". . . Twitter Platform primarily serves Twitter’s interests, in stark contrast to the era of API growth . . ."
and his rueful close in "I believe that Twitter as a medium is and should be distinct from Twitter as a business."
Twitter got to where they are based on open standards, dev empowerment, etc.
Thanks to their money in the bank and lack of need for a biz model they had ample time to own it, be the standard, become the medium - what is a Twitter exactly if not a Twitter?
Going to be very interesting to see where this next step takes them.
Well worth reading the whole thing. It’s not too long, and he’s already Readability’d it for us, if you will.
But if you don’t have time:
• #NewTwitter will make Twitter more mainstream, and will provide advertisement opportunities.
• #NewTwitter is a shift away from being a platform towards being a destination (or “place of consumption,” if you will), and as such further complicates relationships with third-party developers.
• Twitter should become decentralized, but it’s clearly in Twitter’s best interested not to.
• Alex circulated a document internally advocating decentralization, and executives disagreed (understandably).
Ultimately Twitter is a business that needs to generate ROI for its investors.
The open API gave them plenty of ideas for functionality and "holes" in their system, and Twitter has made it clear that they need to fill those holes.
The multiple points of entry also showed them that they needed a seamless user experience, which is why they launched their own native Blackberry app and acquired Tweetie, launched an iPad app and now relaunched their web interface.
Many of the twitter.com upgrades were built by someone else on some other site first, but it is in Twitter's own interest to use those ideas to get more traffic to twitter.com and ultimately use the traffic to generate revenue.
It is sad to think that Twitter fostered its ecosystem only to strong arm it and chop it down when it was ready to figure out its monetization strategy. But those are the perils of building a business off of someone else's platform.
As a Twitter application developer myself I knew this was a risk, but I have become painfully aware of it over the past few months. I think their current trajectory will eventually kill off a good part of the ecosystem that helped them get to this point.
His comments on the difference between Twitter and Facebook are particularly interesting:
"Previously, developers took data out of Twitter and into the context of their own applications and services. The new design flips this on its head, bringing rich embedded content into the site from a host of brand-name web properties. (It’s worth noting that Facebook has done much the opposite: they started out with a very centralized build-it-within-our-walls model, then gradually grew their tendrils out into the web with Facebook Connect.)"
I'm curious how this relates to his later observation that "walled gardens" are doomed to fail. In both instances, the walls are being chipped away, but in the former, the content is being brought in, while in the latter, content is being let out.
This is a nice piece, and makes some very good points.
Completely unrelated: I enjoy reading Alex's writing in the same way I enjoy reading John Gruber and Paul Graham. He's eloquent, thoughtful, and passionate.
Disclaimer: I never "got" really Twitter. The service still strikes me as mostly useless.
Now this could just be me but he says:
"Twitter needs to decentralize [...] my belief that all communications media will inevitably be decentralized, and that all businesses who build walled gardens will eventually see them torn down."
So I heard we had this decentralized way of sharing which they call "the web" which is completely decentralized and without walled gardens. And we had these technologies called RSS and ATOM for receiving updates on those.
So...he's effectively suggesting we go back to the web and RSS but with everyone writing shorter messages?
I think they did this by providing a consistent format (both data format and by enforcing limits), a central view and a usable api. RSS/ATOM just didn't get this right. Twitter forces people to get their message across in a small amount of characters. RSS/ATOM again is inconsistent in mainly due to it trying to be a portable medium (reading full stories in a reader). I guess the twitter model is more natural?
Perhaps there is a place for adopting a twitter based format for future rss feeds and making viewers around this.
I personally like the idea of sites having a dedicated twitter account to post updates to, I don't use rss feeds but like getting information via twitter.
I don't think Twitter can ever really replace RSS/ATOM.
I have a handful of sites I follow (as in really read every update) and Opera having integrated RSS I get a nice reminder popup whenever there's an update. For me Twitter wouldn't work for this as the amount of incoming messages on Twitter is usually to large and it would be distracting.
To be honest I don't really see any problems in the RSS system. I don't really see why enforcing a data format or enforcing limits makes it more usable? I kinda like getting comics delivered straight to my browser's RSS feeds and being able to read blog posts directly in there as well.
Now Twitter would make for an exceptional social RSS where I could find one-of articles related to my interests. Which I guess seems kinda similar to FriendFeed (which I've never used since no one I find interesting is on there) or Hacker News (but without the discussion and a slightly different topic selection).
The main problem I have is that people keep insisting on posting completely uninteresting tweets (such as what they're doing) or using it as a conversation tool. This kind of noise tends to drown its use as a social RSS feed for me.
The method I described would only have the same amount of traffic as an RSS feed as it would just be updating headlines.
I guess it just depends on use. I never got into using an rss reader but follow a bunch of projects/people/companies on twitter and can quickly browse their updates, and follow a shortened link. I like the portability of twitter over different platforms too.
