20030821: DSSERVE DIRFILE= for POINT datasets (cont.)

>From: "Alliss, Randall J." <address@hidden>
>Organization: TASC
>Keywords: 200308201138.h7KBccLd008934 McIDAS-X MKRAOBID DSSERVE REDIRECT

Randy,

>has umicdas2003 been released yet?

Yes, I sent an announcement to the mcidas-x email list that we maintain
and announced it on our HomePage:

http://my.unidata.ucar.edu

I see that you are not subscribed to the mcidas-x email list.  Since this
is bad in terms of you getting notifications about new releases and bug
fixes, I just added you.  Don't worry, the volume on the list is very
low.

>yes...the ibm is running 2003 SSEC version.

Bingo! ;-)

>i will try and build mkraobid.pgm on IBM.

OK.  Also, it would be a good idea to install Unidata McIDAS-X v2003
on all of the machines you run Unidata McIDAS on.

Tom

>From address@hidden Fri Aug 22 10:00:51 2003

Hi again,
I missed the boat when commenting on using STNLIST TYPE=R and/or
MKRAOBID to get a list of stations to use with UALIST.

It turns out that *you can* use STNLIST to come up with the list of
reporting RAOB stations. You do it by adding the DATASET=, DAY=, and
TIME= keywords in your command. For example,
STNLIST TYPE=R CO=US DATASET=RTPTSRC/UPPERMAND DAY=21/AUG/2003 TIME=18
lists the three U.S. stations that issued 18z RAOBs yesterday. Specify
TIME=12 (instead of TIME=18) and it lists the 91 U.S. stations that
issued 12z RAOBs yesterday.

I'm not sure how I was numbskulled enough to forget about DATASET=, DAY=
and TIME= in STNLIST since I'm one who suggested including them in
STNLIST... :^(

Barry

>From address@hidden Fri Aug 22 10:05:35 2003

I also should have remembered to tell you that using DSSERVE and
DIRFILE= with MD files only works with version 2003 or later (as was
mentioned in the Other Modifications section of
http://www.ssec.wisc.edu/mug/x/03mayx.html).

I believe that if you try it with a version prior to 2003, the DIRFILE=
value is ignored and commands will use MCPATH so commands would access
MD files in $HOME/mcidas/data and other MCPATH dirs (and would therefore
appear to, as you said, "produce files from another dataset").

Barry