Linear Collider Forum

Home » Software Tools » Marlin et al » new ilcsoft developers release v01-17-05
new ilcsoft developers release v01-17-05 [message #2302] Fri, 07 February 2014 03:54
Messages: 41
Registered: March 2009
Dear all,

we have prepared another patch release in the v01-17 cycle of developers versions, namely v01-17-05

A lot of fixes and changes have been introduced, please take a look at the already compressed release notes below.

Use the ilcinstall tool with the appropriate configuration files in order to install ilcsoft v01-17-05.

As announced, we have added another OS in the list of reference installations, Ubuntu 12.04 in 64bit. This is experimental for the time being and needs still some work.

Reference installation in afs are available at:

/afs/ # SL5 64bit
/afs/ # SL6
/afs/ 5 # Ubuntu 12.04 64bit

Please report any problems and questions regarding this release in this forum.

Christoph, for the ilcsoft team.

#   iLCSoft release - v01-17-05

Developers release that provides many smaller fixes and a 
number of new features as well as a number of new packages.


 Changes in v01-17-05  w.r.t. iLCSoft v01-17-04

 MarlinTrkProcessor v01-10
  -  added FPCCDTracking code by Tatsuya Mori

MarlinUtil v01-08
     - improved FPCCD classes (T.Mori)
     - added HelixClass_double 

Kaldet v01-13 (by F. Gaede)

 - modified ILDVXDKalDetector:
   - modified to optionally change the relative position of
    the measurement surface within the sensitive layer
     => needed for example in the FPCCD

  - optionally users can add
     <parameter name="relative_position_of_measurement_surface" type="double"value="0.15"/>
   to the VXD parameter section in the GEAR file

 dd4HEP v00-06

2014/02/04  Frank Gaede
  - fixed environment scripts and this${PackageName}.sh
    to also export library pathes for external libraries (CLHEP, LCIO, gear)

  - added some simple tests to ILDExDet, ILDExSimu and noviceN04
    - provide cmake script ( configured as run_test_${PackageName} ) 
      to call test binaries after initializing the environment
      (needed for automated/nightly tests)

  - updated Readme.txt wrt build procedure and running the tests

2014/01/17    Markus Frank
  - Prepare implementation of Alignment using Release and Grap features of TGeoVolume
    and TGeoNode. In principle this was done in a backwards compatible way, but 
    I cannot really check 100% the #ifdef's.
  - Fix handling of XML collections using XercesC.
    (Bug reported by P.Kostka).
  - Fix several compiler warnings issued by Coverity.

2013/12/17    Markus Frank
  - Whole load of fixes to keep coverity quiet....
  - Fix bug in DDG4 field interface component
  - Allow the usage of units, positions and lorentzvectors and in DDG4 properties.
    Example using a position property (ROOT::Math::XYZVector):
    Idle> /ddg4/Gun/position (1*mm,1*cm,0.01*m) 
    Geant4UIMessenger: +++ Gun> Setting property value position = (1*mm,1*cm,0.01*m)  
                       native:( 1 , 10 , 10 ).

  - Support for std::vector<T>/std::list<T>/std::set<T> properties.

  - Support for include directives of the format
        <include ref="file-name"/>
    for the top level <lccdd> tags
       Parent tag    Included root tag  Data content
    -- <lccdd>       <define>           See below.
                     <lccdd>            Nested includes (with DTD validation for Xerces-C)
    -- <define>      <define>           Set of constants
    -- <detectors>   <detector>         Single XML subdetector description
    -- <detectors>   <detectors>        Multiple XML subdetector descriptions
    -- <display>     <display>          Set ov visualization attributes

    DTD validation is only performed using Xerces-C for root tags <lccdd>.

    <lccdd> include directives are executed BEFORE any other directive!
    Other include directives are executed BEFORE the corresponding named directive!

2013/11/14    Markus Frank
  - Reformat files in DDCode and DDG4 to remove the <TAB>s from the source files.
    Hopefully the content is now more readable.
  - DDCore: Remove empty and named constructors for shapes. Empty constructors
    were simply a dangerous hook to memory leaks, named constructors are 
    deprecated since along time.
  - DDG4: add support for messengers to access and modify properties from
    the Geant4 prompt. For the test hit classes and the simple sensitive 
    detector a ROOT I/O plugin was provided.
  - DDG4: Split into several libraries:
     -- pure    link library without any plugins
     --   'Old' plugins. The old stuff at some point
                               will have to disappear.
     --  New framework plugins
     --     LCIO output plugin for new framework

2013/11/03    Markus Frank
  - doc: Add CHEP2013 paper about DD4hep as a start of the documentation 
    section. More to come hopefully.
  - DDCore: several small improvements:
    - Segmentations are no longer Ref_t's.
    - Base internal implementing object directly on the segmentations 
      classes from Christian.
    - Rearrangement of some code from Handle.h to Primitives.h
    - Allow to attach extensions to LCDD
  - DDG4: Fix Frank's simulation problem, which he caused himself
    introducing his famous factories....
  - DDG4: First attempt to fix simulation problem with mignetic field.
    Ready for testing.
  - DDG4: First commit of a independent simulation aimed for LHCb and LHeC
    + other newcomers. Documentation needs to be written.
    - Modular setup of all Geant4 action routines as plugins, including the
      Run-, event-, tracking-, stacking-, stepping- and generator-actions 
      as well as the physics lists with particle and physics constructors.
    - Setup is possible in XML or with C++ either as an AClick in root or
      standalone. Python based starter to come.
    - Missing is a modular description of the G4 UI and VIS manager(s).
    - Documentation is to come. Before documentation is not present
      the usage is discouraged.
    - This version of DDG4 uses boost::spirit. It will compile without
      the BOOST include files and the old functionality is preserved, 
      but it will not work.
      To use BOOST add to the cmake statement the following tags:
      cmake <regular tags> \
      Where BOOST_INCLUDE_DIR should point to the boost header directory.
      BOOST libraries are not used.

