Michael Cowpland -- A Mind for Business: Interview with Corel CEO Dr. Michael Cowpland
by Dwight Johnson
Mar 17, 1999
In this interview, Dr. Cowpland treats us to the strategic thinking that has brought his enterprises again and again to the forefront of the hottest technology business opportunities.
Dr. Michael Cowpland is renowned for his commitment to developing cutting-edge technology and his ability to foresee trends in an industry characterized by frequent change. He is the founder, president and CEO of Corel Corporation, the second-largest vendor of personal productivity applications in the world.
Incorporated in 1985, Corel Corporation is recognized internationally as an award-winning developer and marketer of productivity applications, graphics and Internet software.
Corel Corporation was among the first large corporations to catch the vision and potential of the Linux operating system, announcing on May 11, 1998, its plans to develop a suite of business applications for Linux as well as to deploy Linux on the innovative NetWinder series of computers manufactured by its subsidiary Corel Computers Corp.
At the Atlanta Linux Showcase in October, 1998, Dr. Cowpland excited the Linux community by announcing that the first product in its Linux office suite, WordPerfect 8, would be free for personal use, stating "We believe the development of software for emerging operating systems such as Linux will serve to create a fair playing field for all software developers."
On December 17, 1998, Corel made good its promise and released WordPerfect 8 for Linux for free download over the Internet. It quickly became the top download at download.com and to date a sensational 600,000 copies of WordPerfect 8 for Linux have been downloaded from the Corel and other ftp sites.
I first met Dr. Cowpland at LinuxWorld Expo in San Jose, California, March 2, 1999 at the Corel exhibit booth and we walked briskly off the exhibit floor flanked by three of his staff and into a meeting room in the hotel adjacent to the San Jose Convention Center. The four of us seated ourselves around the table. The energy level in the room was electric but everyone was completely relaxed. Michael is a tall man in his mid fifties and very fit. He spoke spontaneously and at length about every question I put to him. At no time did he have to pause to think. For me, it was an exhilerating experience.
Dwight: How did you first become interested in Linux?
Michael: It was early last year. We were looking for an operating system for the NetWinder and we considered QNX and a few of the embedded systems and then one of our developers mentioned Linux. We looked into it and the more we looked into it the better we liked it. We looked inside the source code and couldn't believe how clean and well written it was. So that was the initiation and we adopted it for the NetWinder and it worked out super. Then we went to the user group meeting in Ottawa where we launched the NetWinder and we had very strong feedback. So at that point we had seen enough that we could say this was going to be an operating system of the future and we will be porting all of our applications over to Linux.
It would have been the end of '97 when we first got into it and then early in '98 was when we committed to porting all of our applications and we ported WordPerfect 8 and that came out recently and we had the 600,000 downloads and then since then we've found that to get the next applications the best way was to greatly increase the effort on Wine as opposed to doing them one at a time. That way we get 90% of each app done automatically and then the compiling enables us to tune each app a certain amount but the bulk of the work is already done.
Dwight: So you were already planning the NetWinder? Didn't you announce that about last May and when you did announce it you had already made the decision for Linux?
Michael: Yes, we had to have the operating system up a long time before that -- probably late '97.
Dwight: But that was right from the beginning planned as a thin-client solution?
Michael: Yes, it was basically a thin-client as well as a network computer and then it became a Linux computer.
Dwight: Didn't you think of it as Java at first?
Michael: Initially we thought it was Java running on an OS and the OS was going to be somewhat irrelevant. And then we suddenly found that Linux was more important than Java. So it was a transformation.
Actually, it was funny. But it dawned on us. We thought, who needs Java anyway? It became almost like a little frill on the side. And that's the way it is now; we have a JVM. It's an important tool to have -- there's a lot of MIS middleware being written in Java. So, it's definitely got its place. But the emphasis rapidly switched from one to the other.
Dwight: How is Linux and open-source software currently being used in-house at Corel and what are Corel's plans for further deployment?
Michael: We're running plenty of our servers on Linux.
Dwight: Your Web servers?
Michael: Yes, and I've got two computers on my desk. I've got one Linux and one Windows, which is good -- first-hand networking back and forth. And I think we'll be rolling it out on a very rapidly increasing basis from here on out in a dual platform mode where a lot of people will have a dual boot or just two computers.
Dwight: Do you think it will become the primary OS for a significant percentage of your employees?
Michael: I think it will over a period of time. It'll probably take 18 months. We'll have all the apps so there won't be much reason not to use it. What most people use it for is the Web and the office suite and graphics and we'll have all of those taken care of. So I'm sure we'll be maximizing its use. Initially it will be parallel and then I think it will become the preferred system.
