Inprise Chief Talks about Borland's Future

1999

Dale Fuller

Dale Fuller is interim CEO and President of Inprise Corporation. There's still a sense of transition at Inprise, where the name has changed from Borland International, to Inprise, and the resurrection of the Borland name under the guise of the Borland.com Web site. The Borland name is now back in more prominent use, from the flags in front of the corporate campus to the Borland/Inprise logo on their Web site.

Before coming to Inprise, Fuller was President of WhoWhere, an Internet directory service acquired by Lycos, and was a Vice President of Apple Computer, where he was in charge of their PowerBook division. Dale spoke with Matt Carter last week at their headquarters in Scotts Valley, California.

Dale Fuller: I've only been here for four months...

DF: We had lost our way in the past few years and lost sight of our roots: providing the necessary solution sets for people to get their jobs done fast. Our focus is on the developer community. Whether you are a Fortune One or Two company with a gigantic IT department, to your very small independent consultant who is doing something for an small insurance agency, we build that scale of products from the low end to the very high end. We provide that kind of scalability to developers, you can get started with one product and our products can scale as you scale.

And in the space of the Internet, if you can't scale, it's the kiss of death. I would say that most applications in the world today cannot scale.

DF: Being in the Internet space, I started WhoWhere, the ninth-largest Web site at the time. There, we learned what scaling was all about, when we were doing 100 Million page views in an hour and designed the site for a million page views a month. It just grew like a bat out of hell.

We could not keep up with it, and we had to abandon the traditional enterprise design methodology of design, develop test, deploy. We were running in real time, and things were breaking in real time. We had to fix them in real time and grow them in real time. In our first rev of products we designed for 10,000 users and when it hit 100,000 people in three months we were panicking. We had to design in real time and watch it grow.

And you can't design for that speed of growth. Look at EBay: they found that lesson last month and will find that out next week when they go down again. The wild thing is that they don't know when they will go down again. In the past we could predict how many people would use our products: one person or 10 people or 1,000 people. But now on the Web, suddenly you can be running millions of sessions concurrently. The bad news is: now that you may be generating the majority of your revenue from the Web, you better make sure everything is running in real time, otherwise you are out of business. That's what I mean by scaling.

DF: I see it as time moves on, the things we've developed in the past is the glue that holds all these things together. For example, we have a very large enterprise system here. We use Oracle and our ORBs (Object Request Brokers) and we needed to have an interface for sales people to enter in orders remotely. So we looked around and bought a product called Clarify.

Clarify fit right, all the specs were right, it was going to work very well. Problem: after we bought the tool and got it going, we saw that as it scaled up, the bandwidth requirements per session were about 3 megabits. So if you have a sales guy out in the field with a 56k modem, it was now taking him 30 minutes to enter in an order that should take him two minutes. But, because of the strength of Clarify, we wanted to still use it.

So we had a guy in our IT department look into it. He quickly created an HTML page that our guys would dump their information into and it went straight into a database. He then used Delphi to take the data from the database and dump it into Clarify. This took two days, versus having Clarify rewrite their stuff for optimum performance and scalability that would take lots of millions of dollars and lots of years.

We don't have time to actually do the design develop, test, integrate, deploy as before, because we have to move at Internet speed. That's where our products fit in: you use them to glue together applications very quickly. And with our products all tying together, it makes it a lot better. You use our ORB which interfaces with Oracle 8i, with Sun servers, with Visibroker. And a lot of our products are hidden in the system: we OEM our technology to other companies, for example, our JBuilder product is built into Oracle 8i. We have a lot of experience of making these things work and solving problems.

DF: I think Java is a great development tool. If you look at the scalability and performance of Java, because of the speed that you can develop with Java versus C or C++, you can get to market a lot faster with your products. And now performance of computers is becoming a moot point for programmers. If you remember back to the late 70s and early 80s, we were really restricted to 64K. But then memory became cheap, and nobody was concerned with spaghetti code. Now, we're running into this problem called the Internet. The Internet has bandwidth problems and problems with scale, meaning that now my Web site can be bombarded with millions of people at the same time, and I need every one. So I need to be able to manage and handle that.

DF: (Yelling) "They've gone to the dark side!"

DF: It's enhanced it. You look at the entire market, and let's face it, Microsoft has a gigantic piece of the market, they've become a standard. So our customers are telling us that "we need a way that we can get from this Microsoft stuff to this back end stuff." Microsoft doesn't want to do it, the back end guys don't want to do it, so ours is a total response to our customer base. How do we bridge this gap?

