Deferred assignment is lost in APERTURE calculation

[roman-martin]

//---------------------Lattice Structure----------------------------------------
QUAD: QUADRUPOLE, L = 1.245;
THE_KICKER: TKICKER, L = 0.18, HKICK := k0nl_S62MK1I, APERTYPE = RECTANGLE, APERTURE = {0.055,0.0325};
MYSEQ: SEQUENCE, REFER = CENTER, L =  12.9;
    THE_KICKER,           at = 0.57;
    Q1: QUAD,             at = 9.2 , K1 := k1nl_S00QD1D/1.245;
    Q2: QUAD,             at = 11.8, K1 := k1nl_S00QD2F/1.245;
endsequence;

k1nl_S00QD1D :=  -0.2606126634154173 ; 
k1nl_S00QD1F :=  0.2606789390631277 ;
ON_KICK = 0;
k0nl_S62MK1I := 1.0e-3*ON_KICK;

//------------------------------------------------------------------------------
beam;
use,sequence=MYSEQ;
SAVEBETA, Label=bstart, PLACE=#S, SEQUENCE=MYSEQ;
Twiss, betx=1.0, bety=1.9;



!------play around here by uncommenting aperture command -----------------------
!APERTURE, range=#S/#E;
!-----/play around here---------------------------------------------------------



ON_KICK = 1;
Twiss, betx=bstart->betx, bety=bstart->bety, alfx=bstart->alfx, alfy=bstart->alfy, range=#S/CELL_63;
plot, haxis=s, vaxis1=betx,bety, vaxis2=x, colour=100, table=twiss;

The above lattice has a kicker with an aperture and a deferred expression for the kick.
If run as is, the 2nd twiss will show a kick and a non-zero x in the plot. If the APERTURE command is uncommented, there will be no more kick and a zero x in the plot. I observed that this is only the case if the kicker itself has an aperture.