base and docs.  The high-level plan and timing for retiring Gen 2 will be discussed on Friday.  Or you could argue it's fully implemented when MakeWarpTask  reaches feature-parity with makeCoaddTempExp (after  DM-17062 ). (I don't care either way; the spirit of this RFC was already in motion and will become reality whether we close this RFC now or not.)  
            Hide
            tjenness Tim Jenness added a comment -

            Either we close it or we add a new "is triggering" ticket that indicates the extra work that is blocking this. Both of these options remove the ticket from my radar.

            Show
            tjenness Tim Jenness added a comment - Either we close it or we add a new "is triggering" ticket that indicates the extra work that is blocking this. Both of these options remove the ticket from my radar.
            Hide
            tjenness Tim Jenness added a comment -

            John Swinbank do you want to make a decision on this RFC? It's in limbo. As I said previously, we either close it or add another triggering ticket.

            Show
            tjenness Tim Jenness added a comment - John Swinbank do you want to make a decision on this RFC? It's in limbo. As I said previously, we either close it or add another triggering ticket.
            Hide
            swinbank John Swinbank added a comment -

            John Swinbank do you want to make a decision on this RFC?

            Not particularly, but thanks for checking.

            If you insist: I'm marking this as implemented, on the basis that

            • we don't intend to schedule any further work to address this RFC specifically;
            • John Parejko's wish will come true at some indefinite point in the future;
            • holding this RFC open until then just seems like an exercise in box ticking.

            John Parejko, if you disagree with that, feel free to tell me about it.

            Show
            swinbank John Swinbank added a comment - John Swinbank do you want to make a decision on this RFC? Not particularly, but thanks for checking. If you insist: I'm marking this as implemented, on the basis that we don't intend to schedule any further work to address this RFC specifically; John Parejko 's wish will come true at some indefinite point in the future; holding this RFC open until then just seems like an exercise in box ticking. John Parejko , if you disagree with that, feel free to tell me about it.

              People

              • Assignee:
                Parejkoj John Parejko
                Reporter:
                Parejkoj John Parejko
                Watchers:
                Christopher Waters, Hsin-Fang Chiang, Jim Bosch, John Parejko, John Swinbank, Kian-Tat Lim, Meredith Rawls, Paul Price, Tim Jenness, Yusra AlSayyad
              • Votes:
                0 Vote for this issue
                Watchers:
                10 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved:
                  Planned End:

                  Summary Panel