I'm still not completely clear on what's ok and what isn't.
They seem to be saying that special-purpose clients would still be ok. For example, a SXSW iphone app that filtered and ONLY showed tweets with the #sxsw tag, and perhaps appended the tag to your own tweet. Not a path to riches, but you can see why such an app may be of use in certain situations. It seems to be allowed under these rules, but I'm not 100% sure.
What they've done is introduce uncertainty, which is worse for the ecosystem than anything else.
Yeah, I think Twitter fucked up this PR pretty badly. I get that it's time for them to make money, and that changes are necessary to that end. Yes, some devs are going to get screwed in the deal, that's a matter of course, but the last thing they should be doing is spreading FUD about their own platform, which is exactly what this announcement was.
I get that they were trying to be transparent, but that's greatly overrated in this case. API guidelines should be crystal clear. There's no way for them to have made them crystal clear from the beginning given the rapid iteration of the business plan. So what Twitter should have done is just iterate the guidelines along with their business plan. Sure it's gonna piss people off when changes come, but there's no way to prevent that without freaking everybody out in advance.
Well, based on this article, anything is allowed. It's just a matter of can you monetize it, knowing that Twitter is focusing on its own client as a core product. An SXSW app doesn't seem too monetizable to begin with, but no reason why you can't make it.
Having said that, I am confused as to what happens if you do make a new competing client. Is this really just friendly advice to developers to not compete against Twitter, or will Twitter block your API access?
Not at all. This sounds more like friendly advice to developers making it clear that they plan to compete in this area. Because theirs have the benefit of being the official apps, you may have a harder time gaining traction.
The writing was on the wall with the Tweetie acquisition and newtwitter, but now there's a blinking neon sign there.
I think this is great advice from Twitter and developers should follow it.
Twitter is totally correct--if you haven't entered this space by now, and you still want to, you are probably overlooking much more interesting opportunities.
Why would you develop a new Twitter client anyhow? You might make $5 a seat from a few thousand users who will expect you to maintain your product in perpetuity.
Seems the signals are mixed about what is ok and what is not ok, and people are going back and forth on that.
More importantly, what happens when something is not "ok", and for what level of ok-ness?
1. A client that spams people's Twitter streams without asking? Obviously not
ok, loss of API token.
2. A client that alters people's stream to include affiliate links? Presumably
not ok, loss of API token? (Questionable)
3. A client which "enhances" people's stream by making semantic clarifications
to content? Example: Imagine if http://twitter.com/english50cent could be
programmatically generated from @50cent. Is this not ok? It's certainly
innovative. What are the consequences? Where is the line?
4. A client that displays the Twitter stream verbatim but comes up with a new
visualization? Sounds like this is not ok too, since it might "confuse"
consumers? Do you lose your API token for this?
5. Generally, someone builds something that Twitter didn't think of and gains
mainstream usage, what is the consequence? What is the risk of Twitter making a
policy against it and shutting it down due to inconvenience to them?
The Twitter team invented the medium — 140 character limited messages with a very specific style of graph-based visibility.
The rest of Twitter's mechanics were invented by third parties. @replies were not invented by the Twitter team, nor RT's, nor hashtags, nor search.twitter.com (acquired), neither was the Twitter iPhone app (acquired), neither were url-shorteners (t.co copied tinyurl, bit.ly, etc).
Many of these mechanics came specifically from third-party consumer clients. Someone came along and made a client where words prefixed with a hash were searchable on Summarize, and now Twitter's entire business model seems to be based around that.
What would have happened if Twitter said "don't make any more clients, we got this" before all this happened?
What's going to happen if Twitter says this now, before who-knows-what happens?
Twitter's greatest strength has been in recognizing value and innovation created by the community of users, but suddenly this changed. I really hope Twitter reverts their policy/suggestion, and apologizes to the community. Right now it comes off as very arrogant, presumptuous, and stifling to innovation and further progress.
Fair point. Yet another example of the original Twitter team being able to find value and capitalize on a constraint that today would probably be ignored or even frowned upon.
90% of users are already on the official client (see the link to the announcement in the article) so the "fragmentation" argument doesn't hold water. I'm guessing this is the nice letter asking devs to stop developing competing client apps and the next letter will be more forceful.
Reading between the lines here it looks like their official client is going to be making them money in some capacity so they don't want people using 3rd party clients that might bypass that mechanism (whatever it is: ads, selling click data, whatever). That's just a guess.
Also notice their TOS is getting stricter. You cannot display data from alternate services alongside data from Twitter's API. Again, that only seems to make sense from an advertising perspective. Future restrictions are hinted at, too.
I used to wonder why the hell companies like Twitter would operate without any solid plans for making a profit, incurring huge costs and just taking on a massive amount of users. In hindsight it looks like you create an awesome free site to get a bazillion users and then once everyone is entrenched you can start to make it suck more in order to get paid.
>Also notice their TOS is getting stricter. You cannot display data from alternate services alongside data from Twitter's API.
I read it a couple of days ago, but isn't that limited to not being able to provide alternatives for the suggestions for who to follow, alternative trending topics, etc?
Can anyone explain why Twitter even matters beyond app discovery?
If I create an app that posts to Twitter so that someone else can consume that data, why not just cut out the middleman after the initial announcement?
Are you suggesting that the important tweets are the automated tweets by non-twitter-specific applications? This is bizarre to me, because I typically only follow people that are tweeting interesting things per se. Discovering apps almost never happens; to the contrary, people sending a lot of automated tweets get unfollowed by me rather quickly.
I dunno. It may be that I can derive more from a longer blog post, with more information, than an SMS that imparts nothing useful to me. For in-depth information, it's nearly useless and I, for one, don't need/am not interested in a running train of thought.
I do think it would make a great discovery aggregator, now that it's up and running.
Just yesterday I was unable to access twitter from my university. It blocked me by saying something like 'API calls exceeded'. A university have generally a single IP address and thousand of users. What will twitter do about this ?
They seem to be saying that special-purpose clients would still be ok. For example, a SXSW iphone app that filtered and ONLY showed tweets with the #sxsw tag, and perhaps appended the tag to your own tweet. Not a path to riches, but you can see why such an app may be of use in certain situations. It seems to be allowed under these rules, but I'm not 100% sure.
What they've done is introduce uncertainty, which is worse for the ecosystem than anything else.