One of the benefits of me using it first-hand is first-hand visibility of why we needed to do our own distribution because we could see that it is a very good system but getting it set up is difficult. You need an expert to set it up. We'll take care of all of that. Similarly, when you browse the network it is not as easy as Windows right now. We'll take care of that.
We can see what's there and what's missing and there's not much missing. We'll fill in the missing parts.
Dwight: How will Corel market the Linux versions of its Linux products?
Michael: A combination of typical Web techniques such as the free downloads we've offered of WordPerfect 8 which has now had 600,000. It's been a huge success. I don't think we ever thought it would be anything like that many. And conventional distribution where we're already selling existing Linux products through people like Ingram and CompUSA. It's a two-pronged approach. Some people like to have the box and the documentation and the CD-ROM. And others don't mind doing the download. And I guess a third component is corporate licensing where we have direct relationships with people who buy licenses. And that would be particularly good for new users such as the Mexican school system where they are adopting it, as I mentioned in the keynote, for all their schools.
Dwight: I knew that Mexico was adopting Linux. So Corel is going to supply some software?
Michael: We haven't actually finalized that but we are definitely going to be working hard in that direction. I think the tender is coming up in the next few months, so we are actually actively working down there right now.
Dwight: Who is competing with you down there?
Michael: At this point, it's hard to say because you're dealing with the Mexican government system. A lot of it will be who has the attention of the government in being able to do the total job. I think we will be very well positioned with our comprehensive approach to Linux.
Dwight: How, in your view, does open-source software fit in to the evolution of computer software in general? Is it the direction that all software is headed or will there always be a market for commercial, closed source software?
Michael: I think it's definitely a mixture of the two where its like a pyramid with the bottom layers being the OS and that's the common denominator, so everybody uses that. And that's what gets the most benefit from the global open-source effort because every single user has a vested interest in it being really good and keeping it open. Once you get into specific applications, you then fragment that effort. So that's why applications will still stay mainly commercial except for the lightweight ones. Because groups of open-source people can be in pockets of a certain quantity which for the operating system becomes an aggregation that's pretty impressive, but if you take any individual application, that's a much smaller population that has the incentive to do it. It's very expensive to develop a fully-fledged application like a WordPerfect or a Quattro Pro. And you don't find that just happening easily. And, in fact, I think a rule of thumb is that if you open-source a project the net result is you get about 20 or 30% actual extra developers free. It's not like 200 times more. I've talked to people like Michael Tiemann of Cygnus who's been in open-source since the beginning and asked him point blank -- because he's got some open-source stuff and some not open-source -- and I said, how do you decide which is which? Basically, its the typical self-interest of your organization. If it makes sense to go open-source, you do. If it doesn't make sense, you don't. It's not particularly doctrine, it's more like pragmatic.
But I said, okay, when you do open-source, what's the net benefit? And he said about 20 or 30%. And I have heard that before too with the Mozilla project, because I think intuitively, when you first hear about it, you think you're going to get thousands of times more effort because it's the whole world. You might have 20 people working on it but there's 2,000 out there working away. It's going to be like magic. But that's not the case. In reality, it's all fragmented down and you get a certain cooperation and effort and it's a bonus but it's not going to make the whole thing stand on its own. The bottom line is you need someone who's got an economic interest in applications to carry them forward if you need the applications to do the extra things people want. Quite often those extra things are not the kind of stuff that enthusiasts want to write. They might be drivers or they might be color separation details that might be a little boring.
It's like having the interstate highway. The highway is free but you have shopping centers too and you don't expect all the goods to be free in the shops.
A big developer like us, we have value and money that we can invest, and we invest that by buying other technology and aggregating it to the end result. And we are now doing the same thing with open-source where we are putting a lot of actual real money into the Wine development because it's good for us. And the same thing with the distribution where we're vesting in that. If we got it free, it would be fine, but in this case we need to put our own investment in there and we've got good resources to put behind it. But in other cases, in Windows, we have had to have commercial arrangements. They wouldn't even allow us to make that open-source.
The model of software is evolving anyway because of the Web where we're doing more and more activity on the Web and that's changing the model where more software will be free. Free download -- and then you want people to create traffic because traffic is worth more than applications now.
Dwight: How is the Internet changing computing and how is Corel responding to those changes?
Michael: It's totally flipping it around and we're responding as fast as we can because we recognize it's a completely new model. We're immersing ourselves in the Web totally to use it in every way possible. We're putting major investment in all areas of that. We've actually talked to three major Web outfits in the last week and the insights we've got are incredible. One of the comments we got from Netscape was they said they spent four years developing enterprise software and ended up being sold to AOL for $9 billion as a media company because of their advertising. They were actually showing us on their Web site where one little button for Citibank VISA they got $60 million for. You have to sell a lot of software to get that just for the advertising hits. So it's just flipped the whole idea around. We're thinking free and then come to the Web site for subscriptions, for ongoing useful stuff and that can be a whole new model -- Free training.
