Hi Ruben, This problem is consistent with the underlying problem which lead to your 'old problem of wrong offsets'. Both issues should be fixed in the ROOT CVS repository as of 3/27/2004 or later. Cheers, Philippe. -----Original Message----- From: Ruben Shahoyan [mailto:Ruben.Shahoyan@cern.ch] Sent: Tuesday, March 30, 2004 7:33 AM To: Philippe Canal Cc: roottalk@cern.ch Subject: RE: [ROOT] Problem with Automatic schema evolution? Hi Philippe, it seems there is another problem: when I first generate the file with old branches and NaDigit1D class version 1, then convert it to file with new banches (1 TClonesArray/branch) of digits v_2, and try to read it it in a new root session, apart from old problem of wrong offsets, I get a memory leek due to the lot of TStreamerInfo's generated: For 1000 events with kNSensors=70 I get after the reading the following lines in the gObjectTable->Print() ... TStreamerInfo 70030 70030 84 5882520 5882520 TClass 86 86 140 12040 12040 TObjArray 70197 70197 40 2807880 2807880 TStreamerBase 70028 70028 104 7282912 7282912 TStreamerBasicType 210059 210059 100 21005900 21005900 TStreamerString 8 8 96 768 768 TStreamerBasicPointer 4 4 120 480 480 TStreamerObject 5 5 96 480 80 TStreamerObjectPointer 4 4 96 384 384 TStreamerObjectAny 2 2 96 192 192 TStreamerObjectAnyPointer 70002 70002 96 6720192 6720192 Such thing does not happen if I generate -> rewrite -> read from the very beginning the digits of v_2. .... TStreamerBase 27 27 104 2808 2808 TStreamerBasicType 56 56 100 5600 5600 TStreamerString 8 8 96 768 768 TStreamerObjectAnyPointer 1 1 96 96 96 TStreamerBasicPointer 4 4 120 480 480 TStreamerObject 5 5 96 480 480 TStreamerObjectPointer 4 4 96 384 384 TStreamerObjectAny 2 2 96 192 192 TClonesArray 70 70 48 3360 3360 Cheers, Ruben
This archive was generated by hypermail 2b29 : Sun Jan 02 2005 - 05:50:07 MET