Description of Problem:

Evolution gives a "could not lock" /var/spool/mail/<myusername>
error whenever I try to access the spool from
evolution, if I use a precompiled RPM from
ftp.ximian.com. 
The spool is NFS mounted. NFS server and client
both run RedHat 7.1 and have nfs locking properly
configured. 

If I compile my own copy of evolution, everything
works fine. Reinstalling the RPM causes the 
problem to recur. 'make install' in the evo
source directory solves it. 

This occurs with both 0.12 and 0.13 and is solved
the same way on both. 



Steps to reproduce the problem:
1. Use a local NFS mounted spool (either mbox spool-format or standard UNIX) 
2. run evo. 
3. Should be able to access spool

Actual Results:
Get error message "could not lock" /var/spool/mail/<myusername>

Expected Results:
Should be able to read spool directory

How often does this happen? 
Always 

Additional Information:

I suspect this is related to compile time
settings set with ./configure. ./configure
seems to be auto detecting information on 
mail spool access, then building evo to
correspond with that. When I rebuild evo, the
proper settings are identified. When I use 
the ximian-compiled version, I'm getting the
wrong settings and the spool dir can't be
accessed. The solution in this case is to 
do this configuration i.d. at run-time or
installation-time.