Mailing-List: contact new-httpd-help@apache.org; run by ezmlm
Precedence: bulk
Reply-To: new-httpd@apache.org
From: "William A. Rowe, Jr." <wrowe@lnd.com>
To: <new-httpd@apache.org>
Subject: RE: [patch 2.0]  htdigest  Linkage mechanisims
Date: Thu, 30 Mar 2000 13:17:58 -0600
Message-ID: <001401bf9a7c$aac5dfa0$345985d0@corecomm.net>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
In-Reply-To: <4.3.1.2.20000330133806.00b08360@pop.ma.ultranet.com>
Importance: Normal

> From: Greg Marr [mailto:gregm@alum.wpi.edu]
> Sent: Thursday, March 30, 2000 12:43 PM
>
> At 07:15 PM 03/29/2000, William A. Rowe, Jr. wrote:
> >Perhaps it's time for a walk through of the matter, (like people keep
> >saying, code is the meat) so let's take a simple patch, htdigest,
> >which
> >won't compile today anyway.  I hope this explains the whys of my
> >approach
> >with a -real world- example.
> >
> >We change some compiler directives [don't know why /D "_MBCS" was in
> >there
> >in the first place; left it alone]
>
> MS puts it there by default on all new projects... really annoying.

If this is the sole reason, then we kill it.  I'm also not sure that /D
"_CONSOLE" does anything for us.

> >add the libraries that are _actually used...
> >   # ADD LINK32 + libcmt.lib oldnames.lib apr.lib ws2_32.lib
> >
> >tell the linker that's all the libraries...
> >   # ADD LINK32 + /nodefaultlib
>
> I still don't see the need for this.  It's been my experience that a
> conflict in linking between libcmt and msvcrt is generally because
> some files are compiled assuming static c library and some with
> dynamic.
>
> >add the map output for debugging gp faults and user bug reports...
> >   # ADD LINK32 +
> > /map:"../build/win32/release/support/htdigest/htdigest.map"
>
> Can't you just use /map, since the default name is the same as the
> exe, just with a different extension?

Possibly.  I've been meaning to take a look at how flexible the build
environment is in accepting $() parms within the editing blocks.  Who
knows, that might resolve a whole set of issues, in itself.

We will need to set up something to package the .map's into the release
candidate, though, since we will need those *explicit* files to really
help troubleshoot offbeat bug reports.