From Thomas.Wainwright at noaa.gov  Mon Jul  2 12:17:52 2007
From: Thomas.Wainwright at noaa.gov (Tom Wainwright)
Date: Mon, 02 Jul 2007 09:17:52 -0700
Subject: [Mapl-support] Introduction to MAPL--comments
In-Reply-To: <468718B3.7040303@mit.edu>
References: <46841EF6.5010406@noaa.gov> <468718B3.7040303@mit.edu>
Message-ID: <46892530.9090206@noaa.gov>

Chris,

I'm using g95 0.91!, Jun  5 2007, with gcc v4.2.0.  The MAPL libs compile fine,
but I get the following error building the application code:
	In file geopk.F90:252

	      pointer (ptr_inbuf8, inbuf8)
              1
	Error: Unexpected array reference at (1)

The same error is repeated for line 253.

I think the pointer() function is a Fortran extension that is not implemented by
g95.  I think gfortran supports it, but I haven't managed to get a clean build
of esmf with gfortran 4.1 or 4.2 (but that's a different support list).

--Tom Wainwright

chris hill wrote:
> Tom,
> 
>   Thanks for the feedback.
> 
>   So far MAPL and the fvhs examples have been used with the Absoft and 
> Intel compilers. I have it on my list to test with the latest gfortran, 
> which is supposed to have improved quite a lot in the last set of revisions.
> 
>   Do you know which version of g95 you are using?
> 
>   We will try and get the documentations sorted out ASAP. Please let us 
> know if/when you find more typos.
> 
> Chris Hill
> Tom Wainwright wrote:
>> Dear Max and co-authors,
>>
>> I didn't make it to this year's ESMF community forum, but one of my staff (Jim
>> Colbert) did, and was impressed with the potential of MAPL.  I decided to see if
>> I could use it, following the tutorial instructions on the wiki.
>>
>> First, thanks for the documentation on the wiki!  It provided enough to get me
>> started.  First, I downloaded the fvhs tree, and tried that.  I made progress,
>> but couldn't finish the compile on my system (Linux-x64/g95/openMPI) due to some
>> pointer-related errors deep in the numerical code (g95 doesn't like geopk.F90,
>> lines 252,253).
>>
>> The latest (June 12, 2007) version of "An Introduction to MAPL" appeared about
>> that time, so I decided to try the simpler Example 1 in that document, by
>> pulling all the source code in the fvhs/src/Application directory and replacing
>> it with the Example 1 code.  After much tweaking of the top-level GNUmakefile
>> and .esma_arch.mk, I got it to compile and run.
>>
>> However, there are a few errors in the "Program Example1" code ("An Introduction
>> to MAPL" pdf, p. 7-8) that you might want to fix in the next edition:
>>
>>     use ESMF: should read:
>> 	use ESMF_Mod
>>     use MAPL ...: has several syntax errors, should read:
>> 	use MAPL_Mod, only : SetServices=>MAPL_GenericSetServices
>>     clock = ESMF_ClockCreate(...: needs '&' at end of first line,
>> 	and "rc=STATUS" should read "rc=RC"
>>
>> Thanks again for MAPL and the documentation!  On to Example 2 next week.  Once
>> we get it all figured out, I think it will greatly simplify getting our
>> biological code into ESMF.
>>
> 
> _______________________________________________
> Mapl-support mailing list
> Mapl-support at mit.edu
> http://mailman.mit.edu/mailman/listinfo/mapl-support
> 

-- 
Tom Wainwright
NOAA Northwest Fisheries Science Center
Newport, Oregon
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The contents of this message are mine personally and do not necessarily
reflect any position of the Government or the National Oceanic and
Atmospheric Administration.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~