They only did it 'cos of fame - A.P.I.
March 12, 2011 2:05 PM   Subscribe

"Developers ask us if they should build client apps that mimic or reproduce the mainstream Twitter consumer client experience. The answer is no." - Not long after saddling it's own iOS client with some unpopular new "features" Twitter is saying no to the development of new competing clients. Existing clients such as Twitterific and Echofon should be unaffected.
posted by Artw (42 comments total) 4 users marked this as a favorite
 
This is my surprised face.
posted by axiom at 2:09 PM on March 12, 2011




He also claims that the diversity of the application ecosystem and the differences in presentation between individual client applications is "confusing" for end users.

If one end user uses one third-party client, how is that end user going to be confused? He or she gets a consistent experience.
posted by kenko at 2:09 PM on March 12, 2011 [5 favorites]


is "confusing" for end users

I hear this a lot in my life. We have a large organically developed iPhone app that has some popularity, and we're embarking on and android version. I'm driving the design to a simplified - useful - UX. But no, this is potentially confusing for users - who may have at some stage of their lives seen or dreamt of the iPhone version, and may not realise they are using a completely different platform and their brains might explode.
posted by the noob at 2:21 PM on March 12, 2011 [9 favorites]


I wouldn't mind the ads, if I could turn off "trending." After the earthquake, I was trying to follow everyone's tweets, but I kept getting angry seeing GODZILLA trending. (also: really? Assholes.)
posted by ColdChef at 2:21 PM on March 12, 2011 [3 favorites]


They're throwing out the baby with the bathwater. Twitter has a learning curve by itself. It's not the difference between the apps.

I suppose I was confused by the available app clients when I first looked at them. But honestly it would just take a faq. "We have our client, others have theirs with particular features."

btw for people who follow lots of links, I really like using smartr
posted by stratastar at 2:26 PM on March 12, 2011


s/confusing/unprofitable/g

Funny how often this seems to work.
posted by Kadin2048 at 2:34 PM on March 12, 2011 [4 favorites]


After the whole #dickbar fiasco I removed my titter client. I've been using Osfoora exclusively now.
posted by cjorgensen at 2:38 PM on March 12, 2011


Of course, as power-abusing profit-maximising jerk moves go it's strangely half assed, as it leaves the existing clients intact. I haven;t decided yet if that's wimpy or our strangely cunning and insidious.
posted by Artw at 2:42 PM on March 12, 2011


This is an example of companies privatizing profits and socializing (or crowdsourcing) costs. Twitter's back-end infrastructure is built on open source components, which means no licensing costs & more plentiful and therefore cheaper developers who have experience with those technologies. The open APIs means that they can crowdsource R&D to the developer community who will thoroughly explore the problem space for free. In-house R&D is expensive, you have to pay for the failures as well as the successes, but with this model, Twitter only pays for the successes. Hundreds, maybe thousands of Twitter projects are created, a few are acquired: Tweetie, Fluther, Summize. Lesser ideas are simply stolen.

What's interesting is that developers aren't protesting to extract compensation for unpaid labor. They're upset that Twitter is shutting down certain focus areas of their crowdsourced R&D department, and the affected labor force demands the ability to work for free.
posted by AlsoMike at 3:08 PM on March 12, 2011 [7 favorites]


I haven't looked very deeply into this, but it sure does read like a really stupid move from an organization that has so far been pretty smart about not making the one-company-protocol nature of their system all that bad in practice.

Of course I guess this is where we hit the inevitable badness-in-practice of protocols that are a single entity's application.
posted by brennen at 3:11 PM on March 12, 2011 [3 favorites]


Twitter could have become a decentralized medium for communication, something general like SMS or e-mail. Instead, Twitter decided to become just a website. This post by a former twitter employee sums it up best: they decided to emphasize Twitter with a capital T over the concept of twitter. I'm kinda surprised that the API was so robust for so long.
posted by The Devil Tesla at 3:15 PM on March 12, 2011 [9 favorites]


Not at all sure this does leave existing clients untouched. They specifically mention they'll be monitoring them closely to look for ToS adherence, which I read as "for pretexts to shut down just like we did those clients last month".

