[daip] [Non-DoD Source] Re: AIPS ATLOD.f Compilation failure on Ubuntu

[Sargent, Andrew J CIV USN NAVOBSY DC (USA)]

It seems this was just my misunderstanding of how these files are supposed to work - after getting input from Megan we were able to successfully configure the installation!

Thanks again,
Andy

On 10/28/20, 2:03 PM, "egreisen" <egreisen at nrao.edu> wrote:

    On 2020-10-28 13:30, Sargent, Andrew J CIV USN NAVOBSY DC (USA) wrote:
    > I spoke a little too soon - I tried it with the plus sign and two
    > blanks and that seems to have let the compilation finish (the minus
    > sign and two blanks did not).
    > 
    > On 10/28/20, 1:28 PM, "Sargent, Andrew J CIV USN NAVOBSY DC (USA)"
    > <andrew.j.sargent2.civ at mail.mil> wrote:
    > 
    >     Hello Again,
    > 
    >     I actually tried that as I was replicating another machine we have
    > setup, but I’m still getting the same error. Any other thoughts or
    > suggestions?
    > 
    >     Thanks,
    >     Andy
    > 
    >     On 10/23/20, 5:51 PM, "egreisen" <egreisen at nrao.edu> wrote:
    > 
    >         On 2020-10-23 17:12, Sargent, Andrew J CIV USN NAVOBSY DC (USA) 
    > wrote:
    > 
    >         Your DADEVS problem is one of formatting:  It must start with a 
    > minus
    >         sign and 2 blanks (or a + with 2 blanks).  The latter says use
    > this data
    >         area for all computers the former says use on NUIT only.
    > 
    >         Eric Greisen

    What are we talking about here - compilation which has nothing 
    whatsoever to do with DADEVS or running aips or some aips program which 
    does require DADEVS?  My advice had to do with the latter.  The DADEVS 
    processing script reads DADEVS.LIST (or other versions of the dadevs 
    data files) looking for the name of the host in the text of the data 
    area pathname.  The one we have here has 178 entries for a large number 
    of different computers and one wants (by default) to pick only those 
    appropriate to the local computer.  This on primate we want "-  
    /DATA/PRIMATE_1"  but not "-  /DATA/EVAN_1" by default (one could ask to 
    include evan's files too).  If there is a + sign, then the script will 
    take the stated data area no matter what its contents might be.  If 
    there were a + sign in the AOC's DADEVS.LIST file, then all 100 
    computers would include the same data area - generally a very bad idea.  
    Read the help file for AIPS - or Chapter 2 of the CookBook - to see the 
    many options available for data areas.  We now have a lustre disk farm 
    at the AOC and many people have a $HOME/.dadevs.always which is included 
    ahead of the site-wide DADEVS.LIST file.  In that a + sign is more 
    useful.

    Eric Greisen