From: pavenis AT lanet DOT lv
Message-ID: <B0000106636@stargate.astr.lu.lv>
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>, djgpp-workers AT delorie DOT com
Date: Wed, 27 Oct 1999 15:18:45 +0300
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
Subject: Re: gcc-2.95.2
References: <B0000106605 AT stargate DOT astr DOT lu DOT lv>
In-reply-to: <Pine.SUN.3.91.991027131905.28265E-100000@is>
X-mailer: Pegasus Mail for Win32 (v3.12a)
Reply-To: djgpp-workers AT delorie DOT com

On 27 Oct 99, at 13:22, Eli Zaretskii wrote:

> 
> On Wed, 27 Oct 1999 pavenis AT lanet DOT lv wrote:
> 
> > One change is that -fstrict-aliasing is no more enabled by default (I 
> > myself put '%{!no-strict-aliasing: -fstrict-aliasing}' in specs to
> > have that still enabled).
> 
> Is this a good idea?  If the GCC maintainers decided to not enforce 
> strict aliasing, why should we?  AFAIK, this feature can subtly break 
> lots of code.

No problems. No such hack in binary archives. I only editted specs 
after I installed it. I really didn't met problems with -fstrict-aliasing.
So if I'll have them I could use -fno-strict-aliasing to compare.
But binary archives of gcc-2.95.2 will not have strict aliasing enabled
by default

> 
> > One question: do we still need gcc-2.95 related files in DJGPP 
> > distribution?
> 
> Probably not.  But I'd suggest to leave it on SimTel until 2.95.2 is 
> uploaded.  With such a high rate of releasing new versions, we don't
> have enough confidence in the current version, so it's good to have a 
> fallback.
> 
> I understand that 2.95.3 is expected in a few (8-10) weeks.
> 

Maybe. Also DJ should update weekly mini FAQ not to point to gcc-
2.95 archives (ZIP picker seems to be Ok, except it points to 
lgpp295b.zip instead of lgp2951b.zip now but I don't think this is a 
serious problem) before gcc-2.95 archives are removed.

Andris