(j3.2006) (SC22WG5.3851) [ukfortran] 09-007/UTI 151

Malcolm Cohen malcolm
Wed Jan 14 03:53:08 EST 2009



Reinhold Bader wrote:
> Since the accessibility rules refer to an identifier and not the entity
> identified (5.3.2 para 1,2), I'd answer this with yes even under the
> current wording.
>   
Since the words in question (not in 5.3.2) are talking about entities, 
not identifiers, I find this argument to be utterly unpersuasive.  (And 
USE association provides access to entities, not their identifiers - the 
identifiers are merely *how* that access is syntactically provided.  See 
11.2.2p2.)

Anyway, the current wording is ambiguous.  Let's not pointlessly argue 
about whether one might guess at the intention and happen to end up with 
the answer you were thinking of, let's get it written clearly and 
unambiguously so that even I can understand it.

...
>   
>> It's also not clear to me just what this restriction is supposed to be
>> giving us - since in many cases it looks like one could wriggle around
>> it (unlike the previous version); that certainly makes it harder to
>> reason about.  Maybe the restriction shouldn't be there at all!
>>     
>
> In which case a rule like "host access beats use access" would be needed, at
> least for this particular situation.
>   
I certainly don't agree with *that* assertion. It's not obvious that 
such a rule would be "needed", indeed it is far from obvious that it 
would even be useful - on the contrary, it would obviously (to me) be 
useless, unnecessary and confusing.  If you have some reason for 
thinking such a rule would be needed, you will need to explain why.

As I said, the more I look at it the more it seems to me that this 
particular restriction is not useful, and that just dropping it (and 
using our normal rules) would be easier.  Maybe there is a case I've 
missed where it would cause a problem, but I want to see it; for a 
start, if it falls into the "can wriggle around" category then it could 
show up a serious hole in the whole feature.

Anyway, enough arguing generalities: hand-waving is cheap, concrete 
examples with reasoning as to why they should be allowed/prohibited are 
what's required if we're going to sort this out.  As I said before, I 
expect someone will be working on it at meeting 187; I'm certainly not 
intending to thrash it out inefficiently via email beforehand, as there 
are rather a lot of other things I have to do.

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





More information about the J3 mailing list