1st BLISS Steering Committee meeting
Beamline Instrumentation Support Software
4/4/2019
News from the BLISS team
+ 3 newcomers (since January, February 2019)
L. Pithan
P. Pancino
P. Guillou
M. Guijarro, C. Guilloud, S. Petitdemange
BLISS team role: BLISS core development & participation to deployment on assigned beamlines
News from the BLISS team
Work organisation : kanban, see our gitlab board
Prioritized
todo
Analysis phase
Coding
Integration tests
Ready to merge
new features
bug fixes
backlog
News from the BLISS team
Sharing knowledge, newcomers integration
Daily stand-ups at 9:30
Peer programming
"Stop and Solve" meetings
Guidelines, good practices
News from the BLISS team
Acting for Quality Assurance improvement
Automatic tests coverage
(from 28% to 33%, +5 points increase since December, 2018)
Systematic code review
Only merge code with documentation
(from 5.568 lines to 10.028, 80% increase since December, 2018)
More integration tests on beamlines
Dealing with the technical debt
Technical debt = Tetris game
(you can't win)
Trying hard to mitigate its effects, have to pay the debt from time to time
Results report, 4/4/2019
Porting from Python 2.x to Python 3.x
High priority items from the BLISS development tasks list
Package and installation procedure
Logging, hw comm. debug feature
Motor control API
Configuration, parameters handling
Missing:
- guidelines
- user dialogs for parameters
BCU training done
BCU training done
BCU training done
Results report, 4/4/2019
User Experience enhancement
Scanning documentation
Scan data handling improvement
New Wago module and interlocks, Tango server
Counters framework
Asked during feedback meeting with scientists:
Typing helper, short names / aliases, user-friendly error reporting, scan output in a separate panel
Meeting with Veijo + Armando to check validity of HDF5 file, to be organised for the end of the month
Idea of refactoring saving - DAU could take over responsibility of saving HDF5 files from BLISS transient storage
!
!
High priority items from the BLISS development tasks list
Missing: documentation
1st BCU training done
(there will be more)
Results report, 4/4/2019
- Integration of F. Picca's library
- common base classes for diffractometers,
- configuration,
- "spec-like" functions,
- 4-circle implementation tested on BM28
- HKL continuous scans tested with Icepap+2D detector
Diffractometers support (work by E. Papillon)
- hkl info in hdf5 files
- real, physical motors position saving in addition to H, K, L
- save / load complete diffractometer parameters, and crystal parameters
- documentation
- test suite for BLISS CI
- Six-circle diffractometer implementation (to be tested on ID03 diffractometer)
!
scheduled in May
done
Results report, 4/4/2019
Nothing prevents the start of deployment, as foreseen
(and it has started)
Navigating the ship between icebergs
Bugs hunting
Technical debt, QA
Changes: new ideas, failed ideas
Many controllers to integrate (motors, devices, etc.)
Training (within BCU, to users...)
Problems and how to help us
Documentation problem
- poor English
- finding the good structure
-
very time consuming
- drawing schemas
- writing content, articles
How much quality do we want for the documentation ?
How to achieve the goal ?
A technical writer would help
Fighting pressure
- accepted projects involving development from BCU raise expectations
Sending a note to BL responsibles to say again BLISS is the top priority ?
Reminder: it would be very useful to keep beamline devices and computers operational, so we can do our testing and to be able to rely on something stable