Other than a few techies, no one understands or uses RSS. In fact, even among most of my techie friends, I'm one of the few who actively reads RSS. It just doesn't work right now in terms of "normal" people being able to use it.
And anyway, saying "we already have the web, why do we need Twitter" is wrong. That's like saying "we already have cellphones, why do we need SMS". Different kinds of information tend to have their own strengths and weaknesses.
Don't get me wrong, I'm not saying "why do we need Twitter?". While I consider it a useless service there are obviously a huge amount of people in disagreement with me, which would imply that it serves some purpose. What I am saying is "Why would we need a decentralized, open Twitter?"
As far as RSS being to complicated goes, that (in my eyes) is a failure of browser authors then. Excuse my disconnect from the mainstream browsers, since I've been using Opera exclusively for about 10 years now. But for me subscribing to an RSS feed is a matter of pressing the orange RSS icon in my URL bar, which shows me a clear overview of the history of the feed with a very large subscribe button at the top. The end result is that the feed shows up in a list under (and looking exactly like) my mail filters (sort of like the folders in most common mail clients) including an unread count and notification for new ones, which seems about as simple as it gets.
The difference is in synchronization. If I subscribe to a Twitter feed, I get the contents of that feed no matter how I access my Twitter stream. With RSS, I have to make sure that my feeds are identical in all the browsers I use. Bookmark synchronization helps, but only for desktop browsers - I can't synchronize bookmarks (yet) between my desktop and my smartphone in a totally transparent way. Even with bookmark sync. taking care of synchronizing your subscriptions, there's still no way of synchronizing read items between instances. So items I read on my lunch break at work won't be marked as such when I get home.
I don't think that is the problem. It's more a problem of explaining the concept and having good readers. Even I, when I started reading feeds, spent a long time trying to find a reader and setting it up (Google Reader, by the way).
Contrast with Twitter - easy to explain the concept, 5 seconds to sign up and get started.
There is just a tremendous difference in usability between RSS and Twitter.
RSS: find the icon (displayed differently on every website), follow a multi-step process to sign up (minimum of three clicks/choices), open another application which you've never heard of before to read this information. Or, click the orange icon on your browser, which is not prominent and is totally undescribed and undocumented. Same deal: minimum of three clicks/choices.
This should be a huge lesson for all developers. You made some shit, call it RSS or Atom, that does exactly what Twitter does, only it's hard to use. People sort of use it, once they can figure it out. Twitter came along and made it one-click to follow Brad Pitt. There are great-grandmothers and 2-year-olds using Twitter, but none using RSS.
The only service I see close to what he's imagining is status.net, check the video, it is almost exactly what he is talking about.
As a heavy user of twitter, I really don't see microblogging as a much needed service that needs to be decentralized. I just want the latest news and links from the web designers and programmers that I follow. #newtwitter seems like it would make the user experience more enjoyable for someone who has leaned on to that standard type of use of twitter.
I also don't worry for developers, as long as twitter doesn't keep track of my unread posts or lets me filter my feed, they have a niche.
I think what Twitter added on top of blogs was just the following/followers lists. They were there to an extent with the blogroll, but blogrolls are very cumbersome to use.
Also, it integrates the reader with the publishing tool.
Is it just me but isn't it a bit too early to say Twitter is trying to wall itself in? Are they taking down the APIs?
From what I can see here they just want Twitter.com to be a viable (and profitable) business. Why wouldn't anyone want that? In fact, I think it's about time Twitter finally stepped up to define just what kind of a service Twitter really is about. If anything it would be a nice stepping stone to introduce to people who don't yet get it's structure to try it out.
Yes, there are third-party developers who have created their own platforms and in a sense helped define the medium. There are many such choices, but really many of the functions they provide overlap. Do we really need that many different Twitter clients with minute differences? If Twitter were to provide the basic functions itself, then I'd say that this is a good time for third-party developers to think up new ways to diversify themselves. Wasn't there just not so long ago someone posted a HTML5 tower defense game up here that has a Twitter scoreboard? Why not expand on that?
It's an interesting question, one that Twitter is at the center of now but one that's not limited to Twitter: Where do we draw the line between a technology company and the culture its technology has spawned?
Twitter.com is not a protocol, it is a service. I think it is also important to note that it isn't a public utility either.
right now it is dangerous to place all your eggs in a single basket. When you only have one service provider you lose control of things such as dependability, namespace, access, etc. These are things that most may be willing to let someone else provide for them, but some may want to control these and other aspects of microblogging. A federated design with multiple independent service providers would go a long way to help fix these shortcomings.
". . . Twitter Platform primarily serves Twitter’s interests, in stark contrast to the era of API growth . . ."
and his rueful close in "I believe that Twitter as a medium is and should be distinct from Twitter as a business."
Twitter got to where they are based on open standards, dev empowerment, etc.
Thanks to their money in the bank and lack of need for a biz model they had ample time to own it, be the standard, become the medium - what is a Twitter exactly if not a Twitter?
Going to be very interesting to see where this next step takes them.