The difficulty there, frankly, is keeping it simple enough. This bit us on the ass with ASAP: it was *quite* sophisticated in its presence and availability management, but we found that people tended not to use the more interesting features. It was just a bit too feature-dense.
So this time, we're playing it a bit smarter -- what I think of as the Google/Apple approach. The feature set is intentionally stripped-down to the basics to start with, expressing the core ideas and giving users the most critical functionality with a reasonably simple interface. As we find out what the users really *want*, we'll build those features in. We lost far too much time in ASAP building systems that were clever, powerful, and just not useful enough to be worth the effort and complexity.
As for calendar integration: not out of the question in the long run, but not simple to do well. We did pretty deep Outlook integration for ASAP, and developed a lot of respect for what a pain in the ass it is. And this is a lot more complex a problem, because most calendar systems aren't designed to express the kind of semantics we need. (And users don't use those features even when they exist.) Distinguishing between "this is a meeting, and I shouldn't be interrupted" and "focus on this project at my desk for a couple of hours" -- well, they're semantically quite different, and have different implications for presence and availability, but they generally look much the same to the code.
Presence management is a remarkably complex art once you get deeply into it: much of Convoq's lifespan was spent on just that, and we had an entire server project dedicated solely to it. Zingdom is only scratching the surface of it so far, because we want to keep things simple to begin with. We'll see where we wind up going with it -- it's just one of the many dimensions we will be playing with. (Frankly, I expect identity-management to be a more interesting and immediate problem, since that's a space in a lot of ferment right now...)
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(no subject)
Date: 2007-07-23 01:26 pm (UTC)So this time, we're playing it a bit smarter -- what I think of as the Google/Apple approach. The feature set is intentionally stripped-down to the basics to start with, expressing the core ideas and giving users the most critical functionality with a reasonably simple interface. As we find out what the users really *want*, we'll build those features in. We lost far too much time in ASAP building systems that were clever, powerful, and just not useful enough to be worth the effort and complexity.
As for calendar integration: not out of the question in the long run, but not simple to do well. We did pretty deep Outlook integration for ASAP, and developed a lot of respect for what a pain in the ass it is. And this is a lot more complex a problem, because most calendar systems aren't designed to express the kind of semantics we need. (And users don't use those features even when they exist.) Distinguishing between "this is a meeting, and I shouldn't be interrupted" and "focus on this project at my desk for a couple of hours" -- well, they're semantically quite different, and have different implications for presence and availability, but they generally look much the same to the code.
Presence management is a remarkably complex art once you get deeply into it: much of Convoq's lifespan was spent on just that, and we had an entire server project dedicated solely to it. Zingdom is only scratching the surface of it so far, because we want to keep things simple to begin with. We'll see where we wind up going with it -- it's just one of the many dimensions we will be playing with. (Frankly, I expect identity-management to be a more interesting and immediate problem, since that's a space in a lot of ferment right now...)