David Nagel Unplugged: Can Palm re-connect?

If anyone has the pedigree to guarantee that the Palm operating system remains a contender in the market for personal information management, converged phone/PDAs, wireless terminals, and other embedded mobile systems, it's PalmSource CEO David Nagel.

Prior to joining Palm as head of its platforms division (now PalmSource), Nagel was a CTO at AT&T and oversaw a huge chunk of that company's R&D efforts. He knows networking. Before that, he ran R&D for the Macintosh OS at Apple. He knows usability. But is executive pedigree enough to reaffirm PalmOS' place in a market that it was largely responsible for establishing?

PalmOS faces some determined competitors. Although late to the game, Microsoft appears determined to dominate the handheld space and has the marketing war chest to make up for any technical shortcomings. Canada-based Research in Motion upped the ante with a wireless package that redefined real mobility to be communications-centric--first with e-mail communications and then with voice.

Symbian, once the power behind the Psion handhelds, has regrouped as the pre-eminent operating system of choice behind most cell phones without a proprietary OS. In the great convergence, these phones take on PDA functionality as PDAs take on phone functionality.

Finally, Sun seeks to ensure that, regardless of what the kernel on any handheld looks like, the development platform of choice for deploying applications across all of them is Java. For Nagel, the heat is on.

During our interview, Nagel left no mobile or handheld stone unturned. He admitted that the perfect converged device doesn't yet exist and batted around a few ideas on what it will look like. But Nagel made one goal of PalmSource's absolutely clear. With the mobile phone market orders of magnitude bigger than the PDA market, PalmSource has to win there and win big. To that end, PalmSource will have to enable many more licensees in the global mobile phone industry beyond Samsung, Kyocera, and Palm (by virtue of the Handspring acquisition) to bring the sorts of the devices to market that will get the wireless carriers excited enough to add them to their lineups.

Now that Palm is just a another licensee of PalmSource, and PalmSource is free to go its own way, can Nagel take it to a whole new level? Or will the operating system provider suffer at the hands of a crowded market with firmly entrenched competition and some of the levers out of its control?

		'At the bare minimum, out-of-the-box, you should be able to read any document someone sends to you.' --David Nagel

ZDNet: I saw that phone. In the clamshell category, it looks like a Kyocera-killer.

Nagel: I prefer to think of it as the PocketPC Phone Edition-killer. I prefer to kill the other platform than my own licensees.

ZDNet: Actually, I don't think the software in the Phone Edition is particularly well done.

Nagel: It hasn't gotten any traction in the marketplace at all.

ZDNet: Actually, I don't think the PalmOS is much better. If you ask me, the software in the BlackBerry phone sets the bar for…

Nagel: Mail?

ZDNet: No, for integration of the personal digital assistant functionality with the phone functionality.

Nagel: Really? It must have gotten a lot better because I thought it fell down on that point. When I tried the BlackBerry, I found that I was never using the address book in it. Never. Or the calendar. I just found it awkward to use. But, I never had the phone edition. I just had the plain BlackBerry.

ZDNet: What's nice about the BlackBerry phone is the way they've integrated all of the inbound and outbound communications into one place. I want to call it a universal inbox, but that's not what it is. It's an in/outbox that tracks all your in or outbound mail, phone calls, and SMS messages and you can very easily respond to a communication from someone that came over one channel using another channel.

Nagel: That sounds like a good idea.

ZDNet: You don't have to hop around like I've had to do on the Phone Edition of PocketPC or the PalmOS-based phones. For example, replying to email with a phone call or vice versa. As long as the person is in your address book, the BlackBerry phone makes that very easy to do.

Nagel: That's actually a very good idea. That's something we could do in the software very easily. What about the new Treo 600? Have you seen that?

ZDNet: I did more than that. I reviewed some of its features on ZDNet.

Nagel: What did you think of it?

ZDNet: With the Treo 300, there were some times where you had no choice but to thumb the screen or take the stylus out and use that to get past a certain point or function. I thought the Treo 600's North-South-East-West jogger button, which moves the user between all the available fields and buttons on the screen, was a great improvement. I also think Handspring improved the way the phone keypad is embedded into the thumbboard.

