Meeting 80 – 27 November 2014

Agenda

  • Actions from meeting 79 (Steve)
  • ICAT development (Steve)
  • ICAT Manager (Christophe)
  • IDS development (Steve)
  • TopCAT development (Wayne)
  • IJP development (Brian R)
  • Installation Script (Leonardo)
  • Issues from ICAT production deployments (Any)
  • 2015 F2F Meeting and Steering Group Meeting (Steve and Andy)
  • AOB (Any)

Notes

Apologies: None
Present: Matt, Tom, Kevin, Wayne, Steve, Brian R,Christophe, Milan, Andy, Rolf

  • Actions – No progress
  • ICAT development: Steve reported that “role” had been added to InvestigationGroup as agreed last meeting. The icat.server 4.4.0 would have been ready to go out except for the new idea about allowing multiple roles which could be added very quickly if people agree. It has been discussed today by e-mail. One more snapshot will be produced with this change and in a few days time 4.4.0 will be released.
    The next release (4.5.0) will include extra RESTFul calls for the benefit of TopCAT 2 making heavy use of the lucene indexing. Steve would like to drop the existing searchText call if nobody is using it. Tom will check with the Mantid people.  These three new calls are currently very specialised. Maybe a more generic way can be found in the future but this is not obvious.  The queries use the notion of “my data” which is defined as the existence of an entry in the InvestigationUser table (irrespective of role) to define the relationship between an Investigation and a User. The meeting was happy with this choice. The investigation call is complete and the dataset call is almost complete. The datafile code has not yet been started. It will be tested on a large scale with a clone of the DLS ICAT. Tom asked if DLS were ingesting via the API which will keep the indices up to date – they are for the bulk of the data but not for the “Investigation”.As the update rules do not work as might be expected Steve plans to check against the UPDATE permissions before the change and against the CREATE permissions after the change.  This is to make sure that update cannot get the system into a state that could not be achieved by a create call.
  • Christophe said that the ICAT manager’s update feature now worked proper;y behind a web proxy. He plans a new feature to allow multiple entities to be updated together.
  • IDS 1.3 is about ready to release as well. This will be done after ICAT 4.4.0 has been released though there is no dependency. A couple of bugs have been fixed since the last meeting and the mechanism to use a cryptographic hash to protect the datafile location has been added. Rolf told the meeting that he had submitted a bug relating to a possible race condition.
  • Wayne reported on his  latest mock-ups of TopCAT 2. He has received a number of questions from Rolf which he will address shortly. Wayne answered a question from Tom about the technology to be used and was told that the plan was to to use AngularJS with (probably) jQWidgets as it is preferable to use widgets that are free and support angular.
  • Brian reported that development of the IJP was going smoothly. He explained that when introducing datafile selection it was actually easiest to start be selecting available jobs and then you will be offered datafiles or datasets according to the type of job.
  • Steve was the only person present who had tried Leonardo’s latest installation scripts. He was a little concerned by the proliferation of bash scripts.
  • Kevin reported that the Vulcan laser ICAT and ingestion was now in pre-production and that plans were in hand for another ICAT installation for Artemis.
  • Rolf told us that a new version of python icat was about to be released once it had been tested with the final release of icat.server 4.4.0
  • Milan reminded us that a report by PaNData on ICAT experience would soon be made available.
  • Andy gave us the good news that the ESS steering committee had decided to use ICAT. They are also going to use ISPyB so we should consider how best to use them together.
  • A F2F and Steering Group Meeting for early 2015 was discussed. A host is needed. ESRF are prepared to do this unless someone else comes forward. It was felt that we should not attempt to co-locate with PaNDAas as this would delay things unnecessarily. Steve will send round an e-mail to see if anybody would like to act as host As Andy was not sure who was actually on the steering group, Steve will set up a mailing list
  • Finally Tom told us that this would be his last meeting representing ISIS as he is starting his new job as group leader in SCD replacing Brian Matthews on the 1st December.

Actions

ITEM DESCRIPTION ASSIGNED
80.0  Find out if Mantid is using the searchText call  Tom
80.1 Seek a host for the 2015 F2F and Steering Group Meeting Steve
80.2 Set up mailing list for steering group Steve
74.0 Look into JDBC connector resilience with Wildfly Steve
61.1 Put the simple ICAT browser into contrib when it is ready Condie
60.3 Finalise a release of IJP and send a note to the mailing list Brian R