Hi Flemming, On Sat, 10 Nov 2001 09:45:55 -0500 "Flemming Videbaek" <videbaek@sgs1.hirg.bnl.gov> wrote concerning "problem in brat code on solaris / rmine001": > While compiling a recent checked-out version of brat > the setup on solaris a few (small problems turned up) > Some of these may have been an operator error. > > a) > ~/newdb/brat/modules/calib/tpc> cvs diff > cvs diff: Diffing . > Index: BrTpcDriftFibCalModule.h > =================================================================== > RCS file: /afs/rhic/brahms/BRAHMS_CVS/brat/modules/calib/tpc/BrTpcDriftFibCalMod > ule.h,v > retrieving revision 1.1 > diff -r1.1 BrTpcDriftFibCalModule.h > 35c35 > < const Int_t kMaxFibres; // max number of fibres (currently 6) > --- > > Int_t kMaxFibres; // max number of fibres (currently 6) > > -- The gcc on sun complained about this const (an error that I know > Steve Sanders also complained about) - Peter could you review what > you want to do here. For some reason, CINT deals poorly with this const integer. I think it may be a problem with an old CINT or something, 'cause using 'GCC 2.96' and ROOT 3.02/01 on Red Hat 7.1 does not give that error. > > b) The root version on solaris is old > -- Kris is it possible to have it updated !! > so I had to comment out the > diff -r1.8 BrRint.cxx > 128c128 > < gROOT->SetMacroPath(0); > --- > > file://gROOT->SetMacroPath(0); -- presumably this goes away. Kris, if you do update ROOT on the Solaris machine, could you _please_ update into /afs/rhic/opt/brahms/pro and /afs/rhic/opt/brahms/new. Cd into the ROOT source dir in brahmlib:~/src/root-3.02 and do ~/bin/confroot make make install for the pro installation, and make clean ~/bin/confroot new make make install for the new installation. You may need to pass additional arguments to the ~/bin/confroot script for the proper X headers and whatever. Note, that there's a similar script for BRAT - yesp, you guessed it, it's ~/bin/confbrat. I really do think, that we should delete all installations that is not in the /afs/rhic/opt/brahms/[pro|new|old] trees, since they cause more confussion than anything else, and as outnlined on numerous occasions, that really posses no problem for user that want to keep thier own installation of BRAT. If I hear nothing, to the contrary soon, I'll start cleaning the house next week. > Had to change the include in BrSplash.h > X11/xpm.h -> /opt/local/include even though I thought to have made the > proper --with-xpm-incdir and --with-xpm-libdir configure flags. Well, the thing is, that the include line is #include <X11/xpm.h> and so if you specified --with-xpm-header-dir=/opt/local/include/X11 the compiler will not find the header, since it will look in /opt/local/include/X11/X11/xpm.h which is wrong. However, if you specify the path as --with-xpm-header-dir=/opt/local/include the configure macro will fail, since it looks in /opt/local/include/xpm.h Clearily, something needs to be done. I'll deal with it ASAP. > c) Manually fixed the path.h file that were created with -e #include > with the -e coming the script that makes the file. I had some problems automatically rebuilding that header. I'll think about it and fix it ASAP. 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 : Sat Nov 10 2001 - 10:51:10 EST