[Rivet] Final particles with negative squared mass

[Andy Buckley]

Paolo Francavilla wrote:
> Thanks a lot,
> 
> ----------Rivet-Bootstrap----------
> I tried to modify a bit the bootstrap script to be able to install rivet 
> on my machine.
> It seems to work for me, but probably it is not general enough.
> The two problems that I have seen are related to the GSL Library 
> (especially the headers)

There is indeed a bug in the released version of Rivet when it comes to 
finding non-standard GSL installations: since most systems have the GSL 
libs and headers in system paths, this wasn't a problem until I started 
testing the 1.1.3 install on "Scientific" Linux 5 (which astonishingly 
has no system copy of GSL!). I've patched the rivet-1.1.3 SVN tag to fix 
this issue, but there's no new tarball (due to sensitivity to HepMC 
versions which makes building the tarball a bit non-trivial). You could 
check out the 1.1.3 tag from SVN and try building that: the next release 
will also include a new --with-gsl= flag on the bootstrap script, but it 
looks to me like you will have to do some installation by hand for now 
(or use it on a standard system like lxplus, if you need results right now!)

> and a wrong location of fastjet
> (at a certain point it starts to use /afs/.cern.ch instead of 
> /afs/cern.ch, but this seems to be fault of a wrong fastjet-config in afs).

That sounds like an artefact of the Genser install process --- they 
initially install into the /afs/.cern.ch tree, then copy across, with 
resulting problems like this. I suggest you report it to them.

> -------------Alpgen----------------
> I have a question about Alpeng+Herwig (or Pythia) via  Agile. I tried to 
> use it, generating the file *.unw with Alpeng, and trying to use it
> in Agile, but it seems to be stuck.
> Do you have a step-by-step example on how to use it?

As far as I am aware, AGILe's interface to AlpGen+Pythia works, but the 
AlpGen+Herwig one doesn't, when using the Genser AlpGen libraries. From 
some debugging that I did with Simon Dean at Les Houches, I believe this 
is the foult of the Genser build again: the hang occurs in a call to 
that library (UPVETO, iirc) rather than any part of the AGILe code, and 
we had it working in the past with our own builds of the AlpGen 
libraries when Giulio Lenzi was doing some multi-jet studies. Again, I 
suggest pursuing this with Genser, since I haven't yet mentioned it.

A word of warning: AGILe is not by any means supported as actively as 
Rivet is... it's a tool to make running certain Fortran generators 
easier, but the support is not complete, nor is the set of generators, 
and we work on it by necessity. So if you're actively trying to use 
AGILe with AlpGen, then you may well have to do some debugging and 
talking to Genser yourself. But we will of course be very happy to 
accept patches!

Best wishes,
Andy

PS. Please send support emails to the rivet at projects.hepforge.org list 
rather than to me: you'll get a much faster reply, and there's a good 
chance it'll be more informative, too ;)