Entry tags:
Some thoughts on Twitter
Okay -- having "followed" a modest number of people on Twitter, looked around some and started to get a handle on it, I'm beginning to get the idea. Thanks to
hakamadare, who I think accurately described it as having a prescriptivist/descriptivist problem that was confusing me. I'm coming around to the opinion that it's mildly interesting but marginal for the purpose that's often claimed for it -- micro-blogging, essentially -- but useful and a *delicious* proof of concept for CommYou for its *best* purpose: informal group conversation.
For those who are curious, here's a little initial analysis of the thing: both what it is and why it's interesting. Keep in mind that I'm still very much a newbie there, but I'm grokking it a lot better now than I did 24 hours ago. This is going to be pretty long, and in-depth. It's important stuff for me, because of what it says about CommYou, but is only interesting those who care about the whole social-networking/blogging space. So in deference to those who just don't care, here's a cut-tag:
What It Is: Let's get the formal description out of the way first. Twitter is kind of like a blogging platform, but you only have 140 characters per post. Yes, really. It was originally designed for use via SMS (texting from your phone), and the length limit comes from that. Technically speaking, posts ("updates") can be longer, but since so many people are accessing it via SMS there is clearly heavy social pressure to stay under the limit, and the UI encourages you to do so.
So you post updates, and other people follow them. You can follow the updates of the people you're interested in via SMS, via IM, at the Twitter website, or via various tools that plug into the Twitter API. You can reply to their updates via direct messages to them (which AFAICT almost nobody does), or via a rough-and-ready conversation convention that has grown up among the users and wound up getting formalized into the tool.
Social networking: Okay, is any of this sounding familiar? It should: Twitter is actually one of the closest tools to LiveJournal I've come across. Yes, it *looks* a lot different, and the details are wildly different. (And it is idiotically over-simplistic compared to LJ.) But the broad spirit of the thing is very similar. In particular, you are formally "following" specific people -- effectively, your friends list. You can control who can follow you -- a rougher but simple version of LJ's access control mechanisms. Your home page shows a list of all the most recent updates from the people you're following -- just like your Friends page.
In other words, just like LJ, Twitter is a blogging/social network hybrid. Even more amusingly, just like LJ, they clearly *thought* they were just building a blogging platform, and sort of accidentally developed a social network because they implemented a following mechanism. The length limit changes the nature of the beast considerably, but mostly that affects the kinds of postings people make: it's less about the in-depth debate, more about up-to-the-minute events. The deep integration with near-real-time tools like SMS and IM makes it very powerful.
It is mildly annoying in that it's Yet Another Frikkin' Social Network to keep track of, and integrates fairly badly with your other networks -- in this respect, I expect CommYou to do much better.
Small-scale Conversation: It's pretty clear that the creators hadn't realized what they were building. Twitter really gets interesting in the fact that you can reply directly to other peoples' posts, and the visibility of those replies is subtle and controllable -- by default, I (a third party) will tend to see replies if I saw the original posting.
The result is that you can have lightweight group conversation inside Twitter. It's fairly crappy, lacking a lot of useful details -- in particular, the total lack of formal threading means that it's hard to have deep conversations, and things will tend to wend around each other a lot -- but functional. I'm noting people who are beginning to use Twitter *instead* of IM, because it fills the same niche better for them. This is *hugely* important, and is the way in which Twitter is most like CommYou. (Albeit at a great remove.)
The thing is, group conversation is precisely where the Web is still weakest. You may not see this, as an LJ user, but broadly speaking the tools are pretty bad. The near-realtime tools, in particular, are very one-to-one: multiparty IM chat is still primitive (in large part because of its lack of context and persistence), and multiparty SMS mostly doesn't exist. So Twitter is filling an important niche, by providing a very crude proof-of-concept social network that has semi-realtime group conversation laid on top of it.
The fact that it's being pretty successful despite being very primitive is *very* reassuring to me, because this is the heart and soul of the CommYou story. My claim is that the thing that social networks need is good mechanisms for conversation among the communities within those networks. Twitter, like LJ, seems to support that.
On the flip side, Twitter also emphasizes something that I've been starting to suspect: I can't ignore the "blogging" side of the equation. It's notable that both LJ and Twitter -- the nearest things CommYou has to recent predecessors -- aren't just about the conversations, they're also about saying What I'm Thinking Now. I suspect that's critical. CommYou is going to be designed to be strong at the subsequent conversations, but I need to make sure I don't do anything to discourage people simply speaking their minds. (At the moment, this mostly affects some of my wording choices in the UI, but there may be feature decisions it affects down the line.)
API: One thing that Twitter got right is implementing an open API early on. As everyone signing up has noticed, their Web UI is pretty terrible. Far as I can tell, not many serious Twitter users use it -- instead, they use tools that were built on top of the API.
