Daily Archives: November 20, 2009

I do not like retweets and spam. I do not like like them Spam-I-am.

My friend Elisa asked me what I thought of the new RT feature of Twitter, since, as she tweeted, I’m an “add value” kind of guy.

Here’s what I think so far: I’m not a big fan. I don’t think I’ll use the feature much if I can avoid it.

But I think I understand the issue that Twitter has on its hands. I just finished writing an article on a security panel composed of federal CISOs last week, where the topic of the secure use of social media dominated the discussion. The conference was on application security, which meant that standards for authentication, identity and verifiability were elevated in importance.

As a DHS security official put it, “one of the fundamental issues with the Internet is that sending a trusted message is difficult. We don’t have any kind of mechanism like that for rapid fire social media communications.”

Trust but verify has been the watchword for me whenever I have decided to RT something for years  now. Is the link safe? Who had control of the account? What other accounts were associated with the tweet and how were they annotated? Tricky, especially in the real-time.

Twitter’s own terms of service are crystal clear about its Impersonation, Trademark, and Terms of Service policies:

“Re-posting other’s content without attribution and without their permission is a violation of Twitter’s rules. Accounts that re-post others’ updates (with or without crediting the author) may be immediately suspended because

  • Re-posting others’ updates as one’s own without giving credit to the original author is tantamount to plagiarism
  • Re-posting or others’ updates, regardless of stating authorship, is a potential form of spam.”

And yet, we have the RT, which reposts the content of others.

Tricky, no? That’s at least part of the reason that some folks have said that a  RT = Spam.

I think the RT caught on and stuck around because it gave attribution and offered a simple, user-generated mechanism for users to share the content of others using microsyntax. I’ve argued that a RT is social media currency.

In aggregate, many retweets of the same message can build both trust in the content and validity of a link or quote. That implied veracity, given by those phatics of trust, can be in turn dangerous. Messages that went viral and contained a link to a malware-laden page or to a phishing syndicate are like viruses circulating around a Twitter biosphere, given the stamp of approval by the brands of trusted senders.

The number of RTs a message receives has also been one of the most useful metrics used to measure the rapid growth of social messaging on Twitter. If you haven’t read danah boyd’s draft IEEE paper on the RT, “Tweet, Tweet, Retweet,” it’s worth your time if you’re interested in how digital ethnography applies to Twitter.

Earlier this week, Twitter rolled another community generated convention into its code with a new RT button. First came the @reply, which was replaced by @mention. Now, we have retweets. Sort of.

(What follows is a conversation I had earlier this week with my friend, Ed Shahzade, on first impressions.)

When @leolaporte tweeted that “Twitter would have been better off implementing its “retweet” feature as “like.” That’s all it really is,” I retweeted him, adding “Spot on” before the RT.

Ed @replied to us: “That’s so far off. It’s republishing, not putting a star on the only copy. It’s like adding to a friend’s newspaper.”

“I disagree,” I @replied. “IMHO, the coding isn’t how the community has used RTs. Implementation amounts to a positive rating, removes annotation.”

“Yes, my chief complaint is lack of commentary(option),” @replied @Ed. “They could have listened better. Bear in mind- fraud prevention was a goal.”

“I understand,” I @replied.

That was because Ed is right, after all. Verifiable messaging is an issue as old as human communication. It’s a serious issue for use in crisis response for government. There are unscrupulous criminals and bored teenagers at play in the system.

But the way that this was implemented may really upset the community and users can still be bad actors under the current system. I’d rather see some kind of system where the highest rated tweets can be ranked by meta juice from likes. Xeni Jardin recently tweeted that she’d like a “spam” button IN the tweet, next to reply, RT and Favorite.

I’m not sure about that feature, given the potential for abuse, but a Flag might be helpful as a way of adding a thumbs down or “dislike” feature. That “dislike,” after all, has cropped up on Facebook, along with the “Like” feature that’s so familiar to Friendfeed users.

Perhaps a RT could be attributed in API and still allow the retweeting user to appear inline? The attribution for tweet sourcing is built into the current code. The issue is that substitutions of those that we do not follow in our timelines are jarring to many.

Part of Twitter’s allure, at least for me, lies in having control over your timeline. Follow, unfollow, block, spam, RT, etc.

I don’t like the new RT but I understand that better trust and authentication mechanisms need to be built in to prevent fraud and ensure better online security for Twitter. I do think an audit trail is key, so that users can easily follow RTs to an original and compare if they like. I do that constantly if I need to verify who said what when. And I like to annotate the tweets.

The reason that it’s important to get the coding right on spam, security and authentication, however, doesn’t rest in my usage. I’m an IT journalist, after all. I find Twitter useful, educational and frequently entertaining but I can live and work perfectly well without it. And, to be blunt, I know it will survive without me.

Many enterprises block social networking platforms like Facebook and Twitter precisely because of such security concerns, along with resilient questions about productivity.  Earlier this week, I reported on a study that surveyed senior IT security executives at federal agencies. Outsourcing, an increasingly mobile workforce and cyberterrorism topped the lists of their security concerns, along with unstructured data resulting from social networking platforms and Web 2.0.

If Twitter is really going to be a “global information utility” for government private industry and nonprofits, security and trust will be at issue. So will privacy. Twitter updated its privacy policy to “explicitly include geotagging and to describe the public nature of most of what people post to Twitter.”

For Twitter to effectively compete and thrive, adding features that don’t map to how its community uses them may not be in its best interest. Since I’m a long-time user and have found it useful, I hope that the feedback they receive is replicated in the code.

1 Comment

Filed under Uncategorized