Re: Pb. of backward compatibility ?

From: Rene Brun (Rene.Brun@cern.ch)
Date: Thu Mar 09 2000 - 19:42:45 MET


Hi Laurent,
You have already received an answer on your mail regarding
forward compatibility.
I just want to add that at the last Root workshop, we presented
a scheme to support automatic schema evolution. We have already
added to the system a new facility called StreamerInfo (not yet used).
The idea is that once we have the full dictionary available
in the Root files and Trees, one should be able to support
complete forward compatibility, as it is already implemented
in teh Root Tress for split mode.
We will claim that this is true only once we will have effectively
implemented the scheme ::)

Rene Brun


On Thu, 9 Mar 2000, Laurent APHECETCHE wrote:

> Hi,
> 
> I'm producing a tiny ntuple with Root V2.23/12 (Linux)
> 
> {
>   TFile* f = new TFile("ntuple.root","RECREATE") ;
>   TNtuple* ntuple = new TNtuple("ntuple","ntuple","r:z:phi") ;
>   Float_t r,z,phi ;
> 
>   r=z=phi=10.0 ;
>   ntuple->Fill(r,z,phi) ;
>   f->Write() ;
>   f->Close() ;
> }
> 
> which I can read back (of course) from 2.23/12, but which I can not read
> back from a previous version
>  of Root (/11,/10,...) : I get a seg. violation :
> 
> root [1] TFile f("ntuple.root")
> root [2] f.ls()
> TFile**         ntuple.root     
>  TFile*         ntuple.root     
>   KEY: TNtuple  ntuple;1        ntuple
> root [3] ntuple.Print()
> 
>  *** Break *** segmentation violation
> 
> Is it that 2.23/12 is writing trees that are not readable by previous
> Root versions (is it advertised somewhere) ? Or am I doing something
> obviously wrong ?
> 
> Thanks for your help.
> 
> -- 
> APHECETCHE Laurent (mailto:aphecetc@in2p3.fr) | TAPS/PHENIX/ALICE
> SUBATECH-Ecole des Mines de Nantes-4 rue Alfred Kastler-44070 NANTES
> cedex 03
> TEL (+33/0) 2 51 85 84 17 - FAX (+33/0) 2 51 85 84 24
> http://www-subatech.in2p3.fr/~photons -
> http://www.phenix.bnl.gov/~aphecetc
> 



This archive was generated by hypermail 2b29 : Tue Jan 02 2001 - 11:50:21 MET