[Rivet] rivet and GENSER

[James Monk]

Hi Anton,

The version of Rivet publicly available in Genser is the previous release, 1.1.3.  I installed it there largely so that the ATLAS interface to Rivet could use it, but it is also available to other experiments.  Please *do not* re-compile, move or otherwise alter that production installation.

Python was deliberately disabled because 1) the included swig python modules were built with the wrong HepMC and/or python version (can't remember which) and 2) the swig installation on SLC is (or at least was) not able to re-build the swig modules.  Since the ATLAS interface only needs the C++ libraries, the lack of python modules did not matter.

As Hendrik suggests, if we wait a short while then version 1.2 is out.  I have also been trying (on-and-off, since it is a frustrating experience!) to install a release candidate in Genser, which is in /afs/.cern.ch/sw/lcg/external/MCGenerators/rivet/1.2.0rc1/ .  There is a more up-to-date tarball candidate on the Rivet downloads page:

	http://www.hepforge.org/downloads/rivet

cheers,

James

On 11 Feb 2010, at 14:58, Anton Karneyeu wrote:

> Hello friends,
> I am going to integrate Rivet to GENSER tests system, and I found two 
> problems:
> 
> * current installation of Rivet and AGILe at
> 
>     /afs/cern.ch/sw/lcg/external/MCGenerators/
> 
>   doesn't have necessary python scripts to run agile and rivet.
>   This is due to configure option --disable-pyext , which was used 
> during installation. That is why I want to reinstall this tools. Can you 
> comment - will it be save for current user of rivet/agile?
> 
> 
> * second problem related with old version of python 2.3/2.4 on 
> slc4/slc5, which does not allow to direct use of several rivet python 
> scripts (make-plots, aida2root, aida2flat) due to use of module 
> xml.etree.ElementTree (which is available starting from python 2.5 only).
>   For this problem I see two solutions:
>     - to patch scripts to use xml.dom module.
>     - to install each time ElementTree module to some local directory 
> (small patching of scripts will be necessary in this case too).
> 
> May be you can propose another solutions?
> 
> Best regards,
> Anton
> _______________________________________________
> Rivet mailing list
> Rivet at projects.hepforge.org
> http://www.hepforge.org/lists/listinfo/rivet