changes in versions 2.2 - 2.4

v. 2.2.0 do not update loop limit icount within a step in subroutines generate and interbunch

v. 2.2.1 initialize qgrid, qgridfft, explicitly to zero, modifed scaling back to boundary in seiler for igeom=4, check icount and increase qtref if too large

v 2.2.2 remove angular dependence from elastic electron reflection, increase qtref if icount becomes too large, correct inconsistency in space charge calculation near edge of chamber, compute energy spectrum separately for electrons incident near the center of the chamber; v2.2.2-c dynamically adjusts the mininum charge of macro-electrons so as to avoid overflow; also the treatment of uniform solenoids has been modified compared with v2.2.2.-b/a; other minor changes affect the electron energy ranges considered in certain output files. v2.2.2-d corrects an error for icoll=2 (missing common block in seiler), and includes the option for additional elastic reflection at low energies modelling the data taken by I. Collins and R. Cimino; selecting inel=2 gives this option which consists of Noel Hilleret's formula evaluated for perpendicular incidence plus an exponentially decaying component centered at zero with decay with 20 eV, such that the total reflectivity approaches 1 in the limit of zero incident energy; inel=1 still selects the elastic reflection according to N. Hilleret (but without any angular dependence). v.2.2.2-e fixed some errors, mainly found by G. Bellodi (corrected treatment of superbunch profile, removed field updates inside on-line Runge-Kutta loops, added missing factor qt in the expression for qtntheta0 to program Ian's and Roberto's parametrization).

v. 2.3 changed exponential decay width of Cimino/Collins paramatrization from 20 eV to 4 eV, speeded up image-charge calculations for elliptical beam pipe (Daniel), corrected error of beam field in the vicinity of beam center for elliptical chamber (Daniel), removed space charge calculation update from on-line Runge-Kutta integration (Giovanni and Giulia), corrected factor in beam weight for falling edge of long uniform bunch (Giovanni and Giulia), increase qtref in seiler routine using qlimit from clean routine (suggestion by Daniel).

v 2.4 corrected error in the vicinity of beam center for round beam in elliptical chamber, added options inel=2 and inel=3, which both model elastic reflection according to the measurement of Cimino/Collins, and option icoll=3, which model proton losses; also added  is a further input line determining a cpu time limit, after which the program execution stops (so as to save files to disk, which are otherwise lost on certain computers).