Founded by Scott Dietzen, the same person behind WebLogic before it was acquired by BEA Systems Inc, Zimbra is an attempt to make email more manageable for administrators and more useful for end users.
The product, Zimbra Collaboration Suite, is an open source email server that can plug into existing email clients such as Outlook.
The product is designed as a collaboration platform that includes the usual grouping of email and calendaring. But it aims to transform email into a more active workflow engine through dynamic links with calendars, enterprise applications, or external services such as Google Maps.
Zimbra currently supports web services to link into enterprise systems or external services. But it is also looking to itself, or the open source community, to develop additional programmatic APIs so programs written in Java, PHP, Python, C/C++, C#, or other languages could invoke Zimbra services.
The server is based on three quarters of the LAMP open source stack. It taps Linux security and its highly efficient file system for storing the huge blocks of data that constitutes email. It uses the Apache web server for its scalability and caching strength.
The MySQL database is used as a metadata store to make the file system searchable. Additionally, it leverages storage networks to consolidate retention of messages and attachments.
The server side architecture is supposed to make the Zimbra email engine very attractive to organizations that are straining under increasingly rigorous data retention policies to make email accessible and manageable.
Content from our partners
For instance, thanks to the database engine, email can be searched, not only by sender or recipient, but also by topic or thread across multiple end users. Or you can search by person and date range, or attachment type, something not achievable with current email engines.
It also allows consolidation of storage, so if many recipients are sent the same attachment, the attached file itself can be stored only once. And, the server has the ability to render attachments as harmless HTML so invasive agents are halted at the secure Linux perimeter before they could infect machines.
In turn, integration with SAN or NAS storage means that the cost of email retention can be reduced because it can now be covered by corporate information life cycle management (ILM) initiatives.
However, like early Java before it, while the meat is on the server, it’s the client side that has drawn most of the attention. The clients are designed using Ajax technology, which enables email messages to contain dynamic links, either to urls or to triggering web services requests. The result is that the end user need not toggle back and forth between email client and browser to retrieve or send information.
For instance, when a meeting proposal arrives, mousing over the request could automatically generate a SOAP messages to query your calendar, with the result displaying as a pop up inside the email message. Or a description of the location of a meeting could dispatch a request to Google Maps to show where the destination is. Similarly, a reference to a customer could trigger a request to the CRM system for an update on either an order or their history of dealings.
When we first met Diezten at Java One back in 1997, most of the hype was around Java Applets. Having emerged before Flash became popular, applets were seen as the mechanism for livening up static browsers.
At that conference, Dietzen’s company, WebLogic displayed an appserver that supported Enterprise Java Beans and other elements of what would eventually become J2EE, before they were formally approved as standards.
For the record, it was easy to talk to people at the WebLogic booth because the crowds were elsewhere. Dietzen’s company made the right technology bets, and with its subsequent acquisition by BEA, the rest became history.
On this go-round, Diezten’s venture has a flashy client, but the value is in data retention and management back at the server. However, with email largely a platform play and IT projects expected to deliver tangible ROIs, it is difficult to imagine organizations opening up their pocket books to replace a core technology that, while imperfect, still usually works.
Diezten says the pressures of Sarbanes-Oxley and similar regulations may alter the odds of success in this admittedly uphill battle.
