SAN FRANCISCO—John Wookey, senior vice president of application development at Oracle, is walking a fine line. Hes charged with building out Oracles next iterations of its E-Business, PeopleSoft Enterprise and JD Edwards EnterpriseOne suites, due next year. At the same time, hes leading the development of Oracles next-generation suite, Fusion Applications, which brings together the best functionality from all three suites—in an entirely new product.
But the hard part, it seems, may not be in the development work itself, but in convincing customers that Fusion Applications—along with Fusion Middleware and the Fusion Architecture—will actually be a seamless upgrade path, versus the major reimplementation that some analysts point to. Wookey met with eWEEK Senior Writer Renee Boucher Ferguson at Oracles Fusion event here last week to talk about the companys 2007 development plans, and how new functionality there—think reports and analytics—will ease the transition into Fusion Applications, due the following year.
Charles Phillips [Oracles co-president] said during his speech that Were not migrating code, this is a new product. But then when you talked about having Fusion-esque capabilities in the next suites coming up [in 2007], how do you reconcile these points?
The difference is that when you think about what applications do, they do things like reporting. They take stuff out and do a report. We think we can actually do a lot of that reporting work sooner than changing all the transactional systems. So we will deliver all the Fusion Reports on top of an E-Business Suite, or PeopleSoft Enterprise or JD Edwards application set, and people can start using them. Theyll still use the other ones if they want, but theyll use the new stuff, and, oh by the way, when theyve done that theyll kind of understand how were porting over to the new world of Fusion.
But we havent changed the underlying transactional systems. When we build the next set of transactional systems, which will feed the data to all those reports and do all that other stuff, were writing new code in this tool set. Thats what were writing new code in. We cant reuse any of the PeopleSoft code. I cant reuse any of the Oracle Forms stuff from the E-Business Suite. Im actually building a new set of applications that focus on supporting those transactional systems in that next generation.
Reports is a separate thing, right, so we can start placing those components. And reports sounds like no big deal, right? But if you talk to any CIO, the bane of their existence is reporting, so think of 3,000 reports, I dont know what half of them do, and at some point you start thinking of an upgrade process moving forward. They kind of think, I know how to move the transactional systems forward, I just look at how the business goals have changed, and I know a couple new capabilities we have, and Ill just get people to move to the new one. I got 3,000 reports—what am I going to do with those things? Ive got to figure that out. So, delivering that early gives them a chance to do that. It is still new code by the way.
Can that new code be migrated when people decide to move to Fusion?
Yes, so then they basically have the Fusion reports, and as they move forward those Fusion reports—obviously we upgrade the data model changes—but those standard reports will be upgraded as part of the standard upgrade. So now youre going from SQR report that works totally differently to a Fusion report with a different technology report, and some data model changes.
Youre basically making a big step, so you can say, Ive got the reporting architecture technology right now, and by the way it turns out its like 10 times greater than the last reporting technology tool, for a bunch of reasons. But now I have a very kind of small baby step to take, just based on some data model changes and report layout changes, that I just move forward very quickly. The users basically arent going to see any big differences moving forward.
Are you going to put a bunch of Fusion-esque things in the next E-Business, PeopleSoft and JD Edwards suites? Or is it just the few youve outlined today?
There are a few key things: the reports, the Web registry—identifying all the Web services were going to have. First of all, we started publishing Web services more aggressively in all the current product lines to ease integration capabilities. So if people did use Fusion Middleware they could actually start building component applications … using services that weve exposed from our existing applications. Were basically going to publish all the Web services were going to have as Fusion, so people can start designing their business processes, which they then will be able to move forward to Fusion applications and transactional systems when theyre delivered.
The other thing that we started talking … is something called Fusion Analytics. So we started a project last year called BI Menuing, because one thing we found is the typical way people get access to their systems is basically they have a hierarchical menus tree and they say, OK, Im going to go do this now, and Im going to go do that now. And one of the things we realized is we built a set of tools called DBI, on the Oracle E-Business side starting a few years ago, and they became primary dashboards for a lot of our customers.
Really what were trying to do is blend those concepts. What we want to deliver is a business intelligent menu approach that says, OK, you guys, heres your major dashboard, heres whats happening in your part of the world, and heres the major things you ought to start going and doing something about. Thats the way you actually get into doing your work on a day-to-day basis. Its kind of like the in box concept, but spread across the whole domain of the business—this thing called BI menuing.
PeopleSoft had some really cool technology to help you do that called EPM, which allowed you to be much more open, so if you had data sources that arent in the applications running, you can actually bring that in. It turns out Siebel has some really strong analytics as a part of that.
But what we want to do is build essentially the way people are going to get access and get into Fusion applications, on top of the current versions of the products that theyre running. So this BI Menu, consistent with Fusion Analytics, would be the way they actually got in to go do their work in a PeopleSoft Enterprise application, or the E-Business Suite. But now the idea is youve changed the way people think about accessing their transaction systems. Its a much more intelligent, rich way to [think about] this is what I should do at my job.
