Re: Error compiling templates

[craig at jcb-sc dot com]

>We are not paying for support NOR we are making commercial use of GNU/Linux.
>For clarity, I retract my statement "DROP our support for" to "DROP our
>plan/idea to support".  I do admire the works of the people behind the
>GNU/Linux software and have kept in mind that monetary compensation is an
>issue.

Ah, okay, thanks for clarifying that.

>We WOULD be prepared to pay for support.  But as it stands, we will not have
>the need.  Our executives have decided long ago that there will be no Linux
>port of our software due to among the many reasons: 1) compiler is broken 2)
>little or no quality "paid" support ...

Hmm, well, if they decided that "long ago", for a sufficiently "lengthy"
definition of that phrase, and if they based that decision on the state
of g++ at that time, then I think many of us can understand why they
came to that decision.

g++, even more than gcc or g77, experienced some tough times a couple
of years ago plus or minus a year or two.  My impression is that the
EGCS project, now renamed GCC as of 2.95 (which would have been EGCS 1.2),
has improved g++, even more than it improved gcc (or g77).

Not that g++ couldn't (perhaps) be said to be "broken", but if your
executives haven't looked closely into how much g++ has improved,
especially into how much work goes into it in the GCC/EGCS project,
they've missed an important part of the equation.  Especially if they
haven't looked into its progress over the past 12 months or so.

(Personally, I don't pay much attention to g++, as I believe C++ will
turn out to be a disaster on at least the scale, though not the short
time-span, of Y2K.  But I don't really have a good choice to recommend,
and can't help but notice what appears to be a huge, and at least reasonably
competent, effort to make g++ the best C++ compiler on the planet.
Moreover, some of the GCC developers who I most admire, for reasons
beyond mere technical competence, do most of their work on g++.)

Finally, while the above covers mostly "free" support (i.e. you can
leverage off that without paying for it), I'd be surprised if it was
still the case (if it recently was) that "little or no quality" paid
support for g++ existed.

But, if that's the case, then that represents a great business opportunity
for someone -- a niche needing filling.

(Try filling that same support niche with those proprietary C++ compilers
I assume you already support.  Step 1: get the source code.  Step 2:
get the rights to distribute it in modified form.  Etc.)

>My strong admiration for Linux is the only motivation I started this
>conversation. But I am convinced that my request will not be entertained.

A closed mind is a terrible thing.  ;-)

        tq vm, (burley)