Looking over the nearing-final versions of several Task docs
Get group feedback on them
Any learnings for template modification
Take a quick look at the LSST SWUG pages and think about what material from there may be of value to edit and transfer to pipelines.lsst.io
Item
Notes
Current single-frame processing pages content
Better to condense the stuff in the description, make it more 'contentful' and less prolix
Link directly to where relevant, e.g. to a page describing the type of calib'd imgs that come out of processCcd
Be concise, make sure the really significant stuff is right at the start of a bullet point
Streamline all the info in isrtask, charimg, calibrate, take out the 'sequential' descrips, these aren't useful
Maybe some of the current content could go into comments in the code – look carefully at this and judge judiciously
Action: review talk by JSick at Aug Tucson mtg for how people absorb info off webpages
Action: when ready, figure out who we want to review this version of the pages (by doing a PR on branch DM-8717)
Argparsers/Taskrunners
Perhaps these will be best dealt with when the docstring-extraction code is written (by JSick and Adam) by making a new rst directive to extract the info and put into the pages
There probably has to be a modified Task Template type for the Tasks that use these functions to incorporate these sections
Action: for now, go through the primary list of Tasks on the doxygen pages and figure out which ones use argparsers
Jim Bosch's comments on DMTN-030 (in CLO)
He asks for a section on init'zn – bit unclear on this request
Action: We all will need to review these carefully, maybe ask for more detail
Definitely useful stuff in both the tutorials and 'info pages' here
But much may be outdated, and we will need to read, test, and curate carefully to make sure we keep/transfer the stuff of value
Because much may be written in a different tone/style than we now want (which is more concise and dense with info), let's not let it bias us as we write new material
Action: for now, let's write a disclaimer at the top of each page pointing people to PLI (pipelines.lsst.io) or CLO (community.lsst.org) for current info