[daip] Re: corrupted RUN files in 31DEC03?

[Eric Greisen]

Stephen White writes:
 > 
 > Hi Eric,
 > 
 > I am doing my annual installation of the latest frozen version of AIPS,
 > 31DEC03, and everything went fine (on both Solaris and Mandrake) 
 > until I tried to run the Y2K test.
 > On investigation, I found some strange characters in the Y2K RUN file 
 > that prevented it from compiling: it looks as though 
 > some "=" signs got replaced by 
 > "~" symbols. If I do a grep on "\~" in the 31DEC03 RUN directory I find:
 > 
 > BASFIT.001:   plcolors(1,9)~ 1 1 1 0.07 0.08 0.03
 > CYGLOAD.001:      IMAGRP=0; IMAGRP(11)~0.52,0,0.1,0.3,0.1,80
 > CYGLOAD.001:   DIFNAM(7)~'IMCL1','IMCLN';   OPTYPE=TNAME
 > NEWPARMS.001:ELSE;FOR GRCHAN=4:1 BY -1;GREAD;PLCOLOR(1,GRCHAN)~RGBCOL; END
 > NEWPARMS.001:END; PLCOLOR(1,9)~RGBCOL; GRCHAN=I; RETURN
 > VLARUN.001:  lsampcal(5)~ '3C138','0518+165','0521+166','J0521+1638'
 > VLARUN.001:  lsampcal(9)~ '3C147','0538+498','0542+498','J0542+4951'
 > VLARUN.001:  lsampcal(13)~'3C286','1328+307','1331+305','J1331+3030'
 > VLARUN.001:  lsampcal(17)~'3C295','1409+524','1411+522','J1411+5212'
 > VLBAPIPE.001:$ Scales the flux/noise parameters (~V2*SEFD) for P_EXT = BaseLine or IMage
 > VLBAPIPE.001:  search(6)~antnum('NL'),antnum('BR'),antnum('NH'),antnum('SC'),antnum('MK')
 > Y2KLOAD.001:      CLBOX(1,2)~2230,2203,3282,3202, 1840,1850,2331,2298
 > Y2KLOAD.001:         IMAGRP=0; IMAGRP(11)~0.52,0,0.1,0.3,0.1,80; FACTOR=0
 > Y2KLOAD.001:         IMAGRP=0; IMAGRP(11)~0.52,0,0.1,0.3,0.1,80
 > Y2KLOAD.001:      DIFNAM(3)~'IMRES1','IMRES2','IMRES3'
 > Y2KLOAD.001:      DIFNAM(3)~'APCLN','APRES'
 > 
 > They all look as though they should be "=".

    ~ is a new verb in aips that allows exactly what is being done
above - filling in several values in an array starting somewhere
inside the array.  All uses above are correct.  Did your installation
RUN POPSGN?

Eric Greisen

 > None of these appear in the 31DEC02/RUN files. I checked the tar file 
 > for 31DEC03 today and they are still present. Thought I should let you 
 > know,
 > 
 > regards,
 > 
 > Stephen WHite 
 > 
 > 
 >