Re: 'const' again

From: Rene Brun (Rene.Brun@cern.ch)
Date: Thu Sep 03 1998 - 22:43:39 MEST


Rutger van der Eijk wrote:
> 
> Hi,
> 
> Again a message about 'const' etc.
> 
> I somewhere need to stream a 'const Text_t*' into a TMessage, which is a
> TBuffer.
> 
> I have
> 
> const Text_t* name = <something>
> TMessage messOut(kMESS_OBJECT);
> 
> and would like to do:
> 
> messOut << name;
> 
> This doesn't work because the operator is defined as
> 
> TBuffer &TBuffer::operator<<(Char_t *c)
> 
> I think this should be:
> 
> TBuffer &TBuffer::operator<<(const Char_t *c
> 
> as for all the other TBuffer input (i.e. <<) operators. Or is there
> (another) reason why I should not use the << operators? (Should I use the
> WriteArray member?... ifso why are << operators public?)

Good point.
> 
> p.s. I'm curious to things discussed at CHEP98, will there be some
> note/report/summary?

I suppose you are refering to the Root Users meeting that we organized
yesterday night. We intend to post a summary of this meeting in the
coming days.
At the conference, many people/experiments have quoted their use
of Root and also many have announced their move to Root, including major
leading experiments in HEP and Nuclear Physics.

Rene Brun



This archive was generated by hypermail 2b29 : Tue Jan 04 2000 - 00:34:37 MET