Hacker News new | past | comments | ask | show | jobs | submit login
It's Time for an Open Twitter (justinvincent.com)
38 points by jv22222 on Aug 18, 2012 | hide | past | favorite | 44 comments



An alternative to Twitter has existed since 2008. It's called identi.ca

http://identi.ca/

The issue is getting people to start using such a service.


People in HackerNews tend to ignore things that have no 'funding', 'VC', 'capital', 'investment', 'startup' or related terminology.

In this very same thread people are mentioning that an open, distributed system is needed or that they'Will contribute by code and kind'.

And identi.ca is not even mentioned in the article.

Disclaimer: Im not trying to be offensive through other posters. They may simply not know about identi.ca. But that proves my point even more. Twitter and app.net have been discussed so much in the past few weeks. Identi.ca barely got any attention.


Identi.ca is not the alternative; it is a site which runs software called StatusNet.

This is like saying Wordpress.com is the alternative to Blogger. The alternative is the software which runs the site.

Most importantly, this software needs to speak a standardized protocol. Luckily, this protocol exists -- it is called OStatus.


If app.net can get the press it has gotten it probably means that the marketplace has reached boiling point. I'm thinking developers and many twitter users are ready for a more reliable and trust-able system.


How does the "marketplace has reached boiling point [with Twitter]", for app.net, differ from the marketplace reaching a boiling point with Facebook -> DIASPORA*, yet the trends speak pretty loudly (http://www.google.com/trends/?q=diaspora) that boiling point & press != success.

I'm hopeful to see app.net take off, but have been unable to find any arguments/discussions talking about how app.net will succeed where DIASPORA failed.

(Ignoring the obvious of having a better name, of course!)


Facebook is (mainly) a single web app and the market knows it as such. Real time messaging is a smorgasbord of different clients on different platforms built by different people. The distributed nature of multiple vendors marketing efforts might give it the edge.

Coupled with a lot of negative sentiments that Twitter users and developers feel might make it an interesting story in the tech press. Diaspora had a lot of press when they first started but they couldn't live up to the promise.

In this case it would be easier to live up to the promise because most of the client code already exists. It would largely be a case of clients adding new endpoints to the frontend and abstracting the rendering functions.

It would also be super easy for users to import their Twitter profile into a new system. Just enter a Twitter username and you could easily get the bio & avatar. Very low cost of transfer. If the press pushed it, and it was so easy to move over, it would have a chance IMHO.

I should also point out I don't think app.net has this opportunity, I think it could only happen for a truly free non-profit system.


If you include trustable, the system and its parent company/organization could not reside in the US.


How about we just use XMPP. You know, the currently existing, extensible, email address syntax using distributed messaging platform?

Why does nobody seem to think of this?


XMPP is probably the best approach. I'm currently investigating this path myself. The disparate set of alternative protocols that app.net is considering are the same ones that diaspora* used. The result is a server that speaks several overlapping protocols.

In order to be successful, what this space really needs is a de facto server, like Sendmail, but one that's easy to set up and configure. Over time, this would make it possible to have a critical mass of compatible servers that speak the same set of XEP extensions. There are many open source XMPP servers. I think one reason this hasn't happened already is that most of them are difficult to set up and configure.


http://onesocialweb.org/ had the same idea (and a fair amount of code!), although it seems to be a bit dormant at the moment.


Does XMPP allow you to follow a person and see all their public messages as they're posted? Does it allow you to see historic posts for a user? Can you search the network for all messages with a given tag, and get a live stream of updates? Can you see trending tags?


XMPP is extensible, so the answer could be 'yes'.

However, the actual answers are 1. Yes 2. Yes 3. Potentially 4. Yes 5. Potentially

Instead of having central aggregation, the free market would drive sites aggregating users messages to improve their features. You could sign up by just adding the site to your list.

Of course, it would require some work and a draft standard. I'm not saying it's perfect. Apparently identi.ca has a federation system too.


There exists an extension protocol for Microblogging over XMPP: http://xmpp.org/extensions/xep-0277.html

Also check out buddycloud (.com / .org)


If we're looking at existing protocols for multicast store-and-forward, XMPP requires extensions for some of what SMTP already has.


Sure, really we'd be better off replacing them all with something designed along the lines of XMPP but by a modern committee focusing on a few limited features.

The idea of it being a single non profit site seems terrible in comparison.