After the whole #dickbar fiasco I removed my titter client. I've been using Osfoora exclusively now.
You and thousands more. And now they're going after third-party clients. Funny, that.
posted by bonaldi at 3:21 PM on March 12, 2011 [1 favorite]


I think they want to see if they can crash this bubble as fast as they possibly can.
posted by stratastar at 3:28 PM on March 12, 2011


I'm the author of a .NET Twitter library; I don't write Twitter clients, but a lot of people write Twitter applications using my library. I love the ecosystem, but the people that Twitter has chosen to interact with the development community are ungreat.

This is a really disappointing development.
posted by DWRoelands at 3:49 PM on March 12, 2011 [1 favorite]


The whole 'dickbar fiasco' is exactly what prompted this. When they talk about 'unified experience', they're not really talking about retweets having different names on various clients or different implementations (because, seriously, what major client does?). What they're actually saying is that their new profit model (the dickbar, promoted trends, potentially promoted tweets, etc.) simply won't work if the third-party clients aren't forced to create / user these features too. The changes in Terms of Service spell it out as such.
A shame, since a lot of the innovation (hashtags, for one) seems to have come via third-party development.
posted by secretdark at 4:26 PM on March 12, 2011 [1 favorite]


The older I get, the less crazy Richard Stallman seems.
posted by enn at 4:29 PM on March 12, 2011 [21 favorites]


What they're actually saying is that their new profit model (the dickbar, promoted trends, potentially promoted tweets, etc.) simply won't work if the third-party clients aren't forced to create / user these features too.

And, of course, a big attraction of third-party clients is that they don't have all the ads / promoted tweets / other bullshit that's making capital T-Twitter less and less useful and interesting.
posted by immlass at 4:35 PM on March 12, 2011 [1 favorite]


Existing clients such as Twitterific and Echofon should be unaffected.

<ominous thunder rolls in background, russm squints and flicks eyes left and right>

for now...
posted by russm at 4:43 PM on March 12, 2011


Ian A.T. Likes Artw's Post Title

Whoops, wrong social network, my bad.
posted by Ian A.T. at 5:19 PM on March 12, 2011 [2 favorites]


Fuck Twitter. Morse at 15 wpm should be enough for anyone.
posted by clvrmnky at 5:20 PM on March 12, 2011


Googles "dickbar"

OH, that thing. Yeah I hate that thing.
posted by emjaybee at 5:20 PM on March 12, 2011


Ian A.T. Likes Artw's Post Title

At least someone remembers the classics...
posted by Artw at 5:35 PM on March 12, 2011


Screw this, I'm switching to Pownce.

What? Why are you looking at me like that?
posted by entropicamericana at 5:57 PM on March 12, 2011 [3 favorites]


The roots of the move are understandable, but it's still troubling. I'd be less worried by a "all third party apps must show promoted tweets" policy than the move they're making here.

I don't begrudge them the need to be profitable, but this is the wrong way to go about it.
posted by dvorak_beats_qwerty at 6:11 PM on March 12, 2011


What they're actually saying is that their new profit model (the dickbar, promoted trends, potentially promoted tweets, etc.) simply won't work if the third-party clients aren't forced to create / user these features too.

And, of course, a big attraction of third-party clients is that they don't have all the ads / promoted tweets / other bullshit that's making capital T-Twitter less and less useful and interesting.


Maybe this just demonstrates that a centralized twitter isn't viable. They built a huge network, became part of popular culture, spent years trying to monetize, and this is the best they could come up with. Meanwhile, people are perfectly willing to pay for quality mobile twitter clients.

Maybe the Appleseed, diaspora, and OSW guys are going after the wrong target. Instead of a decentralized facebook with people sharing pictures, what the world really needs is a decentralized, peer to peer twitter.
posted by heathkit at 6:25 PM on March 12, 2011 [1 favorite]


*Young person in the middle of Fuscia Revolution, checks phone*

Dickbar: "Trending: Security Services Cracking Down on YOU"
posted by stratastar at 6:37 PM on March 12, 2011


Meh. Twitter isn't google or facebook where you can shoehorn in a few billion dollars of ads without really disrupting the rest of the service. Twitter's got bills to pay.

I never understood how these clients apps expected to build real business on top of twitter in the first place. Twitter isn't windows, it's a slightly better version of MSN.
posted by GuyZero at 8:03 PM on March 12, 2011 [1 favorite]


Ugh. I've been really hyped about twitter in the last few months. I think it works remarkably well in certain areas - like The Devil Tesla says, it could have just been a general universal service and done that well - but it's actively breaking itself more and more. To throw a hot-button topic into the mix, I was going to get my twitter name inscribed on the back on my new ipad when I ordered it, but now I'm not convinced I'll still be on it in a year.

