Attempting to build and test under Cygwin

[David E Sigeti]

Dear Sam,

At 07:13 PM 7/7/2004, Sam Hartman wrote:

> >>>>> "David" == David E Sigeti <sigeti at lanl.gov> writes:
>
>     David> Dear Kerberos Folks, I am attempting to build and test MIT
>     David> krb5 1.3.4 under Cygwin on a W2K machine.  The tests fail
>     David> in the db2 utility.  The output is (in part):
>
>I'm sort of impressed you got it to complie.  From time to time people
>say they want cygwin builds working and sometimes get it building but
>for the most part do not offer to submit clean patches to fix problems
>they encounter.  We wouldn't mind cygwin builds working but don't have
>resources to dedicate to it ourselves.

The changes required to build are fairly minor and seem that they should be 
in the code in any case.  For example, a function gets called even when its 
definition is not compiled because of conditional compilation.

However, there is still the problem of the failed test.  Any thoughts on 
that?  I should point out that I will only be running clients.  Is it 
possible that I don't need the database at all?

>We'd recommend working with current snapshots for generating patches 
>rather than 1.3.x releases.
>
>If you do have patches that are reasonably clean, please submit them
>to krb5-bugs at mit.edu.

I just submitted them.  The patch from the current snapshot is huge -- 
apparently, there are many files included in the snapshot that are not in 
the distribution.  The patch from the distribution, on the other hand, is 
very small -- only a few lines are changed.  Since I figure that you will 
want to review the changes anyway, I sent the patch from the 
distribution.  I hope that this is acceptable.

Yours,
David

---
Dr. David E. Sigeti
Phone:  505-667-9239
E-mail: sigeti at lanl.gov
Surface mail: MS-F645, Los Alamos National Laboratory,
               Los Alamos, NM  87545   USA