Re: Direct Quote from the RM

[anon@anon.org (anon)]

If you read below the quote you would have read that I included all 
paragraphs under the heading "Implementation Requirements", 
"Implementation Permissions", "Implementation Advice" and "NOTES".

With that said, Section 1.1.5 paragraph 11, under the  
"Implementation Permissions" heading states that the usages of 
"compile line switch" converts the 'Standard Ada Compiler' to a 
'Non-Standard Compiler' mode of operation that can reject 
compilation_units that do not comply with the "Non-Standard" mode 
or to halt compiling due to too many "errors/warnings".

Note: 'mode of operation' has to do with "compilation environments".

So, adding "-fstack-check" converts the standard GNAT compiler to 
an "Non-standard compiler". 

Also, the "-fstack-check" switch does not reject compilation_units 
or halt process because of to many errors/warnings.  It only halts 
the process if the Storage_Error exception is not trapped and a 
Storage_Error is raised, which could be apart of the algorithm. 
So, it does not meet the definition of a "command line switch" in 
Section 1.1.5 (11).

Also requesting that this switch becomes apart of the default 
suggest that one wants all future GNAT Ada compiler to be 
"Non-Standard" only Ada compile. 

And does a 'Non-Standard' compiler imply "HIGH INTEGRITY and 
HIGH RELIABLE" software, based on the RM? The answer is 'NO!'


In <fi07no$t2$1@jacob-sparre.dk>, "Randy Brukardt" <randy@rrsoftware.com> writes:
>"Adam Beneschan" <adam@irvine.com> wrote in message
>news:d5495389-9ee7-44f5-8a3e-864afdc7edee@s36g2000prg.googlegroups.com...
>> On Nov 20, 4:32 am, a...@anon.org (anon) wrote:
>> > For Randy Brukardt.
>> >
>> > You did not need to post three copies of you post! I read the first one.
>> >
>> > Since, you want to sub-divide the RM that's fine. From 1.1.2 using (2
>> > and 3) states the core starts in Section 1 .. 13. So using
>> > "Section 1 General" paragraph 1.
>>
>> Randy, I think he's got you there.
>
>Not sure why you think so, none of the language talking about compilation
>environments is echoed there. And the original discussion was about
>compilation environments, not about the language itself.
>
>> Perhaps I should make a formal
>> request in Ada-Comment that 1.1.2(3) be changed from "Sections 1
>> through 13" to "Sections 1.1.3 through 13".  And maybe I'm the right
>> person to submit this, since I seem someone who worries way too much
>> about the implications of taking the RM's language hyper-literally.
>
><grin> I don't think there is anything wrong with 1.1 or 1.1.1. It *is*
>unusual that Section 1 isn't marked as Redundant in the AARM, because it is
>just the normal non-normative introductory text. (Most clauses in the RM
>start with introductory text that doesn't include anything normative - it
>just makes the purpose of the clause easier to understand for the reader. A
>big difference between the Ada standard and many others is the attempt to
>make it readable to ordinary programmers.) The real problem is that 1.12
>doesn't clearly note that introductory text is non-normative - but that's
>probably because there are some counter-examples.
>
>> Maybe this should be brought up at an ARG meeting, for comic relief to
>> release some stress before delving into the next Baird issue.  (I
>> notice you finally changed the !recommendation section of AI05-0051,
>> by the way...)  :) :)
>
>Nawww, Tucker just did his homework and provided a new version, which
>necessarily included a real recommendation. Of course, if you read his
>e-mail that ends the !appendix of the e-mail, you'll note that the effect of
>the AI hasn't changed any. :-)
>
>                           Randy.
>
>
>