Hi Hiro et al, I've update BRAT. See more below. On Fri, 19 Apr 2002 17:03:58 -0500 (CDT) Hironori Ito <hito@students.phsx.ukans.edu> wrote concerning "minor bug fixes.": > I fixed minor bugs in brat to compile with gcc3. Her is the > files for changes. > > modules/calib/mult/BrSiPedCalModule.cxx ---> got rid of ios::noreplace > modules/calib/mult/BrTilePedCalModule.cxx ---> got rid of ios::noreplace > modules/digitize/BrTileDigModule.cxx ---> casting Int_t and Float_t > modules/rdo/BrMultRdoModule.cxx ----> casting Int_t and Float_t > > ios::noreplace is not in the part of ifstream option. Removing this might > have very serious negative impact but I got rid of it anyway. It is easy > to check the existance of the file before opening it, but I did not bother > to change. (Not right now) The GCC 2.95 had the ios::noreplace in the libstdc++ library, but when the ISO/IEC C++ standard finally came out, it wasn't there, and so starting with GCC 3.0, the GCC steering commitee decided to follow the standard strictly, meaning that very few extension are present in the new libstdc++ library. > It seems that gcc3 is very picky about Int_t, Float_t, Double_t. > There were a lot of warnings. People should be carefull when > calling some methods. The thing is that the ISO/IEC C++ standard is very strict on types and type safety. Java developers should feel right at home, while C or Fortran developers may get a bit frustrated. Even though it means using some casts here and there, I truely think that _all_ calculations should be done in double precision (Double_t), and then cast to storage type as needed. Please, when you make casts, do it the C++ way; that is Int_t(adc[j]) rather than the old C way, which is (Int_t)adc[j] The C++ way is in principal more type safe. BRAT has been tagged: Version: 2.3.9 CVS Tag: BRAT-2-3-9 Changelog: ---------- 2002-04-22 Christian Holm Christensen <cholm@rcas0014.rcf.bnl.gov> * modules/util/BrMainModule.cxx: Now, if fVerbose > 3, write event progress every 100'th event, and if fVerbose > 5 every event. For really slow jobs, it's nice to get a bit more output. * modules/io/BrGeantInput.cxx: Minor fixes * modules/digitize/BrSiDigModule.h, modules/digitize/BrTileDigModule.cxx, modules/digitize/BrTileDigModule.h, modules/digitize/BrSiDigModule.cxx: Some fixes for the modes using the calibrations rahter than parameters. Also changed a lot of if (DebuLevel() > ...) cout << .... to Debug(..., ..., ...) and similar for Info. * help/brat-help.el, help/config.mk, help/configure.in, help/Makefile.am.term, help/README, help/root-help.el: Added the function root-clas to make a general ROOT based class skeleton and some fixes. * configure.in: Bumped revision 2002-04-22 staszel <staszel@rcas0014.rcf.bnl.gov> * db/geometry/BrMagnetVolume.cxx: GetSwimStatus: introduced a copper shielding in the front of D1. * data/geant/BrGeantHit.h, data/geant/BrGeantHit.cxx: Added two methods GetLocalLine and GetGlobalLine which return respectively, local and global line associated with a given hit. * data/geant/BrGeantTrack.cxx: Added method GetTrackLine that returns track line at it's vertex. * data/geant/BrGeantTrack.h: Added method GetTrackLine that returns a track line at vertex. Yours, Christian Holm Christensen ------------------------------------------- Address: Sankt Hansgade 23, 1. th. Phone: (+45) 35 35 96 91 DK-2200 Copenhagen N Cell: (+45) 28 82 16 23 Denmark Office: (+45) 353 25 305 Email: cholm@nbi.dk Web: www.nbi.dk/~cholm
This archive was generated by hypermail 2b30 : Mon Apr 22 2002 - 13:05:29 EDT