Nagel: Yeah, they've done a better job on that. Before, on the Treo 300, those buttons were too obscure.

ZDNet: But now they've colorized those specific buttons.

Nagel: Right, they colorized the buttons and made them much more prominent.

ZDNet: Also, if you use any of these data-capable wireless handhelds for data operations, they can eat up a battery pretty quickly. If someone like Handspring is going to put a data-capable device in the field, guess what? Someone might actually use it for data.

Nagel: All the time.

ZDNet: I think it was a big mistake not to have a replaceable battery.

Nagel: I actually don't understand why they didn't do that. What do you think of the merger between Palm and Handspring?

ZDNet: I don't think Palm was any more or less capable of producing something like the Treo.

Nagel: They were about a year behind.

ZDNet: My sense is that, since the wireless carriers are in complete control of the distribution channel for converged devices, and because Handspring has some good relationships there, that's what Palm was really after.

Nagel: [Handspring] has had those relationships for a long time. Looking at it from a platform point of view, [PalmSource] has relationships with almost 60 carriers through our licensees. So, there's quite a bit of acceptance of the Palm platform. But, you're right. Getting those relationships with the carriers is a much bigger barrier than anyone could ever imagine.

ZDNet: That's what I'm hearing. When you talk to the device manufacturers, they'll tell you that the most difficult thing is getting their devices added to the carriers' lineups. Palm may have had a relationship with the carriers with the Tungsten W, but it wasn't being marketed as a telephone. When Palm came and told me that it really isn't to be treated as your primary phone, I thought that was ridiculous. Why should I have to carry two big wireless devices and maintain two separate wireless accounts if it all could be merged into one?

Nagel: The W was never positioned very well. People really didn't know what it was. The [802.11-eeuipped] Tungsten C is much better in that regard. It's pretty clear what it is.

ZDNet: I keep wondering whether maybe we should break these devices down so that you have small radio for WAN like a CDMA 1xRTT or GPRS radio, and one for LAN like a Wi-Fi radio and then you can have a bunch of devices like PDAs, laptops, and whatnot that use Bluetooth to share those radios at will.

Nagel: All of this is sorting out. It's going to take another cycle or two for what I call the canonical form factors to emerge. I do think there are different people who like different things. Some people really want a phone and want a little bit of paging, instant messaging, or email and some people come from the other side like the BlackBerry guys, and I think there are some form factors that haven't emerged yet.

ZDNet: One thing that's imperative to me in a converged device is the ability to collaborate on Microsoft Office documents. If I was working alone, I could probably change my desktop environment to make my mobile one more efficient. But I collaborate with others and I can't ask them to change. The lack of any built-in ability to do this with the Palm OS when compared to the Pocket PC is a sticking point for me.

Nagel: I agree with you, and that's one of the things we just announced. At the bare minimum, out-of-the-box, you should be able to read any document someone sends to you. Now, if I want to edit it and play around with it, or do something other than talk about it, then I should buy the application. So, we're going to start building in support for the native file formats so all those documents can render on the device.

ZDNet: Well, if you look at a PocketPC-based device, you can receive Word and Excel documents as email attachments. They can be opened, edited, saved locally on the device, and reattached to an outbound mail. I find this to be far more useful out of the box than just being able to view a document. So, I know it's a matter of personal taste. But, when I think of the ultimate converged device, it needs to integrate well with my software environment, it needs to have a thumbboard, and it needs to have removable batteries.

Nagel: So, in your search for the best converged device, have you found anything that comes close [to what you're looking for]?

ZDNet: I think the Hitachi G1000 comes close, but it's not perfect. It's based on PocketPC, so it can handle Word and Excel documents and it has built-in mail and VPN clients.

Nagel: Is it a candy-bar form factor, or is it more like a PDA?

ZDNet: PDA. In terms of what's already out there, it most closely resembles HP's iPaq. But it has a thumbboard. However, Hitachi neglected to embed a telephone keypad into the thumbboard. That was an oversight.

Nagel: I agree. That would have made common sense.

