Re: Relation between P1788 and P1788.1
Yes, that's a good point about USB.
Also, I see what you are saying,
that, with an environment (circuitry, system, and language)
that conforms to standard B, with standard B a subset
of standard A, there can be programs conforming
to standard B but not to standard A. Certainly, standard
B might have incompatible extensions. I only need to think
back to the state of Fortran 77 compilers prior to Fortran 90.
Is it possible to, and if so, should we, put language in
1788.1 that any extensions with functionality the same
as 1788 should conform to 1788? It seems to me it
would be very hard to specify which extensions these
would be.
Nonetheless, the stated purpose of 1788.1 clarifies that
it is meant to be a step towards implementing 1788.
You may also wish to communicate with Ned, who might
further clarify the reasons for introducing 1788.1.
Best regards,
Baker
On 12/08/2014 09:59 AM, Vincent Lefevre wrote:
On 2014-12-08 08:48:26 -0600, Ralph Baker Kearfott wrote:
Vincent,
Yes, that is true. Certainly, that would be true of any
standard that is a subset of any other standard. For instance
a USB 3 conforming device (USB is a standard but not an
IEEE standard) might not function in a USB 2 port, but
a USB 2 device should function in a USB 3 port, n'est pas?
Well, this is not exactly the same thing. I'm not sure about USB,
but one could imagine USB 2 devices with extensions such that
such a device would work in a strict USB 2 port, but would fail
to work in a USB 3 port due to incompatible extensions. It seems
that USB 3 uses additional pins. Imagine that the device uses
these additional pins for some extension over USB 2: assuming
they are ignored in USB 2, the device would work in a USB 2 port,
but they may be a clash with a USB 3 port.
Then, there's the question of the conformance definition for a
device. Are extensions forbidden, even though they are ignored
for the concerned standard? That's up to the standard to be clear
on this point.
In the P1788 context, it would be the language that would define
the conformance of a program.
--
---------------------------------------------------------------
Ralph Baker Kearfott, rbk@xxxxxxxxxxxxx (337) 482-5346 (fax)
(337) 482-5270 (work) (337) 993-1827 (home)
URL: http://interval.louisiana.edu/kearfott.html
Department of Mathematics, University of Louisiana at Lafayette
(Room 217 Maxim D. Doucet Hall, 1403 Johnston Street)
Box 4-1010, Lafayette, LA 70504-1010, USA
---------------------------------------------------------------