(j3.2006) "Contiguous" doesn't quite do what I need

[Aleksandar Donev]

On Friday 23 March 2007 14:31, dick.hendrickson at att.net wrote:
> uses an erronous contiguous declaration.
I think that with the last iteration we eliminated almost all "unsafe" uses of 
contiguous. That is,  it should no longer be possible to have something 
declared as a contiguous be non-contiguous at runtime. For arguments, a copy 
is made if needed. Pointers can only be pointed to 
"compile-time-contiguous" (aka always-contiguous unless J3 renamed my 
proposal at the last meeting). I think this is a good balance between 
felxibility and safety.
Aleks