February 27, 2018, Tuesday

• Multiple inputs
  • Test with MultiQC
    • Results ok but Data Common landing page permission is not set correctly
    • Inputs are not deleted so wrapper script needs to be modified
      • Inputs are not passed because a backward slash at the end of the link needs to be removed
      • How are inputs passed to input_id1? Will it work for ordered paired-end reads?
        • Inputs are passed as an array
      • input_id2, input_id3, ... need to be removed from app JSON
    • Make "+" disappear when maxCardinity is reached
    • Add "-" for removing
  • Need to rethink with mapping output id to filenames
    • It seems folder will create confusion for replicates and have to specify additional parameter if order is critical (e.g. rsem_de)
    • Leaning towards adding output id as prefix, e.g. ss_a.txt, sc_ss_a.txt, also check old notes
    • Filename will get very long but captures what has happened and we only need to decode the first prefix
      • Can make output id short or more meaningful?
      • ?
    • Tricky since output id is not passed to the wrapper script
    • Tricky for workflow diagram and history column
      • We need to display the full name of the output file or output id?
      • Workflow engine needs to search for full name given an output id
• Limit concurrent jobs with a global variable that counts submitted jobs not passing staged step
• issues on de.sciapps.org
  • Updates are not reflected since we have captured a static json. Shall we add refresh button?
  • Pending to Staging is not reflected on table, so it seems that we are not notified?
• Remaining topics for Friday