Hacker News new | comments | show | ask | jobs | submit login
Dark Social: We Have the Whole History of the Web Wrong (theatlantic.com)
142 points by vectorbunny on Oct 12, 2012 | hide | past | web | favorite | 48 comments

It's called "social web" for a reason.

Instant messenger and chat rooms and ICQ and USENET forums and email are not web. At most the OP is talking about the history of the Internet.

{The World Wide Web (abbreviated as WWW or W3,[2] commonly known as the Web), is a system of interlinked hypertext documents accessed via the Internet. http://en.wikipedia.org/wiki/World_Wide_Web}

That was my first impression too but I think his point is that the web was already social because people were already using (and still do) out of band mechanisms to socially share web links. The new social web is just using the web itself to stream those previously OOB messages.

This is in contrast to saying that before the social web people used the web individually, looking up information by themselves but not interacting socially about the web.

That is retconning "social web" to cover every kind of interaction. Applied hindsight-bias. Before the web, there were boards on services like CompuServe, The Source, AOL, and others, that simply had names that you told other people. It's not a new technology to email someone you're going to the mall when before email you would just tell them over the phone or whatever.


The point isn't that people were being social over email. The point is that people were using email as a social layer on top of the web (i.e., they'd email links around). What we now call the "social web" is just the web as a platform taking over functions that before were performed over other internet technologies like email and IM.

People weren't using the social web before the likes of Facebook and MySpace. You, me and the rest of HN were. But everyone else? No, they were not on forums. Being social on the internet and the web was the purview of geeks. Friendster, MySpace and Facebook changed that.

Uggggh. No. No.

Yahoo!, AIM, IRC, whatever came before phpBB, Delphi Forums, ICQ, even Livejournal.

These things were pretty common, not just among geeks. I chatted with a 60+ year old grandma I worked with at a department store on ICQ in '98 & '99. All my buddies had AIM, yes, the non-geek buddies.

I know a carpenter active on a carpentry forum in the 1998-2002 era.

And we were in rural Idaho.

All of my friends had AIM as well. But that was not true of people even five years older. All of those things that you pointed out existed, but they were not the kind of assumption that Facebook is now. Web-based social sites starting with Friendster, then MySpace, and now with Facebook opened up online interactions to most people in the Western world, as opposed to the geeks, or family members of geeks, or the young, or people with special interests.

The point is not that people were not social on the web before. The point is that many, many more people are social on the web now, and that it's becoming an assumption of society. And from what I have witnessed, this is clearly a result of social networking sites.

You're basically making a non-refutable argument since anyone who responds to you here is going to be an HN commenter whose experience is disqualified by your premises.

At least provide some data to back up your claims.

> that it's becoming an assumption of society.

It was an assumption of general online society in the 90s, as near as I can remember.

What exactly is the argument here? Were people (non-techies) using the web/internet before Facebook/MySpace? Of course. Are considerably more people (non-techies) using it now after Facebook/MySpace? Of course.

That's a function of mobile and broadband, not facebook et al (arguably)

I don't think so. Anyone that was serious about a particular area, such as a hobby, profession, or social cause was already using forums 10 years ago. Now these people converged to the mainstream sites such as twitter, but they have always shared information on the web since it was created. My wife, for example, doesn't know a thing about computers, but is more active in forums and niche sites than I have ever been.

I know many counterexamples, particularly people who were in their 20's, 30's and older ten years ago. MySpace and Facebook was their first introduction to the social web.

I follow a beekeeping forum, and just yesterday there was a thread about some people noticing they had been on for 10+ years. These are definitely not the young hacker crowd. That predates myspace and is literally just an example that came up yesterday, I'm sure I can find older forums.

And are you really saying that before web forums, all the way back in the very beginning of the web people weren't being social on the web by sharing links through email? Remember, email is the true killer app of the internet, the web is just a nice-to-have. :)

I'm talking trends, not individuals.

He acknowledges this upfront. I'm not sure why he persists in using this terminology since he does know better. I think that he is correct so far as there is equivalent functionality in the older internet social tools and protocols.

Pedantically speaking, Usenet and email at least are part of the web. They aren't http-based, but the "web" is really anything with addressable URIs, and both usenet and email have well-defined URI formats, eg:


I disagree with you, and so does wikipedia: "The World Wide Web (abbreviated as WWW or W3,[2] commonly known as the Web), is a system of interlinked hypertext documents accessed via the Internet."