Dwight: What is the revenue side of that?
Michael: Traffic. Two avenues: one is by exposing people to our latest features, its almost like a demonstration as well as training. They will realize what they're missing and therefore upgrade to the latest version instead of being quite happy with their old version. So one revenue would be quicker upgrades because of the improved productivity that they will get. Secondly, if we get more Web traffic, we'll have effectively more advertising revenue if we advertise or we're advertising ourselves more.
Dwight: That's interesting because that's not really what I was thinking when I asked the question. What I was thinking was that Oracle has been putting forth their vision for six months now of the fat server and just the browser on the other end. If you go their direction, applications are going to radically change and I'm wondering if you're seeing anything like that happening at Corel.
Michael: Not really, no. I think we're seeing different business models. Some of the stuff Oracle comes up with doesn't come to pass -- like network computers.
Dwight: Certainly the Internet is causing more networking to evolve in the use of the PC in general and you would think that applications themselves would change.
Michael: They will be be changing but we see a different view. XML will become a huge factor because XML gives you structure between businesses in business to business commerce and transactions. And that's why XML is built into WordPerfect 9, which you can't get in Word. The Oracle model is okay if everybody is on one server but what about business to business? You've got to go through XML so you can talk to each other. And that needs applications in between. It's okay if the whole world is just one big business on one Oracle server but it would never be that simple.
Dwight: What's the bridge with Paradox?
Michael: Actually, we expect to turn Paradox into an XML repository because the next big thing is going to be databases which handle XML which act as repositories and then we've got WordPerfect as the ideal client for inputting the data and depositing it.
Dwight: How far downstream is that new Paradox?
Michael: Probably, we'll do that this year.
Dwight: Will that be an upgrade to the current version?
Michael: Actually version 9 is coming out which has got a lot of Java and JDBC stuff in it coming out in April. But then the next efforts we're going to put onto Paradox would be a refresh to really enhance the XML part. We see that as a whole new window of opportunity to differentiate. Because you're going to have Oracle in as a heavy-duty XML file retriever -- very expensive -- and if someone wants to get going on their own for the first 100 users, then Paradox will be ideal. Built right into the Suite.
Dwight: How do you think Microsoft will respond to the threat of a Corel-Linux desktop that is completely free of Microsoft products?
Michael: We've always been number one in their cross-hairs with the office suite for the last two years, so I think they are used to us being there as a competitor. But at the same time, they are realizing that they need a valid alternative because of anti-trust. So they're probably not quite as predatory as they used to be. We've now developed good co-existence where we went up 10% in users in the last 12 months which was phenomenal news to us because they have been trying to squish us down strongly in the last two or three years and it hasn't worked.
Dwight: You develop all of your current products for Windows. Could they make things difficult for you by not providing you with information you need?
Michael: I don't think so, because they have rules they must obey now under scrutiny from the Justice Department and we just licensed VB8 from them for our versions 9 because it is the best scripting tool.
Dwight: What do you think would be the best outcome of the anti-trust action for the future of the computer industry?
Michael: I think it would be best if they were forced to have fixed prices, I mean standard prices, so they couldn't say one price for Dell, one price for Gateway behind closed doors if you do this or if you don't do this. Force them out in the open so that everything is published because they are almost like a utility. I think splitting them up would be too complicated. But if you basically expose everything they are doing -- it's almost got to be like the phone company that publishes rates. Not so they couldn't change, but at least they've got to publish them. Then there wouldn't be so many games going on all the time of behind closed door negotiations and nothing written down.
Dwight: From some of the developers, I've heard some complaints that unless you play ball with Microsoft just right they might withhold information you need. Nothing like that has happened?
Michael: No, it hasn't happened but people are always worried about it. I think Compaq said that they were worried that they wouldn't get updates if they didn't play ball.
Dwight: But don't we have the entire Software Publishers Association lined up against Microsoft on some level?
Michael: That's right.
Dwight: What's that all about?
Michael: It's about the dirty tricks. People don't like the dirty tricks.
Dwight: But mostly marketing tricks?
Dwight: Not on the technical side?
Michael: No. On the technical side, I think they stopped doing that a few years back.
Dwight: There was no issue that Microsoft Word would be able to run better on Windows than WordPerfect because you weren't getting information?
