Re: JException

[David Lawrence <davidl@jlab.org>]

Hi Mark,

    I think we should maybe go with Elliott's version. They both do 
something very similar except Elliott makes use of the 
"backtrace_symbols" function to grab the function names. Craig accessed 
the info by running the addr2line program cleverly through popen. I'm 
guess Craig may have chosen that option because the  backtrace_symbols 
routine only works on ELF systems (e.g. NOT Mac OS X). Anyway, Elliott 
looks to have solved the name demangling problem as well which is a very 
nice feature.

    I would suggest though replacing all of the "SunOS" references with 
ones that use a test for the GNU compiler and print appropriate messages 
if it is not what is being used.

Regards,
-David

Mark M. Ito wrote:
> I forgot to add: Craig's header file can be found at 
> http://clasweb.jlab.org/websvn/prod/filedetails.php?repname=12+GeV+Physics&path=%2FJANA%2Fsrc%2FJANA%2FJException.h 
>
>
> Mark M. Ito wrote:
>> Folks,
>>
>> I was about to check in Elliott's exception class (see url below) 
>> into the JANA tree, when I noticed that there is already a JException 
>> defined there, written by Craig Bookwalter. Can we use it? Or should 
>> we replace it with Elliott's version? I really feel like throwing 
>> something! ;-)
>>
>> Elliott mentioned his class last month: 
>> http://www.jlab.org/ccc/mail_archives/HALLD/halld-offline/msg00563.html 
>> .
>>
>>  -- Mark
>>

-- 

------------------------------------------------------------------------
  David Lawrence Ph.D.
  Staff Scientist                 Office: (757)269-5567   [[[  [   [ [       
  Jefferson Lab                   Pager:  (757)584-5567   [  [ [ [ [ [   
  http://www.jlab.org/~davidl     davidl@jlab.org         [[[  [[ [[ [[[
------------------------------------------------------------------------