eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
With the goal of finding common ground, assuaging fears and moving forward after the PeopleSoft acquisition, this weeks Oracle Application User Group annual conference in Grapevine, Texas brought together for the first time Oracle, PeopleSoft and J.D. Edwards users.
The conference also brought some small revelations about what users can expect from Oracle in the future.
John Wookey, Oracles senior vice president of application development (appointed to the top spot following the controversial acquisition of PeopleSoft Inc. last December) spoke with eWEEK.com, detailing the companys philosophy on SOA (service-oriented architecture) and application development.
That philosophy, backed by concrete development plans, will see early iterations in the next major releases of Oracles software—Oracle E-Business 12, Enterprise 9 and EnterpriseOne 8.12—and come to fruition in Fusion, Oracles next-generation application suite and underlying technology stack.
The big message to the current customers is that the responsibility for us is to try to simultaneously invest … wait, let me back up. If you look at this industry over the past 20 years, theres been evolution in architecture and in technology—file system to database, client server evolved, the Internet became infrastructure. Each was needed because of evolution of technology, or new ways to solve business problems we wanted to address.
We realized there was a category of problems, not with our core applications, but with how applications fit into the broader strategy for business.
We started realizing there are ways to handle issues that our current applications didnt address, through a service-oriented architecture. Not the age-old illusion of plug and play, but ways to think about architecting applications so there is more flexibility in terms of how [users] integrate and extend applications—and to do it in a much more cost-effective, low risk way.
We started working on this several years ago … we needed to have a way that would allow an end customer to take advantage of what we delivered as standard, but who also needed to be able to adapt to their business. With 11iX we are embracing this architecture.
Where does PeopleSoft fit in?
With PeopleTools X, they were working on this [SOA path]—it was exactly the same project, but we didnt have a lot of robust lifecycle management tools that effectively manage applications and move forward with a lot of very slick declarative tools. But with PeopleTools, because they were proprietary, [PeopleSoft] was starting to determine how they were going to move to a Java-based, standards-based platform.
The term Project Fusion actually came from the PeopleSoft team, working with Oracle. They said, You know, we have one project here, not two. We were working toward the same end—what we were going to was fusion, and the name stuck.
When we talk to people about Fusion and its evolution, the classic thing that happens is they think about the new thing, and not reality. Every customer we have is running on the E-Business Suite, PeopleSoft and J.D. Edwards.
So when we started with the next-generation [suite], we wanted to continue to move applications forward, and add critical functionality that would make it more successful.
The other thing is in thinking differently about the architecture—we had this great opportunity to look at how JDE, PeopleSoft and Oracle solved problems, and take advantage of the best of that, and leverage better than the best.
As were doing this, one of the ways you make applications more adaptive, is you leverage what are increasingly more mature BPM [business process management] tools—which is increasingly in the application server—called BPEL [Business Process Execution Language].
We realized that as a foundation design point, there is nothing that keeps us from introducing that in the current version of our software.
In PeopleSoft 9, for example, we have a much more configured business flow together with the applications. So as much as Fusion is a destination point, we also see it as an evolution by taking advantage of [process capabilities].
Next page: “Fusion” capabilities in individual suites.
Page Two
Are there other Fusion-esque capabilities that will show up first in the individual suites?
Theres a broad category. Some are in line with things that have been delivered—we think business insight, designed with adaptive business processes, is really [key]. Were really coupling a pretty significant investment with Business Intelligence, out of the box, with adaptive business process.
Another area is application service-level monitoring. [E-Business Suite] 12 takes tools that are very mature and well thought of, that are looking at whats happening in the system, going from one state of business process to another, and monitoring that, and creating a business issue. Suddenly you realize there is a problem and you have to dig around.
So is the business process management functionality being moved over from the application server layer?
No, its not moving it, its deploying middleware and using business functionality and exposing it in the engine, almost like you have a tool set and you take that and have business process flows. BPM is a little like business processes on steroids. The workflow engine is running in middleware, and there is a design in the application to take advantage of that.
What is different about this functionality from the BPM capabilities that Oracle already offers?