Michael: No. We know that for a fact because with our JBridge technology we had to X-ray the OLE APIs and I actually asked our chief engineer doing that what about all these undocumented APIs, the ones they're supposed to use and there truly are some undocumented APIs but it's ones you wouldn't want to use anyway. They are just for internal communication. I had him print them all out and we could see they weren't sinister ones. They were just there because they needed internal communication and there wouldn't be any point in exposing them because you wouldn't want to use them anyway.
On the other hand, going back a little bit further, when they were fighting DR-DOS, it was known that there were actually bugs introduced where they actually put a bug in so that if it detected DR-DOS it would fail.
Dwight: So it's the marketing dirty tricks -- special placement on the desktop, etc?
Dwight: There has been some confusion about whether Corel is going to develop a new GUI for Linux or an entire distribution.
Michael: It's a bit of both really because we are going to edit the distribution to what's ideal for a mass user.
Dwight: And you're going to package it and sell it?
Michael: Yes. One stop shopping. Because we want to go to the world's biggest computer manufacturers and say, here's Linux ready to roll. It's got a beautiful GUI. You turn it on -- it's easier than Windows even and it's got office apps and it's got graphics apps, it networks beautifully and you can continue to work with all your friends who've got Windows files.
Dwight: So we've got Red Hat, Caldera, SuSE, Pacific HiTech and Debian -- and now we're going to have Corel Linux?
Michael: Yes, and each one has its own focus.
Dwight: When will we see a release?
Michael: This year. We're looking at about six to eight months.
Dwight: And besides the GUI itself, which you have already stated is one of Corel's main focuses for development, what else are you going to improve?
Michael: A couple of things. Browsing the network so that you can easily browse for all the system files on the network. Right now you have to type in parts and be a bit of a hacker. It will be just like you do in Windows, where you can just find what you are looking for by browsing with a GUI.
Dwight: Doesn't KDE go in that direction now?
Michael: In that direction. There's not that much missing, but there's 5% missing that will make it easy instead of difficult -- the mounting of CD-ROMs, the mounting of modems, maybe some print previewing technology which in our latest office suite is way ahead of anything Microsoft is doing. And that's a very good easy to use feature. We could even put that right in the GUI. It's the kind of stuff you can do for all applications and not just our own because printing is such a troublesome thing normally, getting it right. Font management. Net Meeting -- we've got a voice over IP technology. Voice and video over IP, so we could replicate that function that nobody else has done yet. And, in general, we're coming from a space where we're experts with the GUI with our design for illustration programs, office suite applications and set up and install of huge programs. So we bring a lot of expertise to bear on this which the other companies don't really have.
Dwight: Is this going to be a proprietary GUI?
Michael: No. Open-source.
Dwight: Are you going to use some of the GUIs that have been developed?
Dwight: You're going to start with KDE?
Michael: Yes, and we're talking to Troll Tech about their libraries and they've got a browser technology that we will probably include. We've been very good at always working with what's out there as opposed to reinventing wheels.
Dwight: Are you going to supplant Netscape as the browser?
Michael: We'll make the appropriate choice, because it's all open-source and the skill is simply selecting what's best for the user and that's what it's all about in terms of coming up with popular, easy to use programs. There's lots of choices you have to make and you have to have the people who've shown themselves good at doing that. Red Hat has three full CDs of OS. Maybe that's too much for most people. Maybe they only need one CD -- but the right one.
Setting up your Internet connections and setting up your mail and setting up your home network -- all of that kind of stuff are areas where people would want ease of use.
Dwight: You seem very sure that you're going to get all of this stuff done this year.
Michael: Yes, we're good at hitting milestones. Our office suite is coming out on time; Microsoft is not -- they are delaying again. And Sun has a history of delay after delay after delay with Java. We have a history of actually being on time.
Dwight: The application I'm looking for is Ventura Publisher.
Michael: Ventura's been running five-star reviews right across the world. So we're very keen on Ventura. XML will become a another big boost for Ventura. It's already got very strong HTML, XML and we're looking at another special edition as well, really highlighting XML DTDs now that that's becoming so hot. The Gartner Group said, in another three years, XML will be as hot as the Web itself. Which is pretty encouraging because up to now it's been a bit of an enthusiast's market. But I guess the business to business commerce market is going to drive that big time.
Dwight: Will Ventura be one of those applications that is a Windows application emulated under Wine?
Michael: No, we'll compile it too.
Dwight: It wasn't mentioned as one of the applications that would run under the Wine APIs?
Michael: No, I didn't mention it because it is version 8 and everything else is version 9. It won't be this year but early next year.
Dwight: Thanks very much for talking to me.
Michael: Thanks, Dwight. Thanks for your time.
Copyright ©1999 by Linux Today