Be careful of the uncanny valley between "too slow" and "chit-chat".

There's room in the world for "chit-chat": an online conversation that happens in real-time and consists of rapid exchanges of short sentences and sentence fragments. Google Chat, AIM, et cetera are fine for that.

And there's successful media that involve exchanges of whole paragraphs, but not in real time. This is email, Stack Overflow, HN, et cetera.

But, in my experience, there's not a lot of in-between. My own native medium is the paragraph. If I'm lucky I can, with effort, condense something insightful into a sentence. But often, in an IRC or AIM-style live chat, by the time I've finished typing that sentence the conversation has moved on. It just takes too long to compose.

The antidote for that is to stick to short fragments, follow the rhythm of the conversation, don't say anything too tricky or controversial that will require a lot of exposition and framing, and defer longer-form discussion to a different medium, like the blog post or the email. In other words: If the medium is dynamic, "chit-chat" is the natural result.

Most people just aren't good-enough typists or writers to conduct a rapid-fire real-time textual exchange of substantive sentences and paragraphs. If people could watch me compose HN posts in real-time I'd never post here again. It would be infuriating to watch me and even more infuriating to be watched: I type things that are clumsy or flat-out wrong, then I read them in horror, then I delete them. Welcome to writing. That's how it works, for most people.

None of this is meant to discourage you from trying to build a hosted service for private-to-semi-public, limited-participation, limited-time threaded discussion. It's a thing I've wanted myself and have thought about building. But don't get so caught up in the glamour of real-time that you simply reinvent GChat or Skype or AIM. You may have a shiny new asynchronous real-time hammer, but it still isn't the right tool for driving a screw.

You keep talking about products and features and tech stacks. Can you try to express this idea in terms of people?

If you can fill in the sentence "James would rather be at his son's baseball game, but he can't, because he needed to get an answer about X from Cindy, but he couldn't, because of Y" and you solve Y and get James an answer then there is probably space for a business here.

James could buy a mobile? And then go the baseball game and participate in the discussion from his mobile, maybe even use the touchscreen to draw diagrams on the virtual whiteboard.

If he was using GChat, him loggin off would have put an end to the conversation.

If he was using GMail, they were probably just sending each other big mails, like what was conclusively decided after today's meeting, and that could wait.

I think you're missing Patrick's point?

You could describe an alarm clock as a mechanical device which steadily advances its state with passage of time, and for which you can set a trigger to start making mechanical noise at a certain pre-set state.

Or you could start off with "Joe, a young adult, is having trouble appearing on time for his white collar job, which he want sto keep and succeed in, but can't because health problems/kids/partying are keeping him awake at night, so he can't reliably wake up in the morning. This device will make him get up such that he has enough time to get into the office and be a good employee in the eyes of his boss".

Closer to home, hacker news is not a visually desaturated lisp-coded online forum thingamajig, hacker news is a place for tech/business hackers to meet and talk shop. Quora is a place for knowledgeable people to make a measurable, satisfying difference by spending 5 minutes to answer a question.

Who are the people with the problem? What is the problem? You won't get traction without solving these questions.

Yeah, I had the same idea (give or take) except I couldn't figure out how to make money on it. If that's not your goal, then the idea makes for an excellent open source project.

Also, you might want to have a look at http://dialog.gs - a beautifully designed flop that is. Guy was using the same position as you - small, portable, disruptive, fits between the chat and the email, etc. It appeared to have flopped because he couldn't get people on it, which is something to think about.

I guess Branch is trying to do this already. http://branch.com/

A pity about their clueless home page. Please tell me what you're going to give me in exchange for my email address.

I understand the importance to a business of reaping the email address up-front. But there needs to be a clear quid pro quo. For all I can tell from the home page, I'm signing up for Amway, and the next message I get will be from one of my dear cousins trying to shame me into buying some overpriced soap.

So can anyone here tell me what Branch is about?

You could check out bulletin.branch.com for some context.

This serves as a demo of what they are trying to do. http://branch.com/b/obvious-and-branch-partner-up

I hit that link before griping. [EDIT: the bulletin's root link, that is, not the deep link you provided - thanks for that.] I was that kind, at least.

It links to a blog. This margin is too small to contain my full lecture about why you do not want to introduce your product to the world via a blog, but here's the condensed version: The top of the Bulletin is an out-of-context article about Twitter integration with the Mystery Product. Then the next half-dozen entries are beautiful photos of people being hired to work on the Mystery Product. Then I got bored with scrolling and gave up.

I'm not going to hunt for your call to action. If I wanted puzzles I'd be training on http://web.mit.edu/puzzle/www/

They seem to be doing exactly what I'm trying to do. I think they lack a feature or two, but it could be on their roadmap. Thanks for telling me about this. I don't know if I should feel happy or sad. :|

You should be happy. It means your idea is somewhat validated in that someone else is willing to put money and effort into making conversation better on the internet ;)

*disclaimer: I work at Branch, and we are indeed focusing on many of the problems you raised in your original post. If you look at the @branch twitter account you'll see many more examples of branches that you'd be interested in reading.

I understand. But, I'm just a learner trying to find a pet project that I could be excited about. This was it, but knowing that there are money-backed teams somewhere in the states working on this is discouraging. I was hoping to use this opportunity to work hard and learn async web programming, say node.js/socket.io or maybe meteor.

If I would have been an entrepreneur, I would have hit you guys head-on with a better implementation. ;) :P

Anyway, all the best guys. I did shoot you a mail, and am also waiting for an invite at the least, if possible. :)

Interesting, inviting people would seem to be the first difficulty.

