Re: [ROOT] quitting ROOT by killing the job?

From: Rene Brun (Rene.Brun@cern.ch)
Date: Thu Oct 31 2002 - 07:31:03 MET


Hi Maarten,

Thanks for reporting this huge error in my last mail.
As you say, one should read:

We recommend to use named macros to eliminate these types of problems.
With a named script, you should NOT  use gROOT->Reset();

Rene Brun

On 30 Oct 2002, Maarten Ballintijn wrote:

> Hi Rene,
> 
> Didn't you miss a 'not' there?
> 
> Cheers,
> 
> Maarten.
> 
> On Wed, 2002-10-30 at 15:03, Rene Brun wrote:
> > Hi,
> > 
> > If you are a beginner with ROOT (and may be coming from PAW), it it likely
> > that you try to reexecute the same "unnamed script" without resetting
> > the global scope environment.
> > If you create variables or objects in an unnamed script, these variables
> > are created in the global C++ scope. You should use the statement
> >    gROOT->Reset();
> > as the first statement of your unnamed macro.
> > 
> > We recommend to use named macros to eliminate these types of problems.
> > With a named script, you should use gROOT->Reset();
> 				^^^
> 				not?
> 
> > See the Users Guide.
> > 
> > Please send more information about what you are doing with an example
> > of the script that you rexecute if my diagnostic above is not correct
> > 
> > Rene Brun
> > 
> > On Wed, 30 Oct 2002, Michele Zaffalon wrote:
> > 
> > > Dear rooters,
> > > 
> > > I am new to ROOT and I have installed it on a Linux Debian with gcc 3.2.
> > > Now as a beginnner I do nothing spectacular and I do get some segfaults,
> > > some functions not defined in the scope and so on.
> > > Sometimes for no apparent reason a script that has worked till a second
> > > before gives some critical error, and the only way to get rid of it is
> > > to quit root and restart again. Still sometimes the only way to quit is
> > > kill.
> > > Is it normal that I can hardly reach the 100th command or am I doing
> > > something terribly wrong? I guess the order of magnitude of the number
> > > of commands before the whole thing crashes would be a good indication...
> > > 
> > > Best regards
> > > Michele
> > > 
> -- 
> Maarten Ballintijn <maartenb@mit.edu>
> Massachusetts Institute of Technology
> 



This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:51:16 MET