Re: [boost] Formal Review for Interval Library (was: IntervalLibrary reminder)

From: Joerg Walter (jhr.walter_at_[hidden])
Date: 2002-09-07 10:12:12

----- Original Message -----
From: "Gabriel Dos Reis" <gdr_at_[hidden]>
To: <boost_at_[hidden]>
Sent: Saturday, September 07, 2002 4:29 PM
Subject: Re: [boost] Formal Review for Interval Library (was:
IntervalLibrary reminder)

> jhr.walter_at_[hidden] (Joerg Walter) writes:
> | > Next, how do we do the
> | > computation? One (or maybe the two intervals) need to be converted to
a
> | > common type, but we don't have such a conversion. It's always the same
> | > problem: we can't afford to convert between interval types with
different
> | > base types.
> |
> | You've lost me. Don't you can afford it, because it's too dangerous?
>
> It isn't clear that converting to the type with greater precision is
> really a solution -- you may get surpises by mixing precisions.  So if
> the user really does want to compute with the widest precision then
> he/she should just say so explicitly.
>
> | > This problem isn't really new. A lot of libraries suffer from the same
> | > limitation, beginning with the standard library (just try to do
> | > 'complex<double> = complex<double> + complex<float>').
> |
> | You're right (I had to check this ;-(. But is this a strenght of
> | std::complex and a weakness of the builtin data types?
>
> I'm not sure we should speak here of stenght or weakness since mixing
> precisions can prove to be trouble some.  In that view, I can
> undersstand that a general purpose library doesn't provide mixed
> precisions; it suffices it provides the appropriate conversion functions.

Ok, but the fact remains, that such a general purpose class doesn't behave
like the builtin data types and it should, IMHO. Would you really like to
abolish type promotion?

> | > You are right, it's not a problem of the interval library. The C++
> | > Standard says that "the effect of instantiating the template complex
for
> | > any type other than float, double or long double is unspecified"
(26.2.2).
> | > Not being able to instantiate complex<interval> because the Standard
> | > decided the only interesting types are float, double and long double
is
> | > really a pity.
> |
> | Does this mean, that we need boost::complex<>?
>
> No (for that sole reason), IMHO.

The only other way I see is, that standard library vendors document how to
extend their implementations and boost supports one ore more different
extensions then.

Regards

Joerg