This doesn't surprise me, but it underlines the need for good APIs for CommYou. This is something I've been planning from the beginning (my thanks to
metageek, in particular, for helping remind me at the right moment that good APIs come from building the system around them in the first place), but I'm probably going to have to step it up. My current APIs are designed mainly around one specific usage model for CommYou. At some point soon, I need to step back, and think about how they should be generalized to be useful for other use cases.
IM Integration: I was floored when I realized, yesterday, that their IM integration is Jabber-only. You can use Twitter with GTalk, LJ's chat system, and other similar Jabber clients, but *not*, as yet, AIM. This is mostly for reasons of difficulty, and I sympathize with them -- AIM integration is a pain in the butt. OTOH, it's a pain in the butt that
learnedax and I wrestled down a year or so ago at Zingdom, so I am mildly amused that they haven't pulled it off yet, given how useful it would be.
There's an important implication for me, though: it might be worth it to just do Jabber integration soon. I mean, my IM integration stories for CommYou are pushed pretty far off, precisely because getting AIM up and running is likely to be a nuisance. But I could probably get GTalk integration working in a week or so. I wasn't going to bother, but if it's looking worthwhile for Twitter, it might be worthwhile for CommYou. That's very interesting, and something for me to toy with.
Conclusions: Let's get real -- Twitter is a primitive toy, that barely scratches the potential of what can be done here. But it demonstrates some of the ideas that are integral to CommYou, particularly the way in which it begins to weave together the up-tempo and down-tempo sides of conversations. I think we can do a *lot* better than this, but it's damned encouraging to see how much can be done even with a very simplistic solution. This encourages me to start getting things out in front of people soon, so I can let the users tell me when I have gotten to the point of being useful, rather than assuming I need full functionality first...
![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
For those who are curious, here's a little initial analysis of the thing: both what it is and why it's interesting. Keep in mind that I'm still very much a newbie there, but I'm grokking it a lot better now than I did 24 hours ago. This is going to be pretty long, and in-depth. It's important stuff for me, because of what it says about CommYou, but is only interesting those who care about the whole social-networking/blogging space. So in deference to those who just don't care, here's a cut-tag:
What It Is: Let's get the formal description out of the way first. Twitter is kind of like a blogging platform, but you only have 140 characters per post. Yes, really. It was originally designed for use via SMS (texting from your phone), and the length limit comes from that. Technically speaking, posts ("updates") can be longer, but since so many people are accessing it via SMS there is clearly heavy social pressure to stay under the limit, and the UI encourages you to do so.
So you post updates, and other people follow them. You can follow the updates of the people you're interested in via SMS, via IM, at the Twitter website, or via various tools that plug into the Twitter API. You can reply to their updates via direct messages to them (which AFAICT almost nobody does), or via a rough-and-ready conversation convention that has grown up among the users and wound up getting formalized into the tool.
Social networking: Okay, is any of this sounding familiar? It should: Twitter is actually one of the closest tools to LiveJournal I've come across. Yes, it *looks* a lot different, and the details are wildly different. (And it is idiotically over-simplistic compared to LJ.) But the broad spirit of the thing is very similar. In particular, you are formally "following" specific people -- effectively, your friends list. You can control who can follow you -- a rougher but simple version of LJ's access control mechanisms. Your home page shows a list of all the most recent updates from the people you're following -- just like your Friends page.
In other words, just like LJ, Twitter is a blogging/social network hybrid. Even more amusingly, just like LJ, they clearly *thought* they were just building a blogging platform, and sort of accidentally developed a social network because they implemented a following mechanism. The length limit changes the nature of the beast considerably, but mostly that affects the kinds of postings people make: it's less about the in-depth debate, more about up-to-the-minute events. The deep integration with near-real-time tools like SMS and IM makes it very powerful.
It is mildly annoying in that it's Yet Another Frikkin' Social Network to keep track of, and integrates fairly badly with your other networks -- in this respect, I expect CommYou to do much better.
Small-scale Conversation: It's pretty clear that the creators hadn't realized what they were building. Twitter really gets interesting in the fact that you can reply directly to other peoples' posts, and the visibility of those replies is subtle and controllable -- by default, I (a third party) will tend to see replies if I saw the original posting.
The result is that you can have lightweight group conversation inside Twitter. It's fairly crappy, lacking a lot of useful details -- in particular, the total lack of formal threading means that it's hard to have deep conversations, and things will tend to wend around each other a lot -- but functional. I'm noting people who are beginning to use Twitter *instead* of IM, because it fills the same niche better for them. This is *hugely* important, and is the way in which Twitter is most like CommYou. (Albeit at a great remove.)
The thing is, group conversation is precisely where the Web is still weakest. You may not see this, as an LJ user, but broadly speaking the tools are pretty bad. The near-realtime tools, in particular, are very one-to-one: multiparty IM chat is still primitive (in large part because of its lack of context and persistence), and multiparty SMS mostly doesn't exist. So Twitter is filling an important niche, by providing a very crude proof-of-concept social network that has semi-realtime group conversation laid on top of it.
