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

20000306: McIDAS-XCD setup and LDM quesitons at Colgate (cont .)



>From: Adam Burnett <address@hidden>
>Organization: Colgate
>Keywords: 200002291726.KAA26578 McIDAS-X 7.60

Adam,

>I will try your suggestions.  I did do the ADDE remote server installation
>per the Web documents.

I tried looking at data that should be accessible by ADDE to your machine,
but I could not get back data in the expected format:

telnet gissun.colgate.edu 500
<OK>
<start a McIDAS session>
DATALOC ADD RTIMAGES gissun.colgate.edu
DSINFO IMAGE RTIMAGES
stdin: not in compressed format
    No Datasets found of Type: IMAGE in Group: RTIMAGES
DSINFO -- done

I exited and restarted my McIDAS session requesting uncompressed data
and reran the DSINFO command and it hung.  This tells me that the ADDE
remote server is probably not setup correctly on your machine.

>In fact, I tried to do everything as instructed.

So, your the one ;-)

>I
>always become a bit confused about the path instructions and where
>particular files should be.  

I realize that this is somewhat less than straight forward.  Perhaps the
following will help:

MCDATA - a pointer to the user's McIDAS-X working directory; this is
         the ~mcidas/workdata directory for the user 'mcidas' and
         the ~user/mcidas/data directory for the user 'user'

MCPATH - a list of directories that is searched by McIDAS when looking
         for data, ancillary data, and help files.  The directories are
         listed in colon-separated format and searched in order from
         left to right

We at Unidata feel (strongly I may add) that the best way to access
actual data files (e.g., MDXX, GRID, AREA, etc.) is through file
REDIRECTions.  There are good reasons for this; the best reason is
illustrated by the following scenario:

o assume that you as a user have setup MCPATH as your method of finding
  McIDAS data files.  For sake of argument, let's assume that your
  MCPATH looks like:

  MCPATH=/home/mcidas/workdata:/data3/mcidas:/home/mcidas/data:/home/mcidas/help

o further assume that there is an AREA file named AREA0001 in the
  /home/mcidas/workdata and /data2/mcidas directories

o next assume that you have read/write permission for the copy of AREA0001
  in the /home/mcidas/workdata directory, but only read premission for
  the copy in /data2/mcidas

o next, assume that you have been using AREA0001 as the container for
  experimentation; for instance, you have been trying to come up with
  a fog product using GOES IR channels 2 and 4.  In the process of
  experimentation, you delete (either accidentally or on purpose)
  AREA0001 (though an IMGDEL command for instance).

The next time you try to write into AREA0001, the programs will run
_with no errors indicated_, but the expected output will not be seen
when you display AREA0001.  For even the experienced McIDAS user this
situation is extremely frustrating to say the least.  The situation gets
even more frustrating/hard to diagnose when the file being looked at
in more than one place is something like LWPATH.NAM, the repository
for REDIRECTions for a user's session.  This file should only be
found in the user's McIDAS working directory (i.e., it should not
be shared ever).

>As for the LDM problem. LDM always worked before.

I just started talking to my system administrator about this.  He is
looking for possible problems/needed patches.

>I'm concerned that I
>screwed something up as I worked through my installation.

Even if you did, it will not be hard to get back to an operational state.

>I did have to
>obtain and install the Solaris 5.0 compilers (as instructed) before I did
>all this.

Did you heed the warnings in Notes and Warnings:
http://www.unidata.ucar.edu/packages/mcidas/mcx/warning_mcx.html

This document warns users of SC5.0 that there is a bug in the Fortran
compiler that necessitates compiling with optimization turned off.

>Maybe I corrupted something?

There is something a little amiss, but it is probably not a biggie.

>Thanks

Later...

Tom