(j3.2006) co-array variable

Bill Long longb
Fri May 25 18:04:20 EDT 2007



Aleksandar Donev wrote:

>On Thursday 24 May 2007 17:45, Malcolm Cohen wrote:
>
>  
>
>>...in that context, saying "scalar" achieves nothing.  Arrays and
>>structures are comprised of scalars.
>>

The phrase says (after replacing co-variable with co-array) "scalar 
co-array".  The examples Aleks has below are not co-arrays by either the 
current or previous definition.  However, with the new definition of 
co-array, x(1) is a co-array.  The original intent of [197:13-15] (the 
result of considerable discussion and compromise) was to not allow array 
elements or structure components in this context.  Such things were not 
co-arrays by the old definition, so the wording used to be correct.  
This section was not updated when the definition of co-array was 
expanded in paper 07-210.  This was an oversight. I don't think 
"co-array of rank zero" is materially different from "scalar co-array", 
though Aleks' intent is correct.  It's wordy, but I think we need 
something like "scalar co-array with no subscripts or component 
selection" as the replacement for "co-variable" on [197:13].  The 
"co-variable" on [197:15] can be replaced by "it" to avoid repeating the 
longer wording.  The editor is welcome (as usual) to suggest better wording.

Cheers,
Bill

>>    
>>
>I believe those words should say "co-array of rank zero" (I thought we had the 
>term "co-scalar" for this but acroread does not find it). That is,
>
>x[5]%atomic_component=...
>and
>x(1)[5]=...
>
>should obey the normal segment ordering restrictions and not be given the 
>"atomic read/write" exclusion.
>
>Aleks
>
>  
>

-- 
Bill Long                                   longb at cray.com
Fortran Technical Support    &              voice: 651-605-9024
Bioinformatics Software Development         fax:   651-605-9142
Cray Inc., 1340 Mendota Heights Rd., Mendota Heights, MN, 55120

            

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://j3-fortran.org/pipermail/j3/attachments/20070525/5f759daa/attachment.html 



More information about the J3 mailing list