Another HepTuple bug fix

Thus far two bugs have been found in the HepTuple code. Both have to do with confusing array sizes. One fix has already been put into D0's version of HepTuple. The other is on it's way. In the meantime, you can make the corrections yourself in your own release area. Here's how:

1. In your release directory, create a directory called "HepTuple" and enter it :

        > mkdir HepTuple
        > cd HepTuple
    

2. Copy a corrected version of the HepTuple package into this directory. If you are on ClueD0, use :

        > cp -r /work/gallifrey-clued0/bellavan/build_t03.06.00/HepTuple/* .
    

   If you are on d0mino, use :

        > cp -r /www-d0/WWW/computing/trigsim/release/HepTuple/* .
    

3. Go into your release area's "include" directory and create a soft link so that your version of HepTuple is picked up:

        > cd ../include
        > ln -s ../HepTuple/HepTuple HepTuple
    

4. Go back to the base directory and build everything:

        > gmake all
    

5. You should be set to run TrigSim, or whatever, with your own version. Don't forget to use the command line option "-localbuild"!

For those of you curious as to what the code change was, it is in the file HepTuple/src/HepRootNtuple.cpp, on line #668. The block needs to be specified when the index number is looked up. So the line:

    Column * ci = findColumn(indexName);
 

    Column * ci = bk->findColumn(indexName);