The Radio observatory sometimes generates lots of parsets of observations lasting just a few minutes. These can last a long time, and usually have the same array configuration, with other parameters changing between the various runs.
In order to prevent the control system from kickstarting the entire AARTFAAC chain every few minutes, currently observations lasting <10mins are dropped from the queue. This lost observation time could be usefully gained if we knew that we have already triggered on an observation of the same array configuration. However, this requires carrying state information over multiple parsing of the parset files. [Not sure, this could already have been implemented!]
The Radio observatory sometimes generates lots of parsets of observations lasting just a few minutes. These can last a long time, and usually have the same array configuration, with other parameters changing between the various runs.
In order to prevent the control system from kickstarting the entire AARTFAAC chain every few minutes, currently observations lasting <10mins are dropped from the queue. This lost observation time could be usefully gained if we knew that we have already triggered on an observation of the same array configuration. However, this requires carrying state information over multiple parsing of the parset files. [Not sure, this could already have been implemented!]