Date:         Mon, 10 Mar 2008 15:04:00 -0700
Reply-To:     "Pardee, Roy" <pardee.r@GHC.ORG>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Pardee, Roy" <pardee.r@GHC.ORG>
Subject:      Re: Strange SQL behavior on PC sas--bug?
Comments: To: "Duell, Bob" <BD9439@att.com>
Comments: cc: "Rubanowice, David" <rubanowice.d@ghc.org>
In-Reply-To:  <331C5406E0EC64408B07D6B74516BE17045C4AAB@BD01MSXMB018.US.Cingular.Net>
Content-Type: text/plain; charset="us-ascii"

Ah--good clue maybe--it looks like our unix machine has an extra hotfix
applied.  This guy here:

ftp://ftp.sas.com/techsup/download/hotfix/ve9bx.html

Nothing on the fix list jumps right out at me (tho that's only after a
quick skim) but maybe it's a side-effect?

Thanks!

-Roy

-----Original Message-----
From: Duell, Bob [mailto:BD9439@att.com]
Sent: Monday, March 10, 2008 2:25 PM
To: Pardee, Roy; SAS-L@LISTSERV.UGA.EDU
Subject: RE: Strange SQL behavior on PC sas--bug?

Hi,

I ran your code on my Solaris UNIX machine and my PC (both running
9.1.3, SP4) and got exactly the same result:  the same as you got on
your PC.

Could it be that you are on different maintenance levels?

For what it's worth, I think it's working correctly (displaying missing
values).  Of course, I would have thought the absence of a GROUP BY
clause would have caused an error (as it would with some databases), but
clearly that's okay.

Very curious!

Bob

-----Original Message-----
From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of
Pardee, Roy
Sent: Monday, March 10, 2008 11:55 AM
To: SAS-L@LISTSERV.UGA.EDU
Subject: Strange SQL behavior on PC sas--bug?

Hey All,

A collegue posted this to an internal listserv here & I suggested
tapping you kind folks about it.  Consider this code:

* ---------------------------- ;

data junk;
    x=5;
run;

data no_obs;
    set junk (where=(x=1));
run;

proc sql _method ;
  select  "I am a text literal" as lit,
          "&sysdate"d as datey format=mmddyy10.,
          count(*) as NumObs
  from no_obs
  ;

quit;

* ---------------------------- ;

When we run that on our unix machine, the literal & macro var get
processed & we get (as we expect):

lit                       datey    NumObs
-----------------------------------------
I am a text literal  03/10/2008         0


but when we run on our local PCs we get missings for the literal and the
macro var expression:

lit                       datey    NumObs
-----------------------------------------
                              .         0

When you look at the _METHOD output, there's an extra sqxfil step on the
PC output.

Is this just a matter of having different options specified on the 2
platforms?  Can anybody reproduce our PC results?

Thanks!

-Roy

Roy Pardee
Research Analyst/Programmer
Group Health Center For Health Studies (Cancer Research Network)
(206) 287-2078
Google Talk: rpardee