 ROOT v05-34-10
- initially forced by dd4HEP requirement
- now uses git

 CEDViewer v01-07-02 (by F. Gaede)
some minor improvements to pyced 
         - added userDraw() method for
           drawing additional objects
         - added example for track cut function
 MarlinFastJet v00-02 (by A. Sailer)
* Add optional feature to store PFOs clustered into physics jets in separate  recParticleOut collection, which can be passed to later processors (S. Redford)
* Add support for FastJet v3
* bug fixes

 MarlinReco v01-08-01 (by G.Grenier)

SDHCAL digitizer : charge simulation : simplifies Polya function. Charge
dispatching on pads : fix integration if using functions (slows the process)
and add new dispatching using sum of Erf functions (recommended option)

        GDML v03-00-02 (by J. McCormick)
forced change, older version disappeared

 Gear v01-04
    - added set methods to SiPlanesLayerLayout (I.Rubinsky)

   - bug fix for  FixedPadAngleDiskLayout::getNearestPad (A.Muennich)
     - Wrap around of angle was not done correctly.

GeneralBrokenLines (GBL) v01-16-03
   - documentation augmented
   - cmake config file exports ROOT dependencies
        - diagonalization of scatters
        - trajectories require at least two points  

 pathfinder v00-06 (by I. Heinze)
  - cleaned up code & directories , fix typos
  - added doxygen comments everywhere
  - added class HoughSpaceAnalyser to save plots of Hough spaces
  - added class HoughMapMaximumCandidateFinder, which does the search of the Hough maps for the maxima, before this was done in every kind of Hough map
  - added a class called HoughMap2DuntIntersectionMethod, which is not used yet, but it is an idea for further development
  - fixed bug in TrackGenerator

  MarlinTPC v00-16
(A. Munnich, P. Hayman, C. Kleinwort, F. Mueller, Li Bo, Ch. Krieger, D. Sankar, Ch. Rosemann) -- lots of bug fixes

(P. Hayman )
- Addition of MM reconstruction (see log for details)

(A. Munnich)
- Added a new flag for the pulses with a plateau. They were previously flagges as anomolous shape. Added a electron particle gun for our LP1 test beam geometry in preparation for detailed simulation.
- First version of analysis processor for dEdx. Needs some more testing and tweaking probably. + root scripts
-  several updates to detailed simulation (check logs for details)
- fixes in ModuleDistortionCalculator.
- Major rewrite of the HitDistributor (detailes see below). It should now work more correctly. Full testing not finished. Charge fluctuation on pads still seem somewhat large, but could also be due to electronics processor.

Modified:  HitDistributor called from ChargeDistributionProcessor
           - Fixed errors, e.g. list of hit pads was wrong
           - Rewrite of code for clearer structure, avoiding the same code repeating in different functions
           - Introduced correct handling and usage of TPCModule
           - Introduced local and global coordinate systems. Now also translated and rotated modules work
- test for the behaviour of the HitDistributor
- test for the new TPCGas class

(Ch. Rosemann)
Changed BField access method from strange to 
 1) initialization of data by ConditionsProcessor
 2) loading of the field with the LoadFieldFromConditionsDataProcessor [takes as option the coil coordinate system]
 3) getting it with 
                gear::Vector3D bfield = marlin::Global::GEAR->getBField().at( gear::Vector3D arbitrary_position);
NOTE: needs a b field map; still to be loaded into the database

C. Kleinwort
- StepWiseHelixGeneralBrokenLineInterfaceProcessor (following changes in Bz) implemented.
Needs track reference point inside magnet. RowTripletBasedTrackFinderProcessor uses as default now 1. hit as reference point. 
- A local helix with bending plane perpendicular to an arbitray magnetic field is now used by StepWiseHelixGeneralBrokenLineInterfaceProcessor to propagate in an changing (only locally constant) field

(Li Bo)
- Added modulesNotUsed in TrackMakingKalmanFilterProcessor 

(Ch. Krieger)
 Changed TimePixMaskCreatorProcessor so it does not need any more an occupancy histogram (which was required but not used before), dead and noisy pixels are now defined via processor params instead of setting them in the processors source code

When you have eliminated the impossible, whatever remains, however improbable, must be the truth. (Sir A.C. Doyle in Sign of Four)
Read Message
Previous Topic:Installation of ilcsoft v01-17-03 - Bug in DD4hep installation?!
Next Topic:CMake with local package
Goto Forum:

Current Time: Mon Jan 22 04:24:00 Pacific Standard Time 2018
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.1.
Copyright ©2001-2010 FUDforum Bulletin Board Software