Usenet 1994 January/usenetsourcesnewsgroupsinfomagicjanuary1994.iso/sources/std_unix/volume.20/text0009.txt

Internet Message Format | 1990-08-02 | 1.6 KB

From:  [e-mail]

In article <[e-mail]> Peter da Silva wrote:
>In article <[e-mail]> [Doug Gwyn] writes:
>>     c) People who are concerned with portability with not stray
>>        outside the domain that the standard guarantees to work.
>
>In the real world, people will write programs using their system and manual
>as a working base. If their system says that mmap() works such-and-such a
>way, they'll use it. They may even think that their program is portable,
>because mmap is in 1003.4.
>
>> You seem to be arguing that extensions should be prohibited.
>
>No, I'm arguing that extensions should be explicit. You should have to perform
>some magic juju so that you *know* you're stepping outside the POSIX consulate
>into the anarchic land of Vendor-Specificia.

This sounds like a job for... the Mindcraft C Portability Verifier!
Among other things, this product lets you check conformance to ANSI C,
1003.1, 1003.2 C interfaces, XPG3, or any other standard for which you
write a definition.  Pick any one of these standards or any combination
of them.

I'll post a full description to comp.newprod.

Vaporware right now, but condensing even as I type.  It'll ship a few
weeks from now.

>Like calling mmap() instead of whatever the POSIX routine is.

I hope we don't have to name a new interface every time a new standard
restricts or changes the syntax of an old one.  Especially when the
old interface is difficult to use portably anyway.
--

    Chuck Karish        [e-mail]
    Mindcraft, Inc.        [phone]

Volume-Number: Volume 20, Number 10