[17491] in Kerberos_V5_Development
Re: Proposed platform assumption changes
daemon@ATHENA.MIT.EDU (Chris Hecker)
Mon Jan 30 13:57:40 2012
In-Reply-To: <tslhazdxfid.fsf@mit.edu>
MIME-Version: 1.0
From: Chris Hecker <checker@d6.com>
Date: Mon, 30 Jan 2012 10:56:35 -0800
To: krbdev@mit.edu
Message-ID: <e364d1f9-d263-4312-9909-01d1de98bee6@email.android.com>
Cc: Danilo Almeida <dalmeida@mit.edu>, Simo Sorce <simo@redhat.com>,
krbdev@mit.edu
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Errors-To: krbdev-bounces@mit.edu
I am almost certainly unique in this, but I build krb5 as a static lib to link into my game, so real msvc support is important to me, not just the ability to dynalink a dll built with mingw.
Chris
Sam Hartman <hartmans@MIT.EDU> wrote:
>>>>> "Nico" == Nico Williams <nico@cryptonector.com> writes:
Nico> I believe Windows builds need to support MSVC.
Why? I believe that the resulting SDK needs to work on applications
built with msvc. But why does MIT Kerberos need to support msvc for
Windows development? I can see a number of reasons, but the issue
really seems subject to enumeration of what a proposal to use mingw
should have to support.
I'm skeptical of "you must support this," but very likely to believe "to
use mingw you'd have to show that x y and z are not a problem."
Part of what I'm saying is that there a fairly small number of people
who are going to build KFW, getting the environment set up is already
tricky, and reducing KFW's build dependencies would probably not be the
most important thing on my list of things to optimize for in future KFW
work.
However, making sure you can debug on Windows and similar issues like
that seem very important.
_____________________________________________
krbdev mailing list krbdev@mit.edu
https://mailman.mit.edu/mailman/listinfo/krbdev
_______________________________________________
krbdev mailing list krbdev@mit.edu
https://mailman.mit.edu/mailman/listinfo/krbdev