ZDNet: It's too big though, which is why I said earlier it might make more sense to break it down into pieces. Maybe the whole thing needs to be ripped apart. Maybe we have one [wide area wireless] radio and one bill that goes with it and then we have all these other devices that share that radio, maybe in your belt buckle or something. Maybe the sharing happens via Bluetooth. This way, when I'm doing the business thing, I can take my PDA or my laptop or even a big voice device with a big battery and share the radio for those, but then when I go to the beach, I can bring a different voice device or have a wireless device in my ear with the dialer on my waterproof wristwatch.

So, to the extent that the most successful handhelds are a marriage between good hardware and good software, how much influence do you have over the hardware? Can you walk over to Palm and say "Look fellas, we need to have a talk about your designs?"

		'I don't want Bill Gates designing my ideal converged device because I don't think he and I think exactly alike in terms of providing value, making tradeoffs.'

Nagel: Yes. I say that metaphorically. Palm and PalmSource are separate companies. I really don't have any control over what they do. Also, we have a fundamentally different design philosophy than most of the other integrated device guys. Certainly, Research in Motion [makers of the BlackBerry], but Microsoft as well with PocketPC. PocketPC has a very well defined reference design, but Microsoft does in effect get on the phone and say you have to do it exactly this way. That's why all PocketPC-based devices more or less look alike.

ZDNet: But, on the flip-side, that would make a very predictable target for developers.

Nagel: Yes, but you can do that in software too without having all the hardware look alike. If we did that, we would never have something like what Fossil just introduced, which is a Palm-powered wrist watch. (Editor's note: Fossil must be hedging. The company also announced it will be building products around Microsoft's Smart Personal Objects Technology otherwise known as SPOT). So, our strategy, and we'll see if it works or not, has been to encourage people to be innovative and differentiate. You have all these requirements and you have to figure how to do it. But you're right: Some clever industrial designer will. Our view is that some clever licensee will figure out how to make an entertainment device, while others like Handspring will work around or build on top. Those licensees are real proud of where they built on top. That's where they can add value and they hopefully will get some margin points for that. If you're competing only on price and there are no design differences, then you end up with a commodity product and you get back into the PC business. That's something we think should be avoided in this personal electronics space because everybody has a different idea of what they want. I don't want Bill Gates designing my ideal converged device because I don't think he and I think exactly alike in terms of providing value, making tradeoffs, or whatever. If you give people more freedom, then you're likely to see more offerings that addresses more diverse needs and you end up with a wider market.

ZDNet: Do you see the different needs of small business and large enterprises as one of those things that defines design?

Nagel: Yes. That's why we are standardizing at the software level on a lot of the capabilities, so that the enterprise developer can assume the stuff will be there. We're even providing ways for people to add value in the channel. The architecture is extendable. It's analogous to a browser with plug-ins. We make a standard plug-in interface so that if you don't want to use the VPN we provide, you can use somebody else's VPN.

ZDNet: Based on some the announcements you've been making, it looks like you're using partners to fill in the gaps. How are you deciding which partners to go with?

Nagel: First of all, we're trying to be a little selective. Something like mail, though, is really complicated. In addition to not knowing who's going to win in the handheld mail space, we don't do the back end stuff and don't intend to for the foreseeable future.

ZDNet: You almost did, right?

Nagel: We didn't. Palm did-- the hardware guys. But it didn't work. It's really hard for hardware companies to sell software. What happens is, the sales guys go out and sell you a bunch of Tungsten C's and someone says, "if you throw in the software, I'll buy it." Once you get into that cycle of giving software away, it becomes a nightmare because there's still cost of goods. Meanwhile, the customer is saying "What you mean? It doesn't cost you anything. It's just a box of stuff." Hardware and software are two very different businesses. That's one of the reasons we're splitting. Now, for software, we're starting to choose partners very carefully. Obviously, they have to be philosophically aligned with us. That's one of the reasons that the Visto partnership is going to work so well for us. They understand that we're going to be sort of universal client. (Editor's note: Visto is a provider of wireless handheld-based email solutions and is now working with PalmSource to build a wireless email client that device manufacturers and wireless carriers can include as their PalmOS-based device's default messaging solution. For more information, see ZDNet's Webcast interview with Visto marketing vice president Tom O'Brien). We can work with Visto on ideas such as the one you talked about, where the BlackBerry collects everything in one place. It's a very good idea and I think we can manage it. The alternative is to pick a single solution in each category and then alienate 90 percent of the customers. The other thing that we're looking at is the out-of-the-box stuff. I think one of the problems we have to solve is how to package things. Most of the IT purchase decisions made in this country are made by small businesses. There's something like 3 million small businesses and, out of those, 300,000 people account for three quarters of all IT purchases. Many of those guys don't have big IT departments. They have a few guys to keep stuff running and they want to buy a box of stuff that works. That is something we have to solve and that's something we're working on.