We're focused on the standards like CORBA, but we're working closely with Microsoft to add COM interfaces to CORBA. Microsoft knows there's this other side out here, which they consider the dark side, that needs to connect to them. The major thing is, everyone needs to coexist, and we have become that company that allows everyone to coexist. We have become Switzerland: we have Oracle standardizing on our stuff, we have Sun standardizing on our stuff, and we have Microsoft using our stuff and investing in us. What I hear now is: "wow Dale, how did you manage to get the arch enemies to work with you together?"

DF: No actually it's been a natural progression. Before I got here, the company was always fighting Microsoft. Now, Microsoft is a standard, so you support the standard or you don't, that's your choice. And we chose to support that: there's a market big enough out there and our customers are telling us that we should support it. We're not saying "we're throwing in the towel and going down that path (of supporting Microsoft)": our customers are asking for it. And there are people who say, "you've joined the dark side": I think our Delphi developers saw at our recent developers conference that we have not compromised our products at all at any level.

Along those lines, if you go on our Web site today, you will see that we have a gigantic Linux effort underway. We actually have most of our products on Linux today: all our app servers work on Linux, Visibroker works on Linux. You've seen us as shows demonstrating our JBuilder products on Linux, and you will shortly see some really cool stuff on Linux from us.

DF: Well (hesitating), we listen very closely to our customers… and we respond.

DF: Business to Business will be XML, and it will be gigantic. For me to do business, I have a community of people: my own employees creating the stuff, my customers buying the stuff, and my partners supplying the stuff that I sell. The standards of CORBA, of COM, and the Web side are really to help build my market and to be an interface to my customers. Leveraging the Web to my partners is really going to be reliant on XML data structures on the back end, and it going to be humongous.

What (technologies such as) Linux and XML do for me, instead of reengineering leverage everything I already have. Some of that I have to throw away, but there's lots of stuff I want to keep. We've already had the experience in the last 10 years of reengineering the enterprise, and that's been a tremendous success, and extremely pleasurable (laughing). So now we're able to leverage that and bring to fruition what we started working on years ago with the Web. That's what JBuilder and Delphi leverage, developing applications quickly and interfacing with my customers back ends.

DF: I think it's a commitment to integrating it into all tools. If you ask 20 people in the industry "What is XML" you'll get 20 different answers, all the way from "it's the replacement to HTML" to "Aw, it's just a stupid data structure". I think the answer is somewhere in between, but the issue is, that you have people who need to develop solutions very quickly. Those solutions will comprise data structures exchanging, and what happens with XML is that data structures can be common, defined, and quick to transfer. And IT shops can commit to XML and boom: go develop.

DF: Well, mission number one was to stop the (financial) bleeding of the company, and to really identify where the strengths (of Inprise) are, and how to get people focused on them. When I got here, the company was burning money, and not on smart things. We haven't had a layoff since I've been here, but we've gotten our expenses down.

Our cash reserves, which were nominal when I arrived, are back up to almost a couple of hundred million dollars, we own this building outright adding hundreds of millions to that, so we're on sound financial ground here. Now I'm focused on what my MIT professor would call "linear placement of aquatic assets", or putting our ducks in a row. Now, when we go out and communicate with people that they understand why you've done what you've done, why this thing is moving forward, and starting to say how does it work. Our big picture goal is moving companies from brick and mortar to click and mortar, because that's where revenue and that's where business is moving fast as possible.

The other important strategy I've learned coming from the Internet space is that you can't do it all yourself. You have to partner. It's all about speed: if you wait for your company to develop it all yourself, or you have the "not invented here" mentality, you're dead. This game is all about partnering. There are things that we are very good at, and we will stay very good at and improve. There are things we are not very good at, and we are going to partner with companies who are very good at those things, and we are going to be very good partners with them. So we won't own the whole business but we will be able to provide the entire solution set so our customer actually knows that he's taken care of in an end-to-end solution.

DF: In five years? "Wow, what a turnaround. The company went from $90 to $2.98 and that, what's his name, Fuller, turned it around and now it's back to $90: that's incredible! They have cool stuff, they are doing cool things." We want to get back to that, and not forget where we came from. Over the last few years, we did forget where we came from, and what our strengths were. Our strengths were working with people. It's all about community, and helping people communicate and collaborate.

DF: There's things we need to add, and things we have that have outlived their usefulness, but we're moving in the right direction. Is it moving as fast as I want it to? It's not moving as fast as a privately held company would, but we're public so we have additional responsibilities.

But that's good news for our developers because we have staying power and we'll be here tomorrow. It's been four and a half months since I've been here and we're making major strides and have some pretty big successes. We're excited.

I'm the interim CEO: I'm only here temporarily, as anyone is. I can be fired tomorrow. I think what's happened in the past is that CEOs here have thought they were rock stars and indispensable. I don't want to forget that.