It would be relatively easy to build a Twitter service on top of email. Participating email providers would give every email account its own listserv account. Followers would be subscribed to the list. Tweet emails would have only a 144-character subject, no body (plus any multimedia attachments). You could still use your email normally, but also access the full Twitter functionality with a special client via the web, phone, or computer. Anyone with a normal email account could follow you, though they would have to have an account with a participating email provider in order to tweet or retweet. If a couple of the big players implemented it, the service would have critical mass overnight.


Why not build it on NNTP?


this wheel already exists. it does not need to be reinvented.

http://www.w3.org/community/ostatus/

http://ostatus.org/about/


I'm no zealot, but arguably it's always time for an open alternative to any service or product.


I like the idea of a 501c3 twitter competitor. It would be like wikipedia business-wise I guess.

Centralization has it's benefits, and they could be maintained without all the crap that's happening with twitter right now.


A 501c3 twitter is almost a brilliant solution. What grantmaker wouldn't want to support an organization dedicated to people's speech and open journalism? Valuing extensibility, openness, and free access, the mission is strong and the market is there. If Caldwell wanted to move this way, and offer $50 sustaining memberships for new features and $100 memberships for developers, nobody would blame him and all of a sudden he's open to some amazing funding opportunities.


Amazing funding opportunities perhaps, but not so great profit opportunities. Let's not forget: app.net is actually looking to make a profit (an excellent idea, if you ask me), not just replace Twitter.


A well run not-for-profit damn well better be making a profit, it's just not turned back to the owner. They just take their allotted salary. Which is a pretty fine idea too.


from http://www.thimbl.net/ "Thimbl, the free, open source, distributed micro-blogging platform. If you're weary of corporations hi-jacking your updates to make money, or if being locked in to one micro-blogging platform tires you — well, then Thimbl is for you!"

http://www.thimbl.net/faq.html "Isn't this just like identi.ca/ostatus.net?

The similarity is only superficial, in that we draw upon similarities with web2.0 platforms in how we communicate our use-case.

Diaspora, Crabgrass, NoseRub, StatusNet, identi.ca and the rest are just web-apps with some sort of federation bolted-on. None of them are truly distributed multi-tier systems like the classic internet applications such as email, usenet, irc, or finger. Thimbl, on the other hand, is just a finger & SSH client that illustrates that "microblogging" has already been possible on the internet for decades. It's done by simply presenting the data differently and making it look like the now-familiar Twitter interface. You don't need anything from the Thimbl project running on your server to participate in Thimbl, just the software that is already in your repository - namely SSH and xinetd/finger or compatible alternatives."


We need something along the lines of Diaspora (Open and Distributed). This would encourage geeks to develop on the platform due to it being distributed it would be hard/impossible to shutdown thus shall be around for ever (Twitter might not be). And due to the openness of the codebase etc people would start setting up hosts (Just like Wordpress or Diaspora etc.)


Because Diaspora has done so well.

The number of people that care about why Diaspora exists and why a Twitter clone in the same vein would exist is extremely small. Small userbase means the service is basically dead on arrival.


Completely this. Developers seem to think that the business stances Facebook and Twitter take can be solved purely with more development. They can't be. The problem is hearts and minds of users.


Actually lack of good development may be the problem. Mike Cardwell's opinion of status.net certainly isn't very high:

https://twitter.com/mickeyc/status/236796994532827136

and a lot of disaspora's momentum was destroyed by their early security issues (someone correct me if I'm wrong on this, I wasn't following diaspora that closely at the time).

Remember that as soon as even a few distributed or open systems get going they can be linked together. At that point any new competitors to FB/twitter have a strong incentive to be open as well because it gives them that precious starting network. Then, like good guerillas, you engage the centralized service in a drawn out struggle. They have to win every time, because as soon as their userbase slips it's never coming back. You only have to win once. (You can lose unlimited companies/protocols as long as more and better ones are taking their place)

EDIT: P.S. I totally stole this strategy from a Pournelle book (Prince of Sparta). "We Helots be coming back for more. Again . . . and again . . . and again." Just because they're the bad guys doesn't mean we can't use their tactics!


Diaspora was technologically stunted. It was ridiculous to expect extremely junior developers to design and build a distributed messaging system.


Are you thinking something like Wordpress meets Twitter?


Take a look at StatusNet. That's effectively a Open Source Twitter Clone.


I think non-profit and open-source might be being confused here. I'm talking about a highly publicized non profit utility style global API. To make it easy build-on and fast to proliferate it should probably be an always-on centralized service.


