(j3.2006) IEEE and the GPU world

[Malcolm Cohen]

> Van Snyder wrote:
>> Bill Long wrote:
>
>>> Declaring IEEE_ARITHMETIC to be INTEGER seems like a hack.
>>
>> It's a hack, but not a new feature.  It works now.  Maybe.

Things work now, but probably not this one.

>> [10-007:444:21-22] says "any entity of the host that has this as its 
>> nongeneric name is inaccessible by that name by host association."  So if 
>> IEEE_ARITHMETIC is not accessible by that name, are the things gotten from 
>> it, but perhaps not mentioned in an ONLY clause, still accessible, or not?

*IEEE_ARITHMETIC the module name* is not in itself "accessible" ever.
*IEEE_ARITHMETIC the module contents* can be accessible.

This is one of the bits of lower-level editorial cleanup that should be done. 
Clearly the statement about IEEE_ARITHMETIC accessibility can only be referring 
to the contents of the module (since modules names are never "accessible" - they 
are plain global identifiers and do not appear in any access list).  Inserting 
"any entity from" or similar words somewhere into those sentences should be 
sufficient to clarify this.

Bill Long wrote:
>>>  It might be clearer if there were something more like USE,exclude :: 
>>> IEEE_ARITHMETIC,  or even UNUSE :: IEEE_ARITHMETIC.

For use association, ONLY (by the USEr) and PRIVATE (by the module writer not 
intending to export the IEEE stuff) are already your friends.

We've discussed USE,EXCEPT before, and it never got traction due to potential 
ambiguities and user confusion IIRC.  Maybe its time has come, but I doubt it. 
I have no strong opinion on it either way (I might have once I remember the 
technical details from the discussions last time it came up).

The story is a bit different for host association.  Well, if the problem is that 
if you stick USE IEEE_ARITHMETIC into the top level of your subroutine then 
there isn't an easy way not to access its contents, then the solution for now is 
"don't do that".  This is almost always easy to avoid doing.  There is more 
scope here for action in a future revision.

It looks to me that we already have adequate support for what people will want 
to do here.  That support can doubtless be improved, but given adequacy there is 
no need for an emergency response.

(And as per my previous message, IEEE_ARITHMETIC itself should not be much of a 
problem due to the low level of requirements it imposes.)

Cheers,
-- 
................................Malcolm Cohen, Nihon NAG, Tokyo.