[17120] in Kerberos_V5_Development

home help back first fref pref prev next nref lref last post

Re: [RFC][PATCH] krb5 => libverto main loop

daemon@ATHENA.MIT.EDU (Luke Howard)
Thu Aug 25 10:48:39 2011

Mime-Version: 1.0 (Apple Message framework v1244.3)
From: Luke Howard <lukeh@padl.com>
In-Reply-To: <1314283154.24488.55.camel@t410>
Date: Thu, 25 Aug 2011 14:48:28 +0000
Message-Id: <AECF331C-F21A-4EFF-8BFF-8999F79C2DE9@padl.com>
To: Greg Hudson <ghudson@mit.edu>
Cc: "krbdev@mit.edu" <krbdev@mit.edu>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Errors-To: krbdev-bounces@mit.edu


On 25/08/2011, at 2:39 PM, Greg Hudson wrote:

> On Thu, 2011-08-25 at 05:24 -0400, Luke Howard wrote:
>> If this for dlopen(), Darwin uses .so. For shared libraries, yes, it uses .dylib.
> 
> You're correct.  This presents a more serious architectural problem for
> verto on Darwin, since it assumes that shared libraries and loaded
> modules have the same suffix.  I'm not immediately sure how to resolve
> this.

Recent versions of Darwin let you dlopen() shared libraries (e.g. looking at the PAM modules in Lion, they're MH_DYLIB not MH_BUNDLE; on the other hand, libtool -module creates MH_BUNDLE). So the old distinction between the two is probably not as important as it once was, and you could install install modules with the dylib suffix and it not be so confusing. Depends too on how far back (pre-Leopard?) you want to support.

-- Luke


_______________________________________________
krbdev mailing list             krbdev@mit.edu
https://mailman.mit.edu/mailman/listinfo/krbdev

home help back first fref pref prev next nref lref last post