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

19990915: McIDAS,WSI access via IDD (cont.)

>From: weather <address@hidden>
>Organization: NMSU/NSBF
>Keywords: 199909151520.JAA06097 WSI NIDS

Robert,

>I somehow overlooked the pricing for the NIDS, I got the
>update and made the changes and am looking at the data now.

OK, please let me know if you find anything amiss.

>The connection to UCAR is pretty bad, took forever to get 
>the updates, but it is improving and I am looking at a radar
>image from LWX now.  I like what I see so far. Maybe
>the connection will improve and I can see more later.

Since you probably don't know how the NIDS ADDE server works, I
can offer the following (some of this you obviously know, but I
include it here for others that might read this message):

o to find out what descriptors are available for the RTNPNIDS
  dataset, run:

  DSINFO IMAGE RTNPNIDS

o to get a listing of the available stations for a particular product,
  run:

  IMGLIST RTNPNIDS/N0R ID=LIST

  Note that the times listed in this list are NOT the latest available,
  but, rather, are usually the oldest available.  The reason for this
  has to do with my implementation -- I was going for speed.

o to display the latest image for a particular station for a particular
  product overlaying county outline and state maps on it, run:

  IMGDISP RTNPNIDS/N0R ID=LWX STA=LWX EU=BREF SF=YES REFRESH='EG;MAP SAT 5 
COUNTY=ALL;MAP H 1 WID=2;BAR SU=BREF'

o to put up a loop of a particular product from a particular station, run:

 IMGDISP RTNPNIDS/N0R ID=LWX STA=LWX ALL=1 10 EU=BREF SF=YES REFRESH='EG 
(GRA);MAP SAT 5 COUNTY=ALL GRA=(GRA);MAP H 1 WID=2 GRA=(GRA);BAR (GRA) SU=BREF'
 LB 1 10;DR 9*1 15

>So the data I am looking at is unencrypted NOAAPORT data?

Yes, that is why I can let you look at it.  Please don't post it on
the net, however.

>Do you think our T1 would be able to handle the increased
>data(Tier II) if we buy it from WSI

Yes.  Tier II (one floater site) would be easily handled.  The NIDS files
are pretty small.

>or recieve the free NOAAPORT feed when/if it starts?

The eventual NOAAPORT feed for NIDS will be for ALL of the NIDS sites.
This is 161 NEXRADS that will be sending 22 products (20 plan view
views that can be displayed by McIDAS and GEMPAK/GARP) every 5 (storm
mode), 6 (precipitation mode), or 10 (clear air mode) minutes.  At an
approximate size of 12 to 20 KB, you can see that it would swamp a T1.
There is so much data that the NWS is thinking about a separate
NOAAPORT channel just to handle it.

>Also do you think we should dump the Sun compilers in favor
>of gcc (I assume you will support gcc on SPARC in the next 
>release..  if we go that route) if we get a new machine (I putting
>in for one in the new engineering plan)?

SSEC has gone in that direction.  They now only do testing of their
code compiled with gcc/f2c.  I build using both vendor supplied and
gcc/f2c compilers on only one system right now, Solaris x86.  I will be
building more and more using the gcc/f2c combination as time goes on,
but I will attempt to keep checking builds with vendor compilers.  I
can forsee that I will be forced to support only one compilation
environment on the least used OSes (e.g. HP-UX, Digital Unix, and
AIX).

>That would of course knock $400
>off the cost, but I don't won't to do it if the gcc compilers
>do indeed produce slower executables and/or are buggy themselves.

I have not really noticed a speed difference on our Solaris x86 boxes.
I think that one would have to be doing huge data crunching jobs before
any difference would surface, but that is just an uninformed opinion
right now.

>Thanks,

Later...

Tom

>From address@hidden  Wed Sep 15 19:34:06 1999

Everything works great on the ADDE server. I am sitting here looking at
tstms out my window that I can also see on the AMA radar from the
server with much less effort than the dial-up account.  If WSI would go
for it, I think that would be great for us.  This summer in Ottumwa I
would have to manually save Intellicast.com loops for the scientist
every few hours then display the animated gifs and save the individual
frames..what a pain.  This was for his post-flight analysis.  I think
we pay about $250 a month during heavy use months(6 months), but only
about $75 a month the rest of the year so they would actually make a
little more money from us.


I certainly will not display these on the web.  

As always I appreciate the help.

Robert
NSBF Meteorology