It all leaves me wondering if a social media service can maintain itself if the users are the customers, but aren't paying. I'd shell out $5 for twitter, but a lot of people wouldn't, and then I'd have no one to follow.

Flickr seem to do fine with pro accounts, but twitter's really got one function, and I don't think it can be tiered in the same way.
posted by roobot at 8:11 PM on March 12, 2011


what the world really needs is a decentralized, peer to peer twitter.

so, usenet with a 140 character limit?
posted by russm at 8:22 PM on March 12, 2011 [4 favorites]


I really *hate* the official Twitter clients. I've tried almost every twitter client but I always come back to Echofon. No other client syncs unread tweets between devices, and Echofon was the first one to support push notification.
posted by mike3k at 8:41 PM on March 12, 2011


We need to move to a less fragmented world, where every user can experience Twitter in a consistent way

No. Nononononononononononononononononononononononononononono....
posted by brundlefly at 8:44 PM on March 12, 2011


I don't think it can be tiered in the same way

Maybe they could offer a cheap dickbar-less service?
posted by immlass at 8:47 PM on March 12, 2011


Actually, if twitter itself supported multiple accounts--like hootsuite--it could offer pro accounts.
posted by crush-onastick at 9:12 PM on March 12, 2011


I am rather fond of tdash, myself. Very lean (completely web based) and seems to work a treat for me. I've been getting all rather Cloud fascinated lately.
posted by Samizdata at 6:16 AM on March 13, 2011


Imagine if they'd asked for a small contribution from each user; they could have pitched it as an annual, voluntary fee that nice folk pay.

Instead, it's dickbars for everybody.
posted by The River Ivel at 2:39 PM on March 13, 2011


what the world really needs is a decentralized, peer to peer twitter.

Back a few years ago when Twitter was first getting picked up outside the true early-adopter set, there was sort of an ongoing argument (read: dick-measuring contest) in some circles as to whether it would be possible to re-implement Twitter "in an afternoon," or some other arbitrarily short amount of time.

The best proposal I ever heard put forward for doing this essentially involved cheating; rather than trying to rebuild Twitter the way Twitter is actually constructed — which is (or was) basically like a big database-driven web site — instead build it on top of email tools like Postfix and Mailman. E.g., to "follow" someone, you would use a client that would send a specially-formatted message to their Twitter address, which would just be a special email address; software on their end would parse the message and add you to the list of followers. When they posted an update, it would be sent out to everyone on the subscriber list. If someone's server went down, their updates just wouldn't get pushed out; nothing else would be affected.

Twitter handles a lot of messages, but compared to the number of emails sent in a day, worldwide, it's still not that much. But email isn't handled by one big server farm somewhere, or even a bunch of server farms operated by one entity, it's operated by anyone who wants email service for their domain ... and the result is a system that's proved to be a lot more scalable and reliable than centralized services like Twitter.

Unfortunately, even though you'd only probably be talking about a bunch of shell scripts sitting on top of an off-the-shelf free software stack, good luck getting anyone to adopt your new system, and that's really the crux of things: IMO, Twitter was successful because it's a really easy service to pick up and start using (same with Facebook). Building an architecturally superior message service is trivial, but getting people to use it wouldn't be.
posted by Kadin2048 at 9:11 PM on March 13, 2011 [6 favorites]


what the world really needs is a decentralized, peer to peer twitter.

so, usenet with a 140 character limit?


Well... yeah. It'd keep people from flooding the service with binaries, at least.
posted by heathkit at 10:26 AM on March 14, 2011


Nothing stops people from flooding binaries.

Man, I gotta go write that base64->twitter encoder.
posted by GuyZero at 3:58 PM on March 14, 2011


Now Biz is working with Aol. http://www.bizstone.com/2011/03/biz-and-aol.html
posted by bonaldi at 7:12 PM on March 14, 2011




Kadin2048 - it scares me that my first response was "what are the downsides to that, and how can they be mitigated"... I'll leave out the engineering discussion, because it's sorta boring...
posted by russm at 12:37 AM on March 15, 2011


« Older Harmony harmony oh love!   |   So therefore, OPEC felt justified in raising the... Newer »


This thread has been archived and is closed to new comments