OStatus.


Kind of but not Wordpress as Wordpress is far too bloated for 140 char status's. I'd relate it more to Diaspora that Wordpress personally.


My best wishes. Will contribute by code and kind. And the day such a service exists, I am closing my twitter accounts.


I wonder how hard it would be to add a stalking function to an IRC server, so you can follow someone and see all their public messages; and then log messages so recent chat can be viewed by looking at a channel... are there any other reasons to use twitter over IRC? :P


Wouldn't you always have to be connected, that way?


And Automattic is the company to build it.


"How can one commercial company be allowed to own a protocol such as http or email?"

Twitter's "protocol" is not a protocol but an API that runs on top of HTTP and they created it. The ONLY responsibility Twitter has as a corporation is to do whatever it takes to make the largest profit. If it failed at this, it would actually be in violation of the law as that is the purpose of a corporation.

The world isn't entitled to anything Twitter builds.


You're missing the point amidst your rage. The author means protocol as an abstract mechanism, not a specific implementation. Email is not a specific protocol either - I can receive email using JSON over HTTP - but a abstract mechanism that is implemented by a collection of those.

And that's not Twitter irresponsibility; its responsibility is to do whatever its owners want it to, following the law. Especially since it's a private company, not a public one.

The world isn't entitled to anything Twitter builds.

Nobody says it is. The article is saying that Twitter is not entitled to be the sole mechanism for distributing Twit-like messages, and that we should build and use a different one.


>> The world isn't entitled to anything Twitter builds.

>> Nobody says it is.

I would say the choice of wording by the OP suggests otherwise.

From the OP:

>> Twitter has the power to topple regimes, but by following advertising dollars they are building a walled garden and making anti-competative (link to: http://thenextweb.com/twitter/2012/08/17/twitter-4/) decisions.

If I'm not mistaken, the OP characterizes Twitter's imposing of developer unfriendly guidelines as anti-competitive, which it is not. If Twitter was being anti-competitive, they would be preventing competing platforms like App.Net from existing - and that's not what's happening. Limiting what is essentially a "free ride" is not anticompetitive.

From the same paragraph:

>> It does not bode well for our world that such a commercially driven company should wield so much power and be the epicenter of global conversation.

As far as I know, all companies are commercially driven. Twitter wields power because people like to use it. But so does Facebook. That the author tried to "build a business on Twitter’s platform for the past 3 years" and then be upset by Twitter's change of heart suggests a little sour grapes on his part.

Further:

>> Dalton Caldwell has an interesting idea, but I don’t think it will end the Twitter monopoly

While Twitter is the market leader, it does not have a monopoly on "status updates". The last time I checked my Facebook feed, it was full of the same type of updates that Twitter provides, and Facebook is pretty popular. Google+ is also a well funded competitor to Twitter.


the OP characterizes Twitter's imposing of developer unfriendly guidelines as anti-competitive, which it is not

I don't see how such characterization, accurate or not, mean that OP says the world is entitled to what Twitter builds.

As far as I know, all companies are commercially driven.

Firstly, I still don't see how does that tie to my post. I wasn't making an apology of the article. That said, OP did use a qualifier ("such"). I'd claim that there are less commercially driven companies, e.g. Canonical.

Twitter wields power because people like to use it.

People like to use Gmail too, but if they kick you out, you can still communicate with everyone there. Twitter wields power because it's a locked-in system.

But so does Facebook.

So?

That the author tried to "build a business on Twitter’s platform for the past 3 years" and then be upset by Twitter's change of heart suggests a little sour grapes on his part.

That's called an ad-hominem fallacy. The fact that he may be pissed does not make him wrong.

While Twitter is the market leader, it does not have a monopoly on "status updates". The last time I checked my Facebook feed, it was full of the same type of updates that Twitter provides, and Facebook is pretty popular. Google+ is also a well funded competitor to Twitter.

Again, I fail to see how is this related to my post.


>> Again, I fail to see how is this related to my post.

So to requote your original post.

>> >> The world isn't entitled to anything Twitter builds.

>> Nobody says it is.

Let me put it more succinctly. My interpretation is that the OP is basically saying that the world is entitled to anything Twitter builds.

My subsequent quotes weren't so much related to rebutting you but to highlight the parts of the OP that I felt supported my point.


I understand your interpretation; I just don't see how is it supported by the article, particularly by the points you've quoted.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: