Date: Mon, 10 Jun 2002 13:44:29 -0500
From: Eric Rudd <rudd AT cyberoptics DOT com>
Subject: Re: ISO C99 double math functions
To: djgpp-workers AT delorie DOT com
Cc: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>, sandmann AT clio DOT rice DOT edu, Kbwms AT aol DOT com
Message-id: <3D04F38D.D1CC5FE9@cyberoptics.com>
Organization: CyberOptics
MIME-version: 1.0
X-Mailer: Mozilla 4.72 [en] (Win95; U)
Content-type: text/plain; charset=us-ascii
Content-transfer-encoding: 7bit
X-Accept-Language: en,pdf
References: <Pine DOT A41 DOT 4 DOT 05 DOT 10206100849160 DOT 32640-100000 AT ieva06>
Reply-To: djgpp-workers AT delorie DOT com

Andris Pavenis wrote:

> On Sun, 9 Jun 2002, Eli Zaretskii wrote:
>
> > The main drawback is that we lose the fast libc routines added for
> > v2.03.  I don't want to lose them.

Nor do I.

> Or other way: it could be possible to have 2 versions of libm -
>         - one fast one using coprocessor
>         - slower one based on fdlibm

My suggestion would be to put the new functions in libc in order to avoid FAQs,
but start by simply copying the appropriate routines from libm that are not
currently present in libc.  That way, users wouldn't have to say -lm to get the
new functions, the math functions I coded a few years ago would still be linked in
by default (there is a big speed penalty in some of the libm routines), and we'd
have the new routines with minimal development effort.

As K.B. Williams (or others) write the new routines, they could be incorporated
into libc, but there would no longer be the pressure to get them all written at
once; the worst ones could be re-written first, and the better ones later (if
ever).

Are there any IP/copyright issues to this approach?

-Eric