(j3.2006) (SC22WG5.4261) wg5

[Van Snyder]

In 5.8 I would prefer "keyword type" to "token type".

The MOLD= clause of an ALLOCATE statement works for non-polymorphic
variables as well as polymorphic ones.  You might wish to mention in 6.2
that allocate(a,mold=b) gives A the bounds of B but doesn't copy the
value.

In 8.3, whether the integer status code is made available as an exit
status is independent of whether the processor supports command-line
execution.  In Unix, for example, one might load a program using "fork"
instead of "call system".  The exit status is equally interesting in
both cases.

In 9.1.2, "type integer this kind" should be "type integer with this
kind".  "has value" should be "has a value" (twice).

A full stop is needed at the end of the first line in 9.13.

A full stop is needed at the ends of the first and fourth paragraphs of
9.14.

In 10.3 I prefer the ther "keyword" to "token".

In 10.6, "optional dummy argument" should be "optional nonpointer dummy
argument".

In 10.8, between the examples, "of each" should be "of each element".

The alternative for the entry statement described in 10.9 isn't thread
safe if the individual module procedures need to share data, which would
be module data.

On Mon, 2010-04-26 at 09:09 -0700, John Reid wrote:
> WG5,
> 
> Here is a draft revision of my summary of the new features of F2008. Would 
> anyone like to comment before I put it on the WG5 site?
> 
> Best wishes,
> 
> John.