Choosing a Linux distro (part 2)

By now I’d done my fair share of reading up on the current state of Linux and how to get it to work as a server back-end for a Windows desktop environment. I was pretty confident that a recent Linux build with recent Samba version would handle the domain control (network authorisation) and file sharing tasks. So my mind turned back to the choice of software to replace Exchange Server as my email and PIM back-end.

There are a number of Linux based email server products that claim at least some level of Exchange compatibility:

• Open Exchange (http://www.open-xchange.com/EN/developer/)
• Scalix (http://www.scalix.com/)
• Kerio Mail Server (http://www.kerio.com)

For links and reviews of most of the competitors try the ServerWatch web site at http://www.serverwatch.com/stypes/index.php/TWFpbA==

I had just tried Scalix with the Xandros Server product that I had been forced to send back. Exchange “compatibility” can mean many things it seems. To Scalix it seems to mean that they might provide (I never found it) an Outlook plugin that makes their back-end server look like Exchange. Of the promised migration tools there was no sight. I was not alone among the user forums in experiencing a distinct level of underwhelment with Scalix!

Open Exchange looks like a possible contender. It’s available in commercial and free variants with an Outlook plugin (the “Oxtender”) that – again – promises to make the back-end look like Exchange to Outlook. I haven’t tried it.

The reason is that I – somewhat belatedly – remembered that I have previously used Kerio MailServer as a (then) cheaper replacement for Exchange Server in several client installations - and all had worked very well. Again the product relies on an Outlook plugin (this time called the “Kerio Outlook Connector”) to connect the Outlook desktop client to the back-end server.

Why all these connectors? Well, Microsoft being Microsoft, OE not only use proprietary protocols to communicate with each other, MS keeps the details of the protocols close to its corporate chest. In the Open Source arena there is a surprising and woeful lack of standards for groupware message exchange. This is a subject I may come back to in a later post as – to this grey-bearded old software designer – it encapsulates a lot of the problems that beset the Open Source community and prevent wider up-take of the technology and product offerings.

So – anyone who wants to provide a back-end email/groupware server and wants any kind of market share needs to address Outlook (I haven’t looked up the figures lately but it is surely the dominant email client in the corporate world) – and that means at the very least writing a connector to bridge the gap.