== Yellowsheet ==

==== HELAS ====
   * $\epsilon^{ijk} $ color qqq vertex (urgent) 
   * spin 2 (full set of HELAS routines + MadGraph ? )

==== MadGraph ====
   * Decay chains -- Force propagators onshell -- %Y% JA v. 4.2.0
   * Maximum number of particles
   * Graphs with many particles
   * 4-point in phase space
   * Allow two-digit numbers in the process number definitions (@n in the proc_card)

==== MadEvent ====
   * Calculate all couplings event-by-event (at least if one of fixed_couplings is F) -- %Y% JA v. 4.1.34
   * MadEvent for the Grid
   * Phase space integration/grid optimization for SCET
   * Get right cross section and error for each subprocess in the <init> block of the event file

==== Interfaces ====
   * Interfaces : QNumbers (Pythia, done by Michel by fall) -- %Y% MH & JA v. 4.1.43 & py/pgs v. 2.0.19
   * Add HERWIG interface (1st oct)

==== Tools ====
   * Now that the !MadDecay is available it would be nice to have a script/tool to decay events by merging 
     production / decay event files

==== Web ====
   * Tools: Special tool server? Possibility to give URL instead of downloading/uploading files? 
   * Web: Allow Pythia+PGS running after complete run - preferably rewrite web run form as pearl script -- %Y% JA v. 4.1.42
   * Add text field editing of cards -- %Y% MH & JA
   * Plotting : replace ExRootAnalysis by MadAnalysis for online Plotting, add text field to edit plotting cards online instead of full web form -- %Y% RF & JA
   * Wiki : Interesting questions/bug reports by mail should always be answered on FAQ and BugsReport pages, the documents page  
     and News page on clusters should be moved to Wiki

==== Towards the Grid ====
   
   * Pass the scripts to bash  -- %Y% JA v. 4.2.2.
 
   * Allow a set up  that avoids using the central disk mounted on each node. 
     This is easy within CONDOR, but should be done independently of the batching scheme.
     [attachment:mrenna Here] is an e-mail by Steve Mrenna that explains what he did at Fermilab.
       
   * Implement !MadGrid (TS). In this case MG/ME runs two independent phases. %BR%
      *  make_madgrid [0/1] [nameproc] : prepares the !MadGrid code. The script runs a special 
       version of survey and refine to get  to find the cross sections corresponding to each of the integration channels.              
       This can be run either on a single computer or on a cluster and will be available from the web too. 
       The output of this first phase is 
         * The information on the cross section 
         *  A self contained package, !MadGrid_nameproc.tar.gz, which is similar to madevent.tar.gz 
          but it is a frozen code.  (i.e. ALL input cards (except for grid_card.dat)  cannot be edited anymore).  
          Only very little information that can be passed to MadGrid via the grid_card.dat (the number of events
          to be returned, rnd seed, a file name). 
      

==== Models ====
   * Clarify role of Calculators with the new FeynRules approach %Y%

-- Main.JohanAlwall - 07 Feb 2008