The fact that it's being pretty successful despite being very primitive is *very* reassuring to me, because this is the heart and soul of the CommYou story. My claim is that the thing that social networks need is good mechanisms for conversation among the communities within those networks. Twitter, like LJ, seems to support that.
On the flip side, Twitter also emphasizes something that I've been starting to suspect: I can't ignore the "blogging" side of the equation. It's notable that both LJ and Twitter -- the nearest things CommYou has to recent predecessors -- aren't just about the conversations, they're also about saying What I'm Thinking Now. I suspect that's critical. CommYou is going to be designed to be strong at the subsequent conversations, but I need to make sure I don't do anything to discourage people simply speaking their minds. (At the moment, this mostly affects some of my wording choices in the UI, but there may be feature decisions it affects down the line.)
API: One thing that Twitter got right is implementing an open API early on. As everyone signing up has noticed, their Web UI is pretty terrible. Far as I can tell, not many serious Twitter users use it -- instead, they use tools that were built on top of the API.
This doesn't surprise me, but it underlines the need for good APIs for CommYou. This is something I've been planning from the beginning (my thanks to
![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
IM Integration: I was floored when I realized, yesterday, that their IM integration is Jabber-only. You can use Twitter with GTalk, LJ's chat system, and other similar Jabber clients, but *not*, as yet, AIM. This is mostly for reasons of difficulty, and I sympathize with them -- AIM integration is a pain in the butt. OTOH, it's a pain in the butt that
![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
There's an important implication for me, though: it might be worth it to just do Jabber integration soon. I mean, my IM integration stories for CommYou are pushed pretty far off, precisely because getting AIM up and running is likely to be a nuisance. But I could probably get GTalk integration working in a week or so. I wasn't going to bother, but if it's looking worthwhile for Twitter, it might be worthwhile for CommYou. That's very interesting, and something for me to toy with.
Conclusions: Let's get real -- Twitter is a primitive toy, that barely scratches the potential of what can be done here. But it demonstrates some of the ideas that are integral to CommYou, particularly the way in which it begins to weave together the up-tempo and down-tempo sides of conversations. I think we can do a *lot* better than this, but it's damned encouraging to see how much can be done even with a very simplistic solution. This encourages me to start getting things out in front of people soon, so I can let the users tell me when I have gotten to the point of being useful, rather than assuming I need full functionality first...
no subject
Early on I had heard about twitter, but didn't look into it. Next thing I'm seeing are LJ posts that are a collection of all the "tweets" from a single day. These folks were using a tool that automatically collected and posted the tweets.
The tool is called LoudTwitter: http://www.loudtwitter.com/
An example off my friends page is:
http://darkmoon.livejournal.com/2238072.html
no subject
no subject
(Also on MySpace, GMail... I don't do Facebook.)
no subject
no subject
no subject
no subject
no subject
no subject
Seeing as how at current the IM world is slowly coalescing into interoperable standards, but not there yet, and social networks are only now glancing in that direction, it is interesting to see developers taking matters into their own hands and building unified clients. We've seen unified IM clients for years, and at this point they are more popular than balkanized clients. Social Networking is likely to go a similar path.
no subject
So I think it'll take a slightly different path here, with *applications* working across the networks (as CommYou is intended to eventually do), but not really unifying the entire Social Network experience for a while yet, at least until there's a more standardized idea of what a Social Network *is*. I suspect it'll happen eventually, but I don't think we even yet have a clear idea of what it's going to look like...
no subject
(Actually, we had vast amounts they don't have, including instant A/V meetings, screen sharing, anonymized POTS integration, ability to use AIM as your desktop client for the thing, and more. Sometimes I marvel at how much we built at Convoq.)
It has one or two extra features (in particular, the feed integration), but mostly it is a reminder that we failed to execute on the marketing side -- that there's a genuine product there. *Sigh*...
no subject
Do you think Convoq _built_ too much at the expense of pushing the product? Feature Creep in development can be a killer. Or, to use a pair of cliches, Perfect Is The Enemy of Done, and Don't Bite off More Than You Can Chew.
Also correct me if I'm wrong, but the impression I got from Convoq was that they wanted the Enterprise market? Was it ever intended for widescale release as a free download to the general public? Maybe its more about the target, and less about the product itself.
no subject
Oh, undoubtedly, but not for either of the reasons you describe. ASAP was *fabulous* in most respects -- both done and rock-solid -- but it was too damned complicated. It intimidated the hell out of people, far as I can tell.
Also correct me if I'm wrong, but the impression I got from Convoq was that they wanted the Enterprise market? Was it ever intended for widescale release as a free download to the general public?
Yes and no. The problem is, Convoq's primary target market was small-business, which is a tricky middle ground. We actually *did* have a free version of the product for a while, but because our management always sort of saw that as a come-on for the commercial one, they never pushed it right and nobody ever used it -- it was a very half-hearted tactic...