With our BPM tool, its not that [processes] arent communicating, but theyre communicating in a flow—step A, step B—and what this does is give you much more flexibility in terms of how you want that business flow to work.
So when youre at this point in a buying process, you say, I want to invoke a different process, and have the system reflect how that executes. Not that the system doesnt communicate today, but its difficult for an end user to say, This is not exactly how our business works, I want that to work better today.
Take billing: People often go out and buy their own billing engines, and its a lot of hard work to integrate that into a flow. A business process management tool makes that easier.
Ive talked to customers that are having difficulty now integrating business flows from one application to another in 11i10—things like customer order data to accounting. What in Fusion—or in E-Business Suite 12—is going to change to make business process integration easier?
The reason why customizations have always been such an issue is you took our billing engine, or another, and every time you [implemented] a released version, it would just sort of lay on top of that. The advantage of a business process tool is that it would identify that [implementation] and go ahead and re-invoke those [customizations], so you dont redo the engineering work.
Oracle has been really very quiet over the past four months—whats been going on? And how is the integration of the development teams going?
Weve been doing the Fusion road shows for the past six weeks, and were here [at the OAUG conference] with 25 different events going on, so a couple of things.
The integration of the teams is going extraordinarily well. We continue to make good progress with our plans—well under way with [E-Business Suite] 12, and pretty close to finalizing plans around [Enterprise] 8.9. Weve completed a major release of World that were announcing. There are a lot of things weve been dong.
Next page: Is Fusion Oracles answer to SOA?
Page Three
Is Fusion Oracles equivalent to SAPs NetWeaver?
I guess it kind of depends on how you talk about NetWeaver—which started as integration product. When SAP created mySAP ERP, they had to integrate to R/3 [SAPs older ERP suite]. Its kind of grown from there. Increasingly its looked as a new entrant into the application server space.
If you were to draw a parallel, [Oracles] Fusion middleware is our core Java engine, BPM, analytics—all the things in a broad category that people call NetWeaver. Its a talking point to create a more flexible point for applications.
Is Fusion Oracles answer to SOA?
Fusion is really about … theres part of what we think is going to be important in the next generation. Service enabling applications is important, but we dont think its [revolutionary]. We didnt invent it; SAP didnt invent it…a bunch of academics got together [and invented it].
Its really thinking about a business function as a service that you can integrate into an overall business strategy. Fusion is more than that. Its attacking core issues in how people deploy applications—and building to standards.
Its about how do you lower the cost of computing by automating the configuration process, and supporting more high availability of the system during upgrades? How do you lower cost and lower the risk to deploy?
The other thing we think is so critical is business insight as a point of design—as we design the application. Whats the real-time data that will help you do your job? We started to think about that [with Fusion].
Did the acquisition of PeopleSoft bring a fundamental shift to that thinking?
Not very much. With some pretty smart people [from PeopleSoft] added to our team, we have a lot more people on the team—and tools with great user interface principals and really good customization extensibility, so its easy to extend applications so you dont have to write code. And how to manage changes—they did a great job there. Theyve been focused in different industries too. But [fundamentally] not a lot of overall changes.
Will the upgrades to the E-Business Suite, Enterprise and EnterpriseOne be the first Fusion products, or is that still expected around 2008?
Still expected around 2008.
How will those upgrades help users migrate to a new platform?
A couple ways. One is in describing a business process, it starts giving inventory of business processes that people want to see in the future successor product. Well start looking at that. You can envision in application sets all the ways people manage buying process, or CRM. The BPM tools give us a map to that, so rather than a simple set of switches, it describes the set of business functions [around a process] more clearly. The BPM tool starts giving us a pretty explicit definition of those things.
One of things we envision people doing, as they move to Fusion, is to say, heres how I want run my business—we already have these things in place, so make sure those thing are working the same in 12, as it does in Fusion. Or when you go to Fusion, say I like the way its working in 12, I want to improve it.
With integration a key message from Oracle, will the underlying infrastructure of Fusion change from what it is now with the E-Business Suite? Will Fusion support more than Oracles database and application server?
On the application server, we cannot be agnostic; we are going to work to the Fusion middleware. The team is going to make sure the web server interoperates with [IBM] WebSphere or BEA [Systems Inc.] so theyve been working hard to make sure they can interoperate with our application server, but were going to design our apps to our server.
