The pains of being ahead of the curve
Feb. 26th, 2008 12:00 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
jducoeurThanks to Garett Rogers for pointing out that Google has implemented "chatback" -- a new feature that allows people to click on links to open a chat session with you.
It's a lovely feature. Indeed, it was a lovely feature when we invented it three or four years ago at Convoq -- I did most of the design, and it sounds like Google's system is almost a direct copy. (Indeed, my ASAP Link was live on my LJ profile until we shut the project down.) It'll be gratifying if it proves useful to folks -- I thought it was a good idea then, and I still do -- but man, I'm getting really tired of inventing new things and have them finally proven out years later by other people. This is something like the third or fourth in just the past year.
Of course, it also leads me to wonder what the status of that particular patent is. I'm pretty sure that the ASAP Links patent wasn't fully granted yet by the time we shut down, but it might still be actively in-process. If so, Convoq's patent portfolio just jumped in value again: IIRC, it was one of our most defensible-looking patents, right up there with the Meet ASAP feature. Once again, I find myself with deeply mixed feelings about the whole patent thing.
ETA: Amusing. The Google version is similar to ASAP Links in the broad strokes -- except that it doesn't work nearly as well. For example, we provided a number of different form factors for ASAP Links, precisely because there is no good one-size-fits-all solution. The Google Chat badges require iframes, so they can't be used in, say, your LJ Profile, which is the place one would *most* like to put it. (Because they can be easily disabled, these links are *much* better for publishing publically than simply posting your handle.) ASAP Links could be used with simple image buttons, or even just links, so they could be used almost anywhere. And of course, ASAP Links were vastly more powerful -- not only could you do text chat, you could start a full audio/video conversation with them.
It looks like, in classic Google fashion, this is a very simplistic first cut. It'll be interesting to see whether they wind up implementing all the nuances that we decided were necessary to make the idea fully useful. And it does make me speculate that, if CommYou doesn't work out, there's something to be said for reviving Convoq's technology ideas (with a greatly simplified UI so ordinary mortals can use the damned thing)...
It's a lovely feature. Indeed, it was a lovely feature when we invented it three or four years ago at Convoq -- I did most of the design, and it sounds like Google's system is almost a direct copy. (Indeed, my ASAP Link was live on my LJ profile until we shut the project down.) It'll be gratifying if it proves useful to folks -- I thought it was a good idea then, and I still do -- but man, I'm getting really tired of inventing new things and have them finally proven out years later by other people. This is something like the third or fourth in just the past year.
Of course, it also leads me to wonder what the status of that particular patent is. I'm pretty sure that the ASAP Links patent wasn't fully granted yet by the time we shut down, but it might still be actively in-process. If so, Convoq's patent portfolio just jumped in value again: IIRC, it was one of our most defensible-looking patents, right up there with the Meet ASAP feature. Once again, I find myself with deeply mixed feelings about the whole patent thing.
ETA: Amusing. The Google version is similar to ASAP Links in the broad strokes -- except that it doesn't work nearly as well. For example, we provided a number of different form factors for ASAP Links, precisely because there is no good one-size-fits-all solution. The Google Chat badges require iframes, so they can't be used in, say, your LJ Profile, which is the place one would *most* like to put it. (Because they can be easily disabled, these links are *much* better for publishing publically than simply posting your handle.) ASAP Links could be used with simple image buttons, or even just links, so they could be used almost anywhere. And of course, ASAP Links were vastly more powerful -- not only could you do text chat, you could start a full audio/video conversation with them.
It looks like, in classic Google fashion, this is a very simplistic first cut. It'll be interesting to see whether they wind up implementing all the nuances that we decided were necessary to make the idea fully useful. And it does make me speculate that, if CommYou doesn't work out, there's something to be said for reviving Convoq's technology ideas (with a greatly simplified UI so ordinary mortals can use the damned thing)...
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
(no subject)
Date: 2008-02-26 05:48 pm (UTC)We did that in 1996
Date: 2008-02-27 02:13 pm (UTC)I'm not sure anybody used it, though. In retrospect, that MIME type approach must've been a pretty higher barrier; it probably would've made more sense to have the Navigator team add a cooltalk: URL. I couldn't do that when I first wrote the feature, before we got acquired by Netscape, but it should've been obvious later on.
Re: We did that in 1996
Date: 2008-02-27 03:01 pm (UTC)The thing about both the Google and Convoq solutions is that they Just Work. The Google version looks to be pure DHTML, which probably works pretty much everywhere that has Javascript (although I suspect that they have some serious frame-boundary limits on the look and feel of the conversation). Not that I've tested it, mind, but that's what I would expect from Google.
The Convoq one was based on Flash, which in practice is almost as good (well over 90% penetration, and more consistency than DHTML), and *boy* did we startle the heck out of people when we pointed out that you could do A/V conferencing with just Flash. One reason why ASAP's following was loyal (if small) was the "magic" effect of being able to email a link to a random person, which they could just click on and *poof* be in a videoconference. No setup or install on the receiving end -- it just worked...
Re: We did that in 1996
Date: 2008-02-27 03:12 pm (UTC)Oh, yeah, it definitely sounds like Convoq did it better. I'm just thinking about patentability.
You're probably right. Looking back, I think we (all of Netscape) were probably in denial about IE; we just assumed it was going to stay behind forever. When we released 4.0, IE 3.0 had just come out (4.0 was nearly a year later), and we were too busy laughing at CaptiveX.
Re: We did that in 1996
Date: 2008-02-27 03:42 pm (UTC)It was a near thing. I still suspect that if Andreesen hadn't shot his mouth off, the world today might be very different...
Re: We did that in 1996
Date: 2008-02-27 04:25 pm (UTC)You mean when he started talking about "crossware", browser-based apps that would make the OS irrelevant? Yeah. Do Not Taunt Happy Fun Monopolist.
It didn't help that the first crossware we released was Netcaster, which required a high-end machine (200MHz in 1997) and still ran slowly. Nor that it was a massive pain to build, because of bugs in 4.0. Remember the <layer> tag? You could construct them in JavaScript, and Netcaster did a lot of that. But it turned out that there was some sort of race condition in the engine, which meant that you couldn't have more than one dynamically constructed layer at a time, or they'd all freeze. So Netcaster had to use a Java object that maintained a queue of URLs to load into layers, and loaded one at a time. I don't think we ever publicized the technique, so anybody else who tried doing anything fancy with layers would've just said, "Ah, this is too broken to use".
Re: We did that in 1996
Date: 2008-02-27 06:47 pm (UTC)I don't remember the exact words, but it boiled down to, "Browsers are the way of the future, and Microsoft will go the way of the dodo in the next decade". Waving the red flag in front of Gates was a very bad idea -- it woke him up *just* in time to execute a fairly remarkable turnaround on MS' part and save the company.
And layers -- wow, y'know, I had completely forgotten that tag...
Re: We did that in 1996
Date: 2008-02-27 07:27 pm (UTC)Yeah, that was it. Too much Kool-Aid.
Yeah, it was a serious false start for DHTML. I remember, while I was on the Netcaster team, I spent some time writing a widget library for JavaScript based on layers. ("Widget" in this case meaning "GUI element".) The idea was that, if we wanted people to write DHTML channels, we needed to make it easy for them; and we needed to provide something cross-browser (since we already knew IE 4 was going to do DOM). It was pretty good, for its time; you could write JavaScript to give you an interface just like (OK, a lot like) an ordinary desktop app. But then Netcaster got cut in the Great Reorg, when 25% of the company was let go because nobody was paying for Communicator any more.