[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20001231: McIDAS FRNTDISP probs (cont.)

>From: address@hidden
>Organization: SMSU
>Keywords: 200012301857.eBUIvlo29945 McIDAS-X 7.7x FRNTDISP

Bill,

re: inauspicious beginning to an email?
>  HEAVENS NO!!  This is a first for me.  You guys have made the installation
>totally painless.

Phew!

>Whatta ya gonna do for an encore?  Fix Windows?

We may be good, but we are not that good.  Now that you mention Windows
I have to ask you if you/SMSU is more of a Windows shop than a Unix
shop?  The reason I ask is that there is a version of McIDAS that runs
on Windows NT 4.0 (service pack 6), but you have to purchase a package
called Interix (now owned by Microsoft) and an X windows package
(eXceed).  I am not (yet) officially supporting this platform, but I
will if enough sites tell me that it is important for their
operations.

re: McIDAS and Unix scripts using error status information
>I'm guessing the level of errorcode might have some implication for me.
>I'll watch that.

Right.

re: WXCORE.PRD containing appropriate lines for fronts
>Right, I had done that.  They're there.

Good.

>So lets see.  We get the data via ldm, ASUS is shunted off to (above).
>So is some other decoder supposed to put the frontal data into the 
>RTWTXT (or whateve that is...I'm not at work) ADDE data set?  I expect
>that's my problem.

The way that XCD works with the LDM is all textual products are piped by
pqact into 'xcd_run':

DDPLUS|IDS      ".*"    PIPE
        xcd_run DDS

The ASUS1 and FSUS2 bulletins are in this mix, so they are sent to the
daily spool file written to by ingetext.k (run by the script xcd_run).
If the initial setup of XCD was correct, then the textual ADDE server
should then be able to find the bulletins in the spool.  There may be
a problem in the XCD setup that is reflected in you not being able to
display frontal data.  I could jump on your machine and take a look,
but we may be beyond that given the following of this message.

>Art Person from PSU got back to me today...no real suggestions, just a kind
>offer of additional pipe testing to see what happens under large data load
>while both of us watch.  Basically, his institution and mine are both
>almost shut down, while our reception continues to be spotty...anyway,
>thanks for watching, and where this was all headed, months ago, was your
>suggestion below.  Data-on-demand is a much more appropriate distribution
>model for us than what the LDM offers.

Support was CCed on Art's message.  This was a good reminder for me to
send you some alternatives.

re: data on demand with ADDE
>Amen brother! I hear you.

OK, now we're talking.  Another site has already made the switch to data
on demand, and the comment from him was he was very happy especially since
his data no longer had gaps in it (reference meteorograms).

re: give ADDE access to adde.ucar.edu a shot
>Okey dokey.  I will try to ramp-up my knowledge of ADDE this week.

The nice thing about this is that you can ADDEize your scripts one at
a time while still running XCD for local data decoding.  The other
nice thing about this is that there will be a network of well configured
ADDE servers that will have all of the data that you could imagine
on them, and their access will be good as long as your network is up.
One last thing is that ALL of the NOAAPORT NIDS (single station NEXRAD
Level III product) imagery will be available to you through ADDE,
and you don't have to dedicate any local resources to ingesting/archiving
them.

>For the past semester, it has been clear that ldm doesn't work well for
>us; we don't need most of the data distributed, and we can't get much
>of the data we need (largely shortcomings on this end and the state of MO).
>I look forward to letting someone else archive the data, while we use
>the little bit we need.  And it is much easier for a non-meteorology dolt
>like me to learn about new products by utilizing the tools mcidas gives
>me for visualization and analysis.  99.9% of the ldm-users chatter about
>things (including NOAAPORT) goes right by me as so much alphabet soup.

ADDE access to data is exactly what you need.  I feel confident that
transition will be relatively painless for you.

>Good show.  Hope springs eternal!  Thanks again.

You are welcome.

>Happy New Year!

Big party planned for tonight and tomorrow (why do when you can overdo :-),
so it should be!!

Tom