Hacker News Reader

Comment by mmastrac | Hacker News Reader

I've heard from people at Google that the reason they don't want to open up posting over the API is because they don't want people's streams filled with robot cross-posts from Twitter, Facebook and zillions of other services. I happen to agree with this point of view.

View on HN

Sounds like throwing the baby out with the bathwater to me. I have a hard time seeing how one can justify avoiding providing something that users of a service are desperately eager for, and which would undoubtedly help grow the ecosystem around said service, all because of one potential negative consequence. And especially one that, at the end of the day, people can simulate anyway by just posting the same content (by hand) to G+, Facebook, Twitter, etc.

Key words: "by hand" -- posting spam by hand doesn't scale. That's what it's preventing.

I have a hard time seeing how one can justify avoiding providing something that users of a service are desperately eager for, ...

How do people know what g+ users are "desperately eager for"?

Have there been studies? Polls?

I'll bet Google's done the studies and polls. And I'll bet they are seriously considering the response: "We don't want API-driven auto-junk filling our streams!"

"We don't want API-driven auto-junk filling our streams!"

Nobody does, but - again - we're talking about throwing the baby out with the bathwater here. Not everything posted via API is "junk."

Not everything posted via API is "junk."

True, but enough of it is to make it a problem. What could Google do to trap the noise but allow the signal? Rate limiting? Pattern matching against Twitter posts?

OK, I don't mean to say that every G+ user wants this, mind you. But that a substantial number do is pretty obvious by the posts to the ticket about this, and observing how people interact with other, similar social networks. Does anybody really think that people wouldn't like to be able to use TweetDeck (or whatever) with G+? Or programatically interface with G+ from their blog, or whatever? I may ultimately be proven wrong, but I'd be shocked it a proper study did not show strong demand for a write API.

Here's an example of why they don't want to turn it on (this is one of the examples they use internally).

https://plus.google.com/107158560758466172135/posts

Charlie Kindel uses a backdoor to paste his tweets to G+ and adds very little to the community. There is no interaction on his posts - it's just added noise.

There is no interaction on his posts - it's just added noise.

Fine, then the people who find it to be noise can un-circle him. For the class of people who are largely interested in the same topics as Charlie, it isn't even noise in the first place. Honestly, I don't see this as much of an argument against allowing the Write API.

Really? Sure, I don't find his content appealing, but I don't think that has anything to do with the manner in which he posted it. Replace all those tech articles and RT's with links to bacon (or zombie) related content and Mr. Kindel has a shot at being voted president, no?

The content might be slightly unappealing recycling of tech news, but the manner in which it's posted is what really hurts there. It'd look crappy anywhere, but especially so on Google+ where posts are long form and links are rich.

Like he apparently had some original ideas about the naming of Windows RT, represented in 4 posts each ~100 characters long (one of which seemed to be an "end of split post" marker). It looked the worst of both worlds. And argument so abbreviated it had been reduced to gibberish, but looking very verbose due to having been artificially split.

Although... those 4 posts about Windows naming could actually be an argument FOR a write API. If he wasn't using a hacked up backdoor, then maybe he'd have an app that could have handled that in one post. But I do see both sides of this. It just sucks when the few idiots ruin it for everyone else. It'd be nice to post via API and it would be nice if people didn't abuse it. ;)

The point is that broadcasting the same message verbatim across every medium and community. removes one's ability to tailor the message for the standards of that community.

Not every message needs to be tailored. And not every message need be the same across every medium. Having a write API doesn't force anyone to do either of those things. It would not give or remove anyone's ability to tailor as needed or broadcast verbatim as desired. Having a write API would allow 3rd party apps to add G+ to your options but doesn't force you to act any way you would not already act without the write API. My point was that having a write API doesn't have to be bad... but some abusers will ruin it for the rest.

I came across the service that cross-posts to G+ by accident. The funny thing is I can't even remember its name right now!

I couldn't turn it off if I wanted to (and I don't; I have no desire to "engage" on G+, but the narcissist in me likes knowing as many people are seeing my posts as possible).

I turned it on as an experiment to see how the people who have "circled me" on G+ would react.

The reaction has been mixed. On one hand VicG used it as an example of what not to do. A few others have made snide comments.

On the other hand, I've picked up hundreds of new circlers and gotten some interesting (to me) comments on some of my posts/tweets.

I have no desire to "engage" on G+, but the narcissist in me likes knowing as many people are seeing my posts as possible

And there is a best argument against adding a write API to Google+

I have a page for my podcast. I want my podcast's web-site to automatically post about new episodes to G+ the same way it does for Twitter and Facebook. If people don't want it, they can just unsub.

Maybe they should allow api-posting to pages, and not to users.

But what is worse: A stream full of reposts, or an empty stream? It's each user's resonsibility to choose what to post (or repost) or not... Google could just give lower priority in the "filter" to reposted content.

Google Buzz had a lot of import sources available at launch. There was so much noise from automated cross posting that I didn't even know that several of my friends were actively using the service until months later.

In contrast, Google+ is almost completely signal for me right now.

Yes, I was waiting for someone to bring up Buzz. I was also a first-day user of Buzz, and guess what? They allowed third-party import, and very quickly out-of-context Twitter updates overwhelmed my stream. Then there was a backlash, and the smarter people turned them off rather quickly. But the vast majority of people kept on using it. I think Google knows exactly what they are doing here.

Not all the G+ streams are empty, so that dilemma only really has one horn. Bot-posting would raise the noise floor so as to swamp the current (small) signal. Better to attempt growing the signal itself before adding in a bunch of noise that the users will have to sort through.

Now, if one could separate off each user from their bots... that would be neat. Maybe default to only circling a user, but show that they have some bots which one can also subscribe to, if one wants.

They have already allowed certain product vendors access to the write API - more details here - http://adwords.blogspot.co.uk/2011/11/third-party-tools-to-h...

This is exactly what happened to Google Buzz.

Reply on news.ycombinator.com