Youve changed the reporting outputs to allow them to define their business process, so now when they move to Fusion, what are they changing? Their underlying transactional systems. So yes, there will still be some changes for them, but youve changed the way people think about getting in to do their work, but the way you get information out of the system doesnt change, and the way you build your business process doesnt change.
So youve made it a much more graceful on-ramp to moving to Fusion applications. By the time you get there, the hope is that as PeopleSoft customers move, they kind of go, Not that much change, because Ive got all this cool stuff the last couple of years, and Im starting to use it now.
But youre talking about actually now changing the process of the way a person gets into their applications?
The process doesnt change, but what will change is understanding where I should start instead. And what we are trying to do is give people information like our notification systems. Youve got an employee expense report, youve got a new hire process, but youll get really rich information about what you care about, and say thats the one I want to go to. Thats the idea. Its a pretty simple concept, but theres a lot of technology that we have now that makes it a lot easier for us to deliver.
When do you plan on delivering the analytics capability?
2007. We talk about 2007 applications, and thats a key one. Our hope is well get it done this year, but were being a little conservative.
Did you determine a user interface for Fusion?
Basically in [E-Business Suite] 12 were going to go much more to a PeopleSoft look and feel because the user experience work that theyve done. They did a lot of work at PeopleSoft in terms of getting very good feedback [determining] what is easier on the eyes—people work very long hours, and what is easier on the eyes in terms of color scheme, font type, the general flow. So we actually learned a lot. When we did 12 we brought all those pieces in.
Then we have a team thats really been driving out UI prototypes from two standpoints: Whats really productive for people to work in; and what is going to be easier for people to transition from, whether theyre Oracle or PeopleSoft. So that changed something were factoring in, trying to make sure that the new releases are as productive as possible, but also that they are very similar to what they had before.
There is something called skinning, a technology called JavaServer Faces that basically segregates exactly how the thing looks, and the business logic underneath. So if you want to design something thats exactly like the way PeopleSoft or JD Edwards used to work, you can do that.
So one of our obvious deliverables is when we deliver Fusion applications its like, Id like it to look like the E-Business Suite, or Id like it to look like JD Edwards, or Id like the look of PeopleSoft. It wont change every behavior, but basically that first impression is, This doesnt look that much different. And, by the way, there will be another one that says, We think this looks better, so if you push this button you tell us if you actually like it better.
Any thoughts on whats going to happen with Nexus, Siebels composite application platform, given its similarity to Fusion Middleware?
If you look at the objectives for Nexus and Fusion, they are remarkably similar. You need to move to much more component-based architecture that allows people to extend. Its exactly the same thing. What were really looking at is, How do we bring some of the Nexus developers on-board and try to accelerate overall the project? But they were in kind of similar points in some ways, a little further ahead in others, and a little behind in others.
I know youre not talking about merging code, but will you take some of the development work theyve done over the past three or four years and bring it into Fusion?
Absolutely. A good example is one of the things we think is very important in the future is that, if you think about skinning the application, and then theres capabilities like where were focused called a DHTML interface, which is kind of like Java scripting but is an AJAX principal, and thats a very common design pattern that weve been moving to. Microsoft has been building much more of whole rendering kit around that in Nexus, and we think that is a very attractive option that we want to have in the architecture.
So thats something that were hoping actually that there are just some nice components that well reuse those. Theyre still in Java class libraries, so theoretically some of the stuff is still usable. But its really the people who designed that, really what they learned in going through this process thats really what we inherit as part of this.
You talk about on-demand applications as being a big part of your business in 2010. In terms of how youre building your applications today for the future, how is that going to move you forward with on-demand?
The biggest area we focus on in terms of on-demand is this idea of superior ownership experience. We think a lot about, How do you make a more productive on-demand business? How you make a more productive on-site deployment is you have to really focus on automation—things like the configuration setup, definition process, how you orchestrate a business process and have the applications inherit that behavior. Thats going to have a huge impact, and make our customers much more productive in deploying applications. And its going to make our on-demand much more successful as well, because [users] can really fine-tune the application, even when were running it on our premises, to an individual customer requirement.
So its not running a separate iteration of an application for on-demand specifically?
Yes. One thing that Siebel has done is they have a separate code line around their on-demand business and they iterate very fast. That is one thing were actually looking at. Were going to keep that code line around CRM [customer relationship management]. Were going to continue to support it. The same thing with Contact On Demand, which is a business they started last year. Were going to continue to iterate those code lines. Well adopt the Fusion architecture, but we actually think theres a reason why they did that and why they wanted to go more quickly.
Check out eWEEK.coms for the latest news, reviews and analysis about productivity and business solutions.