This page describes the process used to add the SYSTEM pseudo instrument and CALVER type. Since the task of adding a new instrument is rarely done it is performed manually adding each stage of the pmap->imap->rmap->reference file hierachy in the correct order. Since CRDS requires that all file references point to official files already in CRDS, ensuring "broken" contexts are not created, only some orders of submissions work. This is the process that was used to create and submit files in a viable order:

Create the SYSTEM CALVER type spec (empty .rmap with required header fields) and add it to the crds/jwst/specs directory to declare the new SYSTEM pseudo-instrument and CALVER type. This makes them valid in CRDS. Update the CRDS server with the new code.

Use submit mappings to submit the spec which doubles as the first empty rmap. Do not use Generate Context or Auto Rename, the SYSTEM imap doesn't exist yet.

Create a SYSTEM imap and assign the CALVER member the official name of the first empty rmap. Do not Generate Context or Auto Rename.

Download and modify the current "edit" .pmap and add the first SYSTEM imap to it by hand using the official name from the CRDS server. Re-upload using Auto Rename but not Generate Contexts. This should complete the process of fully plugging in the SYSTEM CALVER type in the CRDS rules, as an empty type.

Submit the first SYSTEM CALVER reference normally, a .json file, adding it to the empty rmap using Generate Contexts and Auto Rename to generate a new .rmap, .imap, and .pmap.

See also ticket 442 for more info on the overall bootstrapping process of CALVER.

Last modified 2 years ago Last modified on 01/31/17 08:10:31