		'We're getting our act together on the phone side. I don't claim to have the last best solution. But we will have it.'

ZDNet: The real opportunity for PalmSource is to tap in to the Java developer base because of its size-- roughly 3 million developers. Most if not all of them are experienced in building network-based applications, which is important as more and more handheld applications --- especially enterprise ones --- are expected to work over wireless networks with a lot of latency. Why doesn't PalmSource focus on one of the largest and fastest growing developer bases as opposed to the relatively small number of developers focused on the PalmOS? Also, many of ZDNet's readers have complained about how there are different versions of the PalmOS currently shipping with different PalmOS-based devices and how there are software incompatibilities between them. Why does that situation exist? The reason these questions are related is because, with Java as a development target, you get to move away from the question of whether it's Palm OS 4, Palm OS 5, or the next one.

Nagel: So, which version of Java do you have? Which implementation? The problem with Java is that it isn't unified. It really isn't. All the phone guys have their own variants. On my phone, I was hoping to use Java for the network interface; all the provisioning and transaction-based stuff. I don't want to rebuild that every time. I don't want to have to build a central socket for every other player. It would help to have some standardization, but all the carriers have forked off different variants of Java. So, I'm trying to be agnostic on this stuff.

ZDNet: That could be said of older versions of Java, but Sun and the Java Community Process seem committed to leveling that out. Many of those problems have either already gone away, or will be done away with eventually.

Nagel: They also seem committed to have differentiation. Everybody wants their variant to be the standard. Also, I don't know about the complaints and the confusion over different versions of the operating system. The applications that run on OS 4 work almost without exception on OS 5. It's been a pretty good transition and we have actually not had a lot of complaints.

ZDNet: When I write about one handheld or another, several members of ZDNet's audience invariably respond with a technical support request because they think we have all the answers. A recurring question for the Palm OS has been getting old applications to run on the new OS.

Nagel: You're right. It hasn't been perfect. Not every application works. But the vast majority of them do. We do not get a lot of complaints about this.

ZDNet: But what about the way some products are shipping with the old OS while others are shipping with the new one? If all the apps worked, doesn't it make sense for everyone to ship with the new OS?

Nagel: It's not as bad as you make it sound. Of about 28 devices currently in the pipeline, only about three are on OS 4 and they're the ones being sold in China. So, that's one of the things that could have been nightmare for us, but it actually hasn't been one. I'm sure there are some glitches, but for the most part, it has been a pretty smooth transition.

ZDNet: I still think Java is a better direction. Based on what you've said so far, you're clearly trying to target and demonstrate some superiority over Microsoft and PocketPC. There's no one single development community that's more focused on that mission than the Java one. With PalmSource remaining focused on PalmOS, the phone guys working with Symbian and Java, and with RIM going Java all the way, Microsoft is succeeding at dividing and conquering. But, if you joined the battle, the outcome might be different. The other benefit is that Palm users immediately get access to a lot of Java applications that are portable from one device to the next. That's a strong selling point of Java--the compatibility of the applications with a variety of devices.

Nagel: Since the beginning of last year, we've gone from 12,500 applications available for the PalmOS platform to about 19,000 now. In terms of developers, we're up to 280,000. We're getting our share of I.Q. points focused on the platform, so we're not worried about lack of software. We have a large enough share of the market.

ZDNet: OK, why not have your cake and eat it too? You can stay with the PalmOS apps but still make it a great Java platform and then you could see which way the developers gravitate.

Nagel: I have no problem with that. I'd love to work with Sun. But Sun is very difficult to work with. They do not make it easy. We were one of the early members of the JCP. We tried to build a PDA profile, sort of a J2ME grown up a little bit. But we decided ourselves that that was sort of a bifurcation and that it created more problems than it solved. Sun didn't like it. They wouldn't support it, so we just decided that we weren't going to go through with it and that we would leave J2ME be.