Consider this: I want to chat with certain people about a certain topic (such as a product feature at work). I want it archived but only visible to the participants unless I add someone later or make it public. I want a url that acts as an invite and a convienient way to invite someone, in which they receive an email. I also want to be able to pre-invite people ( a group) and then I could paste a link in our Skype chat that everybody in that group gets auto-added to this chat. When I'm done I want to be able to have the URL be a wiki, with export as HTML or Markdown. I want it to support Markdown. I also want to be able to promote a chat message to a post and have it appear expanded and formatted in the wiki.

G+ meets almost all of your criteria, only thing it's missing is the ability to toggle discussions public or private after the fact, but you can at least control that aspect when creating the conversation.

G+ is surely the closest thing to what I envision. Yet, I am not convinced.

I once tried 'chatting' with my girlfriend on google+ using a private channel. The fear of some error happening, and our messages being accidentally shown to public was too much, and we switched back to chat. Google+ us social and public. Not technically. But that is what they are trying to be, and advertising to be.

I have had private chats using G+ posts and I attest that they never turned public. Once you make a post, you can't change it's visibility. You can only delete it, or lock it down - comments or reshares.

Though I think gPlus + Wave would be the ideal such environment for me.

I think there is space (one thing for me working in your favour is that i don't want to sign in with my gmail or facebook.. i want a separate account, if any), but traction is a major problem- if it feels like a ghost town, no one is likely to use it. Getting over that initial hurdle will be huge, especially since you can't really gamify it like quora / so, and no advertising or social proof benefit like FB / twitter.

Thanks for the reply. A few questions.

Why do you not want to sign in with gmail? I had imagined a gmail password-less signup. There would be only one field: your email ID. The login would then cling on to your computer, unless you want to use the website from another computer in which case, you can enter your emailID again, and you're back.

I think an easy sign in is the most easy way to avoid ghost towns. The way I'm avoiding the ghost town effect is by asserting that this is not a social website. The next time you want to have a discussion with bob and harry, you could just add their emailIDs to the discussion, or paste the discussion link on them. They would click on it, click 2 more buttons to sign in, and voila: They're in the discussion in a flash. The next time they visit, they will have their dashboard with the previous discussions they made.

The point is, you don't need thousands of users to be in town, to enjoy this town. Only when I have MANY users, will I then roll out a public discussion feed. Till then, its only micro-site tool to help you talk with the people you know in a better way.

I don't want to use gmail because i don't want to have a global online identity. But it depends on what you see as the primary use case (corporate? technical? casual?). Optional gmail is fine. If it's all private then maybe i'd be OK with it, but it doesn't seem like this is your long-term goal. I'm not certain how the security would work with what you are describing, i guess it requires you to be signed in to gmail constantly, but password-less option sounds good. Friction must be extremely low. Outside of gmail, make sign up just username /password.. avoid the sign-up email if you can. Think about if you really need someones email.

You will of course have to make this 100% perfect and have a lot of great ideas (but those can grow organically later) to get very far with this, IMO, but i wish you luck.

Not gmail - because I don't have one.

I'd like to see the chats be destroyed after a given period of time. People change, but their chats online don't. That's the current problem.

-- also chats can't be logged by search engine, so either blank out, or use robots.txt (ideally both)

This is Friendfeed rediscovered. It's a bit disappointing that Friendfeed is on the death row after Facebook "aqcuhired" Friendfeed.

So... IRC?

A valid question. IRC, however is old technology. There are permanency issues. Unless I have a client configured which saves logs, I can't see old discussions I had.

But, I think you do get the idea. IRC is solving the use case I plan to solve with modern technology, and by getting rid of some of its culture. I don't see any corporate meetings happening on IRC. I don't see startups discussing ideas in IRC. Nicknames, bots, strict environment, lack of ease-of-use and design, mostly limited to technical topics, no idea of multiple topic discussions(from quora). Just few of the many things here IRC stops behind to serve this particular use case.

What about Campfire then?

Nah. I see the words "Plans and Pricing" and "30 day trial" on their landing page.

Ok, I'll play: Convore?

It's gone, replaced with grove.io apparently, which costs money (30 day free trial, 10 bucks a month cheapest tier).

Right, but Convore existed and died. One explanation was "it wasn't actually popular or necessary". It would seem useful if not actually important to figure out how this is different from that service, given that that service had strong backing and even had a temporary working community surrounding it.

http://namesake.com/ might be what you want

Are you talking Campfire with room per discussion topic?

Google/Apache Wave?

The world is probably still not ready for that, maybe it will never be. Wave is too confusing. I'm still following a linear discussion model, such as seen in facebook comments. Just that here, they will be more permanent, and not casual.

The novelty of this idea is that you would be able to use this platform with just about anyone. Your email ID would be hidden, but not your name and pic(gravatar or google).

"Discussions online should be as fluid as if face-to-face."

IMHO, (or maybe "for my purposes") discussions online have been plenty fluid ever since the forum was invented. Beyond that, the quality of discussion depends on the people in it.

Also, (quality) discussion is content, and the big feature of real-life conversation is that there are either no middlemen, or middlemen who get dumped in a snap when they make shenanigans. That is a minimum requirement for anything worthwhile on the web, too, where resources are even easier to come by. If there was a single site where all discussion took place, that in itself would be reason to boycott it. My advice would be to make (or improve) apps and protocols, not websites.