[sork] Will vacation 2.2.1-cvs work with Horde 2.1 ?

[Eric Rostetter]

Quoting Christopher Crowley <ccrowley at tulane.edu>:

> I recently updated the backend mail server. Part of that is a new vacation
> binary.  It doesn't like the "null" .vacation.db, .vacation.pag and
> .vacation.dir files that vacation currently puts in place.

At least in CVS versions, it only puts empty ones in place if configured
to do so.  I'm not sure about the current release version.

I don't think it is well documented, but at least the CVS RELENG and CVS HEAD
(and probably the latest release, but haven't checked) allow you to use
various file formats for the db files.   From vacation/config/conf.php (from
CVS RELENG tree):

// The type of database file to use if using the forwards backend.
// Valid types are:
//
// empty:      An empty file
// hash:       Berkeley DB 2.x hash, little endian
// hash3:      Berkeley DB 3.x hash, little endian
// btree:      Berkeley DB 2.x btree, little endian
// gdbm:       GNU dbm (gdbm) database, little endian

$conf['server']['params']['dbtype'] = 'empty';

So the default is the empty file.  But it can be others.  Start by
generating a vacation setup manually, then use the unix "file" command to
find the type of db file needed, and set the config as needed.

> I have plans to
> upgrade my whole instance to the latest horde later this month.  But, I need
> to resolve this vacation problem now.

Use current release, or if it does have the above, then the CVS RELENG version.

> I am considering dropping the vacation '2.2.1-cvs' into the production
> instance.  I wonder if anyone can tell me if this will work?

Yes, it should work fine.  2.2.1-cvs is RELENG and should work fine with
any release versions.

> I will test it
> first, but authoritative information would be welcome.

Should work fine.

> Christopher Crowley
> Technology Services
> Tulane University
> ccrowley at tulane.edu
> 504.314.2535
> PGP Key ID: 0x7CF18FDB

--
Eric Rostetter
The Department of Physics
The University of Texas at Austin

Why get even? Get odd!