ZDNet: Are you saying that PalmSource followed the rules of the JCP's framework and Sun vetoed it? Sun says it has never vetoed anything at the JCP.

Nagel: No, they didn't veto it. We actually got it approved. But in the end, there wasn't a lot of enthusiasm for promoting an extended platform. But I just want to address the attitude that you described where everything else goes away, operating systems become kernels, and everything runs on a runtime like Java. I just don't believe it.

ZDNet: ZDNet's readers are always interested in knowing what folks like you who have access to all the prototype devices have tucked away in your pocket. What are you packing right now?

Nagel: Actually, right this second, I don't have any. But if you must know, I try to rotate so that I have a good idea of what's out there and what they're capable of. But when I travel, I like to bring a Treo--particularly when I travel overseas. But when I'm here in the U.S., I really love Sprint. So, I've been using Samsung's devices. I've been using Samsung's I300 and I'm just about to get Samsung's new I500. ZDNet I could never get used to the soft-keyboards on the devices like the I300.

Nagel: I actually like it. I have big thumbs, but it still works for me. It does have a couple of irritating features, though. The worst feature is that sometimes, maybe one out of ten times, my ear taps the button that turns on the speakerphone and I get this blaring thing in my ear. But, other than that, I like it. But I also haven't gotten into the habit of doing a lot of email yet. As a communications device, I mostly use it as a phone. Maybe one reason I'm doing this is because the ideal device hasn't shown up yet. Right now, there are all these integrated devices out there, particularly the ones with Bluetooth like the Sony-Ericsson ones, which take hours to get the integration working. I don't want to be a systems integrator.

ZDNet: Are you suggesting that carriers that sell integratable devices like Bluetooth-based phones will actually sell a whole package that includes the phone and the PDA?

Nagel: Sure. It needs to be preconfigured and it needs to work out of the box. Right now, it doesn't. I'm much more likely to buy two devices that are designed to work together and that I don't have to configure after buying them. Also, I'm happy to have a larger data device that I have to keep in my briefcase. So, it has a good keyboard on it and a big battery and my phone needs to be able to attach parasitically when I dock them. So, instead of the problem where the data device drains your phone, how about the opposite?

ZDNet: Well, now you're starting to talk about an industrial design issue, but earlier you said you really don't have that much influence over the industrial design process.

Nagel: I have no formal influence but I certainly jawbone our licensees like crazy. Some of them listen and some of them don't. Look, this is an extremely important area for us. The size of the market for phones is an order of magnitude and a half times the size of the market for PDAs and handheld computers and, frankly, we think this is our future. We're putting a lot into it.

ZDNet: What are your thoughts about the international front? One thing we hear is that the PalmOS doesn't have the sort of international critical mass it needs to succeed on a global scale.

Nagel: You mean Europe? You find this in Europe a lot. How do I put this delicately? I think the Europeans are slightly xenophobic--in the Eurocentric sense. They think Symbian is the obvious solution. The Europeans like this sort of collaboration idea. They haven't been real successful at it in the past so I'm not sure how successful it will be going forward. But Europe is a great case in point. We gained ten share points in Europe over the last quarter. It's the one market in the world that's growing. One reason we weren't in consideration in Europe for so long is that none of our devices worked on their networks until last year. (Editor's note: Europe has standardized on GSM for voice and GPRS for data and until last year, there were no PalmOS-based offerings based on GSM or GPRS technologies). That's changing, especially now with Samsung's new I500. So we'll be getting good visibility. I'm perfectly happy to fly under the radar. We're getting our act together on the phone side. I don't claim to have the last best solution. But we will have it.

ZDNet: Any thoughts on CDMA vs. GPRS?

Nagel: I think the early always-on data CDMA stuff is clearly better. It's roughly twice the throughput, maybe more of GPRS. It seems to be much more seamless. But CDMA also does some pretty goofy things. The SIM card on the GSM/GPRS side is absolutely fantastic. I travel to another country and all I have to do is move the SIM card to another phone and I'm up and running. There is no SIM card in the CDMA world and I don't understand why they didn't adopt that. For me, that is one issue.