Further Thoughts on Oracle Collaboration Suite 10g, Sep 8

In my recent review of Oracle Collaboration Suite 10g I listed some questions that I was planning to ask Rakesh and Jason from Oracle during my scheduled briefing session. Here's the answers as I understand them.

Q1. Requirements for Including External Users in a Workspace
External users have to be listed in the OCS Directory. OCS uses its own LDAP Directory Server for all directory-related tasks, and it synchronizes with one or more external directories in order to capture user information. So, based on that, an organization could automagically bring in internal users from their Active Directory, and have a second and separate LDAP directory for external users.

In reference to directories and directory integration, I asked whether technology from the recent Oblix acquisition was included in OCS 10g. It's not currently, but that is definitely on the product development roadmap.

Conclusion: Pretty standard stuff here. No great surprises, and no great deviation from current practice in other areas. This area will improve via the integration of Oblix ideas and concepts.

Q2. Addressing Pillar 2, for Offline Access
As I outlined in Pillar 2 (Location-Independent Access to Team Data, People and Applications), there is still a need for offline access to the systems and data that people use on a daily basis. Network connectivity is not always guaranteed, nor cost-effective, although that is definitely improving. OCS 10g addresses this need for offline access by using the "natural clients" that people use today. Email messages and discussions from Workspaces can be read in an IMAP email client. Documents stored in Content Services or Workspaces can be accessed offline in Windows Explorer, thanks to technology licensed from Xythos. And calendar items can be shown in calendaring clients, such as Outlook.

Conclusion: This is well and good, but insufficient. By having people go to three different "natural" places when offline means that the experience is "unnatural" compared to the online approach. For example, when working with OCS Workspaces online, messages, documents and calendar items are displayed together in a single unified interface, but when working offline, these are spread across three different clients. The integration is lost, the context is lost, and the flow is lost. I argue that this needs to be improved through the addition of an integrated rich client, one that delivers an equivalent experience both offline and online. If Oracle wants to break into the corporate market, then they should focus their initial development efforts on a Windows client.

Q3. Addressing Pillar 4, for Suite-Wide Calendar Consolidation and Free/Busy Searches
In Pillar 4 (Coordinate Schedules with Team-Aware Scheduling Software), I argue that with the proliferation of calendar instances across collaborative workspaces, the calendaring story has become even worse for end users. Calendar instances generally don't reference each other, and free-and-busy searches do not take into consideration all of the standing appointments that people have.

Oracle has done it right!. I was absolutely delighted at what Oracle has achieved with calendaring in OCS 10g. Due to the presence of a strong calendar server in the Suite, meetings created in Workspaces automatically show on the individual's calendar, and are taken into consideration when a free-and-busy search is undertaken. Sure, this only deals with internal calendaring, but it's a whole lot better than other current approaches on the market. Rakesh and Jason confirmed that some of the Oracle Calendar guys (who came from Steltor) are actively involved in the CalDAV standardization efforts, so they recognize that cross-organizational calendaring is a nut that still has to be cracked. This is all good.

Conclusion: What Oracle has achieved here rocks big time. This is the best implementation of calendaring for enterprise users that I am aware of to date. The Steltor technology is obviously playing a big and important role in making this happen. With respect to internal calendaring, and this being an area that neither Microsoft nor IBM Lotus deal with properly in an age of workspace proliferation, I'd argue strongly that Oracle should be a strong contender in any email migration projects. For customers that are currently on Exchange 5.5 and who are considering Exchange 2003 and SharePoint, you have to do due diligence on OCS 10g.

Q4. Addressing Pillar 6, for Consolidation of Assigned Tasks
Pillar 6 (Enterprise Action Management) outlines the need to ensure that people know what they have been assigned to do, regardless of the system in which the task assignment has been made. For Oracle, tasks assigned in an OCS Workspace will display in that workspace, and all tasks for an individual across a collection of Workspaces will be rolled-up on the individual's portal home page to provide a summarized view. Users can also send tasks by email so that they can be displayed in a task list that can then synchronize with a mobile device.

Conclusion: More work is needed here, and this comes back to the need for a rich client. Tasks should automatically roll-up into the user's task application of choice, so that they can be synchronized directly with a mobile device.

Q5. Addressing Pillar 7, for Collaboration Auto-Discovery
I was so thrilled about calendaring in OCS 10g that I didn't delve into Pillar 7 (Broaden the Network through Automatic Discovery Services) too deeply. Within Workspaces, templates can be set up that that automatically bring in previous reference documents to help a new team learn from previous experiences. When I asked about real-time analysis of topics and customers, and automatic searches across the Suite to find relevant material on a moment-by-moment basis, Rakesh and Jason indicated that these matters were on the roadmap.

Conclusion: There is a minimal degree of Pillar 7 ideas embedded in OCS 10g today, and further enhancements and developments are on the roadmap. Oracle is no different from the current major players on the market in this respect.

Next Action
For Oracle, I see three high priority next actions:

1. Scalability and Reliability Concerns. In a comment on my original post, Richi Jennings (ex-HP OpenMail, ex-Samsung Contact) referenced the "reputation" OCS has for poor reliability and scalability, due in main to its reliance on the Oracle relational database. That surprised me. If there is any truth in this, then Oracle needs to work very quickly at a technical level to resolve the issues and get to an architecture that works. If there is no truth in this, then Oracle needs to point to reference customers that have achieved above-average reliability and scalability using OCS.


2. Integrated Rich Client for Windows. For the reasons outlined above, I think Oracle needs to focus serious development efforts on building a rich client for the Windows platform. In my view, this is an essential part of a strategy focused on breaking into the corporate market, and is needed to unify Workspaces data when users are offline.


3. Go-To-Market Strategy Briefing. The technology of OCS is impressive. However, there are certain things that need to be in place at a go-to-market level in order for this Suite to be successful. I'm looking forward to a subsequent discussion with Rakesh and Jason on this topic.

For me, in addition to the go-to-market strategy briefing, it would be most appropriate to have a hands-on trial of the Suite to ensure that the marketing materials align with the technical realities. How do we make it happen?

Comments

[Christopher Byrne]

It is interesting that you mention the Steltor acquisition. I know of one major university who went into a panic over the Oracle acquisition and went off to look for a totally different calendaring system. Their fear was that Oracle would kill the product as they have been known to do.

[Michael Sampson]

Chris,
Well, in this case, the University made the wrong decision. But there is always much uncertainty when new major moves are announced. As I noted above, it has enabled Oracle to do something that the current majors don't.
M.

[Richi Jennings]

Assuming I'm thinking of the same university as Christopher is, the Oracle FUD was a small part of their decision motivation.

The bigger parts were that they needed to migrate from the by-then-dead HP OpenMail and that Microsoft had made them an offer they couldn't refuse -- the sort that BillG's so good at.

Speaking of HP, rest in peace, Lew Platt.