Opened 16 years ago
#1385 new defect
wrong mask file may be used for chip stage stamps
| Reported by: | bills | Owned by: | bills |
|---|---|---|---|
| Priority: | low | Milestone: | |
| Component: | pstamp | Version: | |
| Severity: | normal | Keywords: | |
| Cc: |
Description
The postage stamp parser has a number of optimizations that attempt to optimize parsing. One of these is that the camProcessedExp data for each exposure is cached. This avoids multiple invocations of camtool.
Since the postage stamp spec says that the 'best available astrometry' is to be used for sky to chip coordinate transformations the latest camRun for any particular exposure is used for the astrometry. These data are cached by the parser and are reused if a subsequent request row needs astrometry from the same exposure.
Recently I switched the resolution of the mask file to use the camera stage masks. This avoids some nasty destreaking management when chip files are cleaned and updated. The camera masks aren't cleaned and thus once destreaked are always available.
However, since the camRun data is cached, for a byid chip request for a chipRun other than the latest for a given exposure the newer mask file will be used for stamps and NAN masking.
Since usually the latest chipRun is the one of interest to users (IPP processing is always getting better) and the latest is the one chosen as the result of bycoord requests this probably isn't a big deal.
