A  strategy for

BLISS

BCU meeting, 25th January 2018

EBS program - for BCU:

3 teams, 1 objective

Abandonning Spec

Welcome BLISS

It's not gonna be easy...

 

... but it's gonna be fun

Strategy for BLISS =

a global approach

1. Renovating Software Deployment

& running of services

How ?

  • Conda to install BCU software
    • blissbuilder becomes: building a Conda package hosted in bcu-ci.esrf.fr
    • blissinstaller becomes: "conda install" command
  • Supervisor to start servers and daemons
    • no need for bliss_dserver anymore

1. Renovating Software Deployment

& running of services

When ?

As soon as possible !

It's already started

 

Conda saved Antonia and me yesterday

1. Renovating Software Deployment

& running of services

1. Renovating Software Deployment

& running of services

2. Towards BLISS 1.0.0

2. Towards BLISS 1.0.0

Version 1.0.0 means:

  • API stability
  • homogeneity
  • basic set of features

2. Towards BLISS 1.0.0

Features list

  • Configuration
  • Motor control
    • controllers, virtual, motion hooks
    • trajectories
  • Sampling & Integrating counters
    • P201, Keithley, Wago, ...
  • MCA
    • all XIAs
  • Lima
  • Temperature controllers
  • Scans
    • ascan, dscan, etc.
    • continuous
    • HDF5 file saving
    • data publishing
  • Metadata (ICAT)
  • MUSST, OPIOM
  • PEPU, Speedgoat
  • Online visualisation          
  • HKL
  • Documentation
  • Stable API

2. Towards BLISS 1.0.0

Starting with 1.0.0:

- semantic versioning

- BLISS Conda package

3. 2018: assessing BLISS for various beamline experiments

3. 2018: assessing BLISS for various beamline experiments

- Adding missing features

- Testing in real conditions

- Preparing shutdown

- Early adopters: first user experiments running BLISS

 

3. 2018: assessing BLISS for various beamline experiments

3. 2018: assessing BLISS for various beamline experiments

Participation in Scrum sprints:

become a BLISS Product Owner for 3 weeks !

= you tell BLISS team what to do

... and then you continue with BLISS

3. 2018: assessing BLISS for various beamline experiments

Guidelines:

- control of an equipment, an experiment or an entire hutch with BLISS

- less Spec as possible (better if no Spec at all)

- no new Spec macro !

- scientists have to be volunteers

3. 2018: assessing BLISS for various beamline experiments

Gilles

Laurent

Gilles

Roberto

Antonia

Jens

Cyril

Final schedule to be defined with scientists (needs commissioning time !) and BCU contacts

3. 2018: assessing BLISS for various beamline experiments

Continuing "early adopters" projects started in 2017:

- ID30 (Tiago)

- ID11 (Sebastien)

- ID15 (Emmanuel)

- MX (Antonia+Matias)

3. 2018: assessing BLISS for various beamline experiments

Manpower issue

2 CODs to the rescue !

Hope to start recruitment campaign soon

4. 2019: Great Deployment

4. 2019: Great Deployment

BLISS to replace spec as much as possible in as many beamlines as possible

 

 

 

Guess what...

4. 2019: Great Deployment

In parallel:

- finalizing transition to Python 3

- users training

5. Graphical User Interfaces

5. Graphical User Interfaces

Online data visualisation

- based on silx

5. Graphical User Interfaces

BLISS shell for scientists

- final goal: integrated environment

5. Graphical User Interfaces

User Applications

- MXCuBE

- BsxCuBE

- ID16

- ID21

- ID17

- ...

5. Graphical User Interfaces

User Applications = web applications

5. Graphical User Interfaces

Manpower issue

 

- no one is responsible for User Interfaces development

- maintenance of existing applications !

- demand will grow within next years (post-2.0 era)

 

Conclusion

"They did not know it was impossible so they did it"

 

(Mark Twain)

A strategy for BLISS

By Matias Guijarro

A strategy for BLISS

  • 495
Loading comments...

More from Matias Guijarro