Ray Ozzie talks peer-to-peer
- 13 December, 2000 17:06
Ray Ozzie, the driving force behind the creation of Lotus Notes: In the B2C (business-to-consumer) space, there's definitely interest. There's a strong uptake amongst a certain set of customers - either in marketing or [those] having a relationship with a small group of people who have a natural affinity to one another.As for [intracompany] ad-hoc workgroups, I don't know if there's interest because we haven't actually been talking to customers in that realm. But to be as blunt as I can, I don't believe that Lotus or Microsoft are going to leave holes in their product line [that allow us to target the market] for internal colaboration. We're focusing essentially on external collaboration in terms of customers or supply-chain applications.
My challenge is to make our platform relevant as soon as possible. I'm going to use every resource at my disposal and every bit of experience that I have to do that. Microsoft will be using every resource at its disposal to either treat this as friend or foe, depending on how it suits them.
There is, and will be, a free version of the client. Part of our challenge is to become a relevant feature on most people's desktops as quickly as possible. We've also identified two versions of the product that . . . have substantial value to the individual or to the enterprise [and] will represent a valid, reasonable purchase that we aren't going to have to twist somebody's arm to buy.
If you start to be a heavy user of the product and you're dealing with lots of relationships or projects, you'll want features that bring activity in those areas in to your consciousness more readily.For instance, when something happens in this tool, put a notifier on my bottom line, put it up on the screen, [or] send me an e-mail. We can also [provide a version with] much better Microsoft Office integration than [in the standard version].
On the enterprise side, there's enterprise systems management: If I'm an enterprise IT guy and I build a directory infrastructure, I don't want people making up names in Groove and starting to use new names to identify each other - I want to deploy my naming infrastructure. I want to make sure that I control what tools they can use in these clients. [For this reason, Groove has] the ability to centrally manage distributed clients. And then there are certain services we provide that essentially mean we'll be selling service-level agreements to companies that are interested in them.
It depends on how you conceptualise the future of phones and PDAs and [other handheld devices]. I am a skeptic in terms of increasing the level of functionality of what we refer to as a phone - I just use the phone to call people. So I'm not really focusing on that as a platform for Groove. I'm really big on things like Blackberry pagers; they're great devices. But the really interesting stuff is [a device that falls] somewhere between the laptop and the PDA, maybe something in the form of a tablet with a fold-out keyboard. That is a perfect device for Groove.
It's totally integrated. There is an expectation amongst consumers that there's an increasing level of commodity function which they should get for free. [So] if you really want to have something that is sustainable, it has to have solution capability - solutions where a third party can connect to the other [relevant] things in the enterprise. Transaction systems, sales force automation systems, [and] document management systems are the enterprise systems which have relevance and data. We have a client-side technology, but it's supposed to integrate with the server side, not just exist.
There are two basic mechanisms. The first one is just to use [available] APIs. Our programming environment is Windows, so it's very easy to take any COM-compliant language and tie it in. The second one is a Knowbot toolkit, which lets you essentially create a bot that has a certain level of function and runs on a server. Individuals just invite the bot in, just like it's another person. The goal of this bot is to suck information from the server, put it into the knowledge repository, and make it available for people to query.
I've never seen anybody, in the whole database business, standardise how bits should be stored on disk and so forth. That's how Oracle and Sybase and Microsoft compete. Our goal is to make sure that we've factored our product in the right way so that we can be aligned with standards efforts as they emerge.
We don't have an integrated JVM [Java Virtual Machine]. You can build Groove applications in J++, but you can't build them in Sun Java because in order to do that [we would have to include] a JVM. [Users can also] embed the Internet Explorer component and write Java within that.
Unlike any other product I've worked on, the Groove client is nothing more than a component framework that's offered up to a developer. So you can develop a completely different UI [user interface] that looks nothing like what we've got. We've even experimented with what it would be like to take some of the controls and move them out to hardware on a tablet device, as opposed to being in software on the client. And if someone wanted to take that client functionality and move it into a different client, it would be easy to do.
We're shipping an information sharing [technology]. Can it be used for bad? Absolutely. . . . I'm not trying to thumb my nose at intellectual property holders - my whole business rests on intellectual property, [and] if my source code is out there, it damages me just like it damages an artist or a publisher. I respect intellectual property, but I'm not going to make Groove so that it can't share my source code.
Somebody's going to hack the OS of TiVo devices so that you can do peer-to-peer sharing of TiVo drives. Once you do that, anybody could record [a TV] show, and then anybody else could watch it. [I don't know] why that hasn't happened yet.