19990924: publishing an ADDE dataset

>From: "Jennie L. Moody" <address@hidden>
>Organization: UVa
>Keywords: 199909241521.JAA10457 McIDAS-X ADDE

Jennie,

>I got a request from one of the people at the recent
>NESDIS meeting in DC.  He is one of the weather service
>contacts, and he really liked the specific
>humidity imagery.  He had spoken at the meeting about
>the fact that soon they hope to be distributing a
>lot of Derived Product Imagery (the kind of stuff at
>CIMSS and NESDIS FPDT) on NOAAPort.  After I presented
>he (Kevin Schrab) and Don Gray expressed a lot of interest. 
>Kevin told me they would be really interested 
>in including this as an additional derived product 
>(I felt a bit tentative, since I feel we are not as "mature" 
>as most of the DPI'bs), but I thought it was a nice idea.  In 
>the meantime, he asked for the web-link to just view the image 
>loops we are making.  Well, after one day of looking at these
>(I sent him the link yesterday), he has asked for 
>ADDE access to the images.  

Great, congradulations.  This should be really easy.

>Thats where I need help.  I am still painfully blind about
>how the ADDE was/is set up on windfall, since Kevin did
>this, and I have not had to really deal with it.  Somehow,
>I do not think we were set up to be a remote server.

You are setup.  I routinely bounce to windfall (and most of the twenty
other sites running the ADDE remote server) to check on IDD data flow.
I especially do this when someone reports problems receiving data.

>But then again, I guess when I was in Boulder one time
>you linked to my system and we did view images from windfall?  

Right.

>I don't know how to give ADDE access to a remote user (she
>said feeling profoundly dumb). (I also don't see any drawback to 
>doing this, but would listen if you had any advice to
>the contrary?).  

ADDE is pretty much wide open.  If someone knows your machine name and
the names of data sets that you have, then s/he can run McIDAS remotely
and look at your holdings.  I do this and you never know.

>I understand from your message last night that you too are 
>working with a full plate right now, so a simple response of 
>"can't help ya now" is fine if thats the case.

Putting together an ADDE dataset is pretty easy, and, the good news, is
that it will get even easier in the future.  Here is what I would do:

o collect all of the imagery that you want to make available via
  ADDE into a single directory (this is not really needed, but it
  is logical)

o have the image files named so that their AREA numbers are contiguous
  or at least lie in a unique range.  For instance, set the AREA numbers
  to be in the 4000s, or something equivalent.  The idea here is that
  these images must be so that their AREA numbers do not overlap with
  any others
  
o crank up a McIDAS-X session as 'mcidas'

o verify that you can find the images in your dataset by either file
  REDIRECTions or through your MCPATH setting

o create a dataset using the DSSERVE command.  For the sake of argument,
  let's assume that the images lie in the cylinder range 4100 - 4200,
  inclusive.  Let's further assume that you want to name this dataset
  UVA (the group name) and HUMID (the descriptor) (change these to
  something you like; these names mean nothing by themselves).

  Create the data set:

  DSSERVE ADD UVA/HUMID IMAGE 4100 4200 "UVa Specific Humidity Composites

That is all that there is to it.  You should, of course, verify that
you can access this dataset before sending the information along:

  DSINFO IMAGE UVA
  IMGLIST UVA/HUMID.ALL
  IMGDISP UVA/HUMID 1 EU=...

>Thanks Tom.

No problem.  Please let _me_ know the data set so I can look also :-).
You will also probably want to make your enhancement(s) for these
images available so that they will look the same to folks in NESDIS as
they do to you.

Tom