Twitter's odd use of the word 'followed' | Hacker News Reader

Twitter's odd use of the word 'followed'

blog.jgc.org

101 pointsjgrahamc21 comments

I suspect that Twitter is keeping a list of people I followed in the past [..]

Being a mere sentence fragment makes "followed by X"'s meaning ambiguous. Twitter can have their cake and eat it too, as "followed by" could be either present or past tense, despite what we'd pragmatically expect it to mean in this case. I agree it's a bit weird, although it seems like one of those instances where Twitter wouldn't care since it probably provides users with some benefit anyway.

So, I'm not sure why Twitter thought @plan28 would care who @jgrahamc follows

I've noticed Twitter doing some interesting things in this regard. I create new Twitter accounts in a separate "clean" browser to avoid logging out and I often get myself suggested as a follow. They must be assuming people within either the same IP range or who are geolocated to a similar location would be interested in each other.

I think it would be simpler than that - I have not checked but I expect I follow both @jgc and @plan28 and if there are enough then somewhere a k-value will be high enough that suggesting a follower of one should follow t'other will work.

The geolocation idea is a nice one though - I will try and add it to my bag of tricks CMS - cheers

Geolocation used that way exposes something like a location ("is nearby") without those users even having a choice or being aware. I'm not saying that's what Twitter does, but if it did, I wouldn't find that nice at all.

The death of privacy is not nice. But it is still dead, but the corpse is warm so hard to notice.

Wait till governments notice what they thought was secrecy was actually privacy. Total freak out time.

As someone who’s done a fair bit of translation on Twitter’s crowdsourced translation dealio (https://translate.twitter.com/), I can say that Twitter’s deliberately ambiguous use of language (for various reasons, e.g., as here, avoiding any explicit specification of tense, as well as of gender, number, &c.) can be very difficult to parse in a precise and accurate manner, especially without enough context.

I was wondering ... what exactly motivated you to dedicate time and effort towards improving the service of a $US10 billion valuated company for free? I'm not trying to be sparky or anything, it just baffles me. Is there any reward structure in place or interactions with other translators?

Translation is one of those things where crowdsourcing makes sense - there's not always a 'right' way to translate a particular phrase, so a consensus between multiple people helps.

Agreed. Although in my experience it can go badly wrong... http://opensignal.com/blog/2013/04/26/an-apology-to-our-kore...

Wow, some people are really immature.

Although this seems like a good example of why, when crowdsourcing translations, you want to get multiple people doing the same language so you can compare their results.

> Wow, some people are really immature.

One man's immature is another man's hilarious. In this case I side with you, but I'm pretty sure if they'd come up with something wittier than "All Koreans are stupid" I could find myself laughing at it.

I think that, regardless of what they write, intentionally defacing someone else's site, especially in a way that they're unlikely to be able to easily detect, is rather immature.

Wait... they crowdsourced the translation of their Terms and Conditions? That seems like, I don't know, a bad idea. For user content or interfaces sure, but T&Cs are essentially a legal document.

Well, mine’s the case of a minority language (Irish) where, even on the off chance that some professional outfit were paid to do the translation, it’d more than likely be of very low quality. Basically either (a) it wouldn’t get done or (b) it’d be really bad, so in the end I’d rather just do it myself, even for free.

"Because of this it's worth thinking of things you do on the Internet as public and irrevocable."

It is prudent to assume that collected data is immutable and not time stamped. Not time stamping data creates value.

My niece has visited my house. Saturday junk mail arrived addressed to my niece "or current resident". A local business paid for that unique lead and for the cost of a stamp. I knew it was junk from afar.

The company selling my niece as a datapoint can probably point to a plausible chain of reasoning showing that my niece once, however briefly, dwelled here. Social networks, smartphone apps, etc. all indicate it.

Twitter's utility for users is a by-product of its purpose. Its purpose is to hit the sweet spot between activity and ruthless monetization. [edit] example: If jgc once clicked on a tweeted link to an article from Time Magazine, then it is plausibly legitemate to say he "followed on Twitter".

I was at a talk which explained that twitter has recorded every user event since it was launched e.g. Follow, Tweet, Unfollow. They calculate their metrics by processing every event in bulk from start to present. So yeah - they never throw any information away.

Because of this it's worth thinking of things you do on the Internet as public and irrevocable.

It's worth knowing that things you do on the Internet are public and irrevocable.

It's worth knowing that things you do are irrevocable. And often more public than you think.

It doesn't seem very mysterious. I guess the UI requires that something is in that box, and to satisfy the edge cases where the user doesn't follow anyone, they fill it with other criteria, and use the 'followed' label a bit loosely.

Just Semantics. Twitter is technically correct by terming "followed by," instead of "follows." As we all know (and many celebrities and gov officials) with Twitter, a delete button doesn't really exist.

The guy didn't delete his account, and is super surprised they keep old data. What did he expect?

Reply on news.ycombinator.com