Why shouldn't he? Plans change, platforms evolve. Simple case in point:
Apple, a few years ago, changed from PPC to Intel chips. Devs whose code
assumed a particular endianness - perhaps because they didn't care about
"portability" - were in a world of hurt during the transition.
<shrug> I don't see the parallel. Difference in endianness is a known
problem. Because "endianness" is a universal concept. Differences in
accessing a web service do not exist because very few OSes actually
support such an access. So, a cross-platform (or, rather, a
platform-independent) solution involving endianness is necessary. A
platform-independent solution for accessing a web service is not possible.
No - not really. I used Boost as an example of a portable library that
could be legitimately referred to here in the general case.
In the general case you can. The OP wasn't talking about a general
case. At least that's what I gathered when I recommended asking in the
OS newsgroup. What was so wrong about it that you can't stop and think
about your objection and simply see that your efforts are off-target?
The rest, the notion that it applies specifically to the OPs question
about web services for instance, comes from your own hyperactive imag-
ination, for which I'm not responsible.
<sigh> My "hyperactive imagination"... Listen to yourself. So, it's my
hyperactive imagination that insists on suggesting the OS newsgroup for
inquiries about web service access, right? And for all other
OS-specific frameworks or libraries. And you keep bringing up an
IRRELEVANT portable library that doesn't address the OP's issue in the
lightest. And scorn me for not agreeing with you... I give up. It's
like arguing with a child, and I have more than enough of that at home.
V