Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin AT sources DOT redhat DOT com
Date: Sat, 11 Nov 2000 23:27:56 -0500
From: Christopher Faylor <cgf AT redhat DOT com>
To: Stephen L Moshier <moshier AT mediaone DOT net>
Cc: cygwin AT sources DOT redhat DOT com
Subject: Re: long double support in cygwin
Message-ID: <20001111232756.A26752@redhat.com>
Reply-To: cygwin AT sources DOT redhat DOT com
Mail-Followup-To: Stephen L Moshier <moshier AT mediaone DOT net>,
	cygwin AT sources DOT redhat DOT com
References: <Pine DOT LNX DOT 4 DOT 20 DOT 0011112221130 DOT 2423-100000 AT moshier DOT ne DOT mediaone DOT net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.6i
In-Reply-To: <Pine.LNX.4.20.0011112221130.2423-100000@moshier.ne.mediaone.net>; from moshier@mediaone.net on Sat, Nov 11, 2000 at 11:00:58PM -0500

On Sat, Nov 11, 2000 at 11:00:58PM -0500, Stephen L Moshier wrote:
>>>A search of the mailing list archive turned up what looks like
>>>another attempt to offer long double functions, over a year ago,
>>>as a version of mathinline.h.  Did that work lead anywhere?
>>
>>Not that I'm aware of.  I don't recall anyone submitting any code.
>
>The posting was by Tim Prince.  If you told him what you just
>told me, to go talk to newlib, then I suppose he would not have
>submitted any code to you.  But I don't see it in newlib either.

Most of us (the core cygwin developers) tell people who are proposing
changes to things that involve newlib that they should talk to the
newlib mailing list.  We are not infallible, however, and sometimes we
may miss a post or forget to suggest this.  Tim has been with the
mailing list long enough that he may already know this.

However, you've prompted me to search the archives to see what you're
referring to.  Tim's code was adapted from glibc.  We can't use glibc
based code.  That is undoubtedly why no one suggested folding his
changes back into newlib.

>>As to why it's not there, it's not there because no one has
>>implemented it.
>
>I don't think that explains it.
>You could have copied the implementations from the old linux libc,
>glibc 1 or 2 libraries, from libio, from gmp, or from DJGPP.  The printf
>in newlib is BSD.  That has previously been modified for long double
>support in linux libc 4, as a contributed item for DJGPP and, one would
>think, by BSD as well.  Why aren't you using any of these six
>or seven codes that various people have indeed implemented?
>What is the policy?

There are licensing considerations which prevent us from taking code
from LGPLed sources (http://sources.redhat.com/ml/cygwin/2000-11/msg00240.html).
I don't know exactly how many of the above this applies to but this is
not the main problem.

To the best of my knowledge no one has provided a copy of the software
that you desire to the newlib maintainers.  If someone does provide a
patch and that patch does not come from LGPLed or similarly encumbered
code then it would probably be acceptable.  It is that simple.

That's the usual policy in a free software project.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com