TANGO at ALBA

1
TANGO at ALBA
Witness of a new dancer

• Alba team Presented by DFC.

2
Experience
3
Experience
WE HAVE TO ADMIT THAT WE CHEATED A LITTLE BIT
4
Controls groups

• Software
• Tango administration
• (Tiago Coutinho.).
• Device Pool
• (Jörg Klora. Tiago Coutinho )
• GUI FrameWork
• (Josep Ribas)
• Controls System admin and deployment Tools
• (Sergi Puso (SYSTEMS SECTION))
• Power supplies
• (Lothar Krause)
• Liberas
• (Lothar Krause, Jairo Moldes)
• Insertion Devices
• (Fulvio Becheri)
• Radiofrequency
• (Ramon. Suñé)
• Timing
• (Ramon. Suñé)

• Hardware
• Fast interlock units, HV splitter
• (Oscar Matilla)
• Timing
• (Oscar Matilla)
• Patch pannels, Hardware modules, cabling
• (J.Vicente Gigante, Alberto Ruz, Peter Readman)
• EPS (vacuum, magnets, RF)
• (Roberto Ranz)
• Radiofrequency
• (Roberto Ranz )
• Electronics for detectors
• (Andej Seljak, Bern Salo)
• Insertion devices
• (Andej Seljak,)

5
Quite a few Projects on going
Event-based PyStateComposer
Hierarchies of Synoptics
Synoptic Tree 0.4
(from a JDraw file to a DeviceTree)
PySignalSimulator with Dynamic Attributes
Visualization profile for each Synoptic
6
Tango dance Appraisal

• Fits well the needs of Alba.
• Medium size labs...
• Tango is easy
• We have a choice of partners. (hardware)
• Common tools.
• Good for support groups. Dance and the tools the
  same everywhere
• Not real time needed

7
Get started quickly more time to get to the core

• Writing device servers is fast, clear and
  reliable.
• It proved its versatility in different fields of
  control
• Analog I/O cards, mathematic simulators,
  archiving databases using the same easy-to-use
  development tools.
• GUI toolkit allow the developer and the expert
  user to work with the controller as soon as it
  compiles.
• Jive, Atkpannel,
• New PyTango binding speed up incredibly the
  development of middle-layer devices
• State Composers, Serial-based, Tcp-based
  controllers, PLC controller, Simulators.
• Many servers for common equipment available
• The community is active adding dservers.

8
Experience

• Considerable participation in the core.
• PyTango. (python servers )
• 64 bits support.
• Device Pool.
• Easy to outsource parts.
• I.e. defining shared libraries
• Python QT.

9
Device pool
10
Device pool

• class E516Controller(MotorController.MotorControl
  ler)
• class_prop 'SerialCh''Type''PyTango.DevStri
  ng
• AddDevice, DeleteDevice
• PreStartAll, PreStartOne, StartAll, StartOne
• PreReadAll, PreReadOne, ReadAll, ReadOne
• PreStateAll, PreStateOne, StateAll, StateOne
• SetPar, GetPar (velocity, acceleration, )
• GetExtraAttributePar, SetExtraAttriubtePar
• AbortOne, StopOne

11
Sharing (borrowing) ds applications

• ATK
• Archiver
• Egiga

12
Abstraction

• TANGO device model nice abstraction for
  distributed systemadvanced features ?
  userfriendly general purpose GUIs
• predefined types
• parameter types to commands (methods) can not
  be mixed
• hardcoded set of states (ON, OFF, STANDY, etc.)
• (maybe) not everything fits into the
  device-orientated view(e.g. data streams?)
• Installation
• Large scales..

?
?
13
Learning curve

• easy to get started
• expertise required to use advanced featured
  effectively

core Tango developer
Tango internals, Cobra/IDL
polling
events
DS
documentation
clients
jive (configuration)
weeks
one
four
14
Conclusion
Tango mature, although constantly evolving
Thank you!