Databases, it hasnt been decided yet. Its Larry [Ellisons] decision. Hes spending time with customers. Thats something were talking about with customers a lot.
Is the database question coming up quite a bit in customer meetings?
They need to know for planning purposes. Weve been very clear that if youre a JDE customer, were still going to support you. For a very, very long time. Theyre very happy to hear that.
How will you give users incentive to switch, if at all?
Any customer that is running our apps, will have a free upgrade to Oracle—if youre running financials of any version today, you wont have to pay any additional [license fees]. If the technology stack changes, were going to make sure people have the incentive to not have the license be an issue.
So I envision JDE world customers will stay on that way past 2013. The way well incent people is by building a product that they want to move to, there really is no the other way to force people to move.
Next page: The data hub concept.
Page Four
SAP was in your backyard last week doing a partner recruitment event—200-something companies attended. But it doesnt seem like Oracle is doing the equivalent? Will you aggressively be seeking partners going forward?
We are starting to do that. The first thing weve been doing is reaching out to the JDE partner network, which is pretty large and hadnt been managed well. Weve been working with them.
One of our Fusion initiatives—and anything between now and then—is creating an ISV and SI program, leveraging what we have in the program, and leveraging that in our general application strategy.
Something has been vexing me about the data hub concept. The idea with the data hubs is to have all data around a specific area in one place, versus strewn throughout a bunch of databases. How is it different to have various hubs for different activities—customer information, or financial information—than it is to have too many databases?
The idea is that it started with a pretty simple concept. When we were building ERP and CRM, we had contention, so we built a new architecture—Trading Partner Model—that is J2EE and service enabled.
We delivered that and marketed it independently, but found a lot of customers used it to integrate [data] to use as a central repository, and second, to use as business intelligence, to get a sense of who customers are, what do they own. So the customer data hub was born out of that … at least for visualization to create a data library, so people have actually gotten a lot of benefit out of that.
[The idea] is to use these as an additive tool. One thing it does is surfacing concerns around who owns data…they start thinking about [data] in a centralized way. So people have decided as much as its another system they bring in, it gives them one point [of reference].
From that they can start evaluating, and use as a starting point to rethinking their information system, to start managing things. These are the systems we need to build long term strategy around.
Its interesting that SAP and Microsoft are on a similar timeline as Fusion, with their ESA and Project Green, respectively. How are you going to differentiate Oracle applications moving forward?
There are a few things. Were going to be standards-based, with our applications written in Java, with J2EE, and around industry standards. Second, were focusing on business insight as a design point for our applications. Not that were ignoring business process—thats important—but without business insight, you cant optimize business process. We think those things integrated are important. But we think [analytics] is a different design point going forward.
The last thing is really thinking about taking out as much cost and risk as possible. The biggest risk today—its just too hard to run these systems. Weve been doing a lot of incremental things [to fix that]. In the future, well make it easier to install and configure [applications] though a set of human interfaces.
You have to have high availability systems—a big part of systems is to run while there are upgrades. Theres a bunch of things you should be able to do always, even if youre getting maintenance.
The basic things: more embedded “knowledge” in apps.
Those we think are just a set of investments we need to make as an applications provider—what were trying to do is be very close to customers, about whats important to them.
Remember that $1 billion investment IBM made in PeopleSoft that would enable PS to embed WebSphere into its suite of apps? Did any integration capabilities—from PeopleTools maybe—ever come from that deal that will make its way into Fusion?
They really hadnt signed the contract. PeopleSoft had done a lot of research of what was out there on the market [in terms of middleware]. They looked at what Oracle had; they looked at what tools different organizations had, and started thinking IBM would be a good company to partner with. They started doing a lot of work with PeopleTools X. Thats kind of nicely leveraged: Where it would have been WebSphere now its going to be Oracle.
Editors Note: This interview with John Wookey, Oracles senior vice president of application development, orginally was published on June 6, 2005. It was inadvertently republished on Sept. 21, 2005.
Check out eWEEK.coms for the latest news, reviews and analysis about productivity and business solutions.