[d2n-analysis-talk] Compiled Library Mystery

[Diana Parno]

This question is mainly for Brad, since the rest of us are stuck.

When BigBite runs are replayed using the analyzer on the d2n machine, the 
resulting root files are filled with multiple (40+) copies of a number of 
VDC efficiency histograms, regardless of what is in the .odef file. You 
can check out /home/dseymour/onlana/ROOTfiles/e06014_det_BB_1072.root for 
an example.

I grepped the names of the histograms in question and they turn up in only 
one place in my onlana directory: the compiled library libBigBite.so. This 
is a symbolic link to 
/home/flay/d2n_analysis/d2n/bigbitelib/libBigBite.so.

The mystery histograms are defined in BBDecData.cxx. Dave was kind enough 
to comment out a few lines giving the histograms a Write() command, since 
I don't have write permissions in that directory. He then compiled it: "I 
use: ./makeall in the ~/bigbitelib directory."

No change. In trying to figure out why there was no change, I ran

ls -ltr ../../flay/d2n_analysis/d2n/bigbitelib/libBigBite.so

which gives a last-modified date of November 20. Oddly, the symlink from 
Dave's replay directory gives a last-modified date of today, whereas the 
symlink from my directory gives a last-modified date of January 15 (which 
may have been when I last set up that link). The symlinks are both 
pointing to the same place. The code file where the changes were made 
shows the correct last-modified date.

So -- any ideas how a re-compile of the BigBite library might have 
happened without changing the last-modified date of the compiled library?

Any idea where these mystery histograms are coming from, if not from the 
BigBite library? I'd thought I'd made some progress toward understanding 
the dependencies, but I see there may be something left to learn.

Thanks for any help you can offer.

Diana