Okay, I'm confused by what this log on my ADMIN configured app server is telling me.  The CONFIG setting shows the processing order as follows and that final "sasv9_usermods.cfg" file is where I have my     -SET  encoding "UTF-8"    statement defined.   Given that SAS should take the final encountered system option as it goes through each supplied file, I would think this is the proper place to make that desired definition to further my tests on reading the special symbol.   proc options option=(config encoding ) value ;     Option Value Information For SAS Option CONFIG Value: ( /app/sas/install/SASFoundation/9.4/sasv9.cfg /app/sas/install/SASFoundation/9.4/nls/en/sasv9.cfg /app/sas/install/SASFoundation/9.4/sasv9_local.cfg /app/shared/sasconfig/gridconfig/Lev1/ADMIN/sasv9.cfg /app/shared/sasconfig/gridconfig/Lev1/ADMIN/sasv9_usermods.cfg /app/shared/sasconfig/gridconfig/Lev1/ADMIN/WorkspaceServer/sasv9.cfg /app/shared/sasconfig/gridconfig/Lev1/ADMIN/WorkspaceServer/sasv9_usermods.cfg ) Scope: SAS Session How option value set: Internal     Option Value Information For SAS Option ENCODING Value: LATIN1 Scope: SAS Session How option value set: Config File Config file name: /app/sas/install/SASFoundation/9.4/nls/en/sasv9.cfg     Yet, the same app server session tells me the ENCODING option is being set from the first .cfg file in the above listed order.   I've also tried adding the option to the launch command inside the app server definition in the metadata repository, to no positive affect.   Not sure what I'm overlooking here, open to suggestions.     ... View more

Thanks for responses to date.  I was curious about the Flitcon tools but alas, the VA firewall had that identified as a black-listed site so there is little hope of getting anyone from our Technical Reference Modeling office to perform a software review on that other toolkit.   I did browse thru that site on my home computer but didn't see anything leading me to believe it approached the capabilities offered by the current Metacoda tools.  I shall take these findings back to our Acquistions office and hope for the best. ... View more

Perhaps you would be better off learning how to combine A and B, rather than fixing the problem later.  If you are using a DATA step, it would be: data C;    if _n_=1 then set A;    set B; run; If you are using SQL: proc sql noprint;    create table C as select * from A, B; quit; You would no longer have that single observation at the end, where the contents of A matches nothing from B. Good luck. ... View more

Or just this way: DATA NEW2(DROP=RC ) ; if _n_=1 then do;    DECLARE  HASH  AA(DATASET:' work.new(where=(score=2) ) ' ) ;  /* only the offering IDs are ready into hash object */    RC=AA.DEFINEKEY( 'ID' ) ;     RC=AA.DEFINEDONE() ; end; set new; if aa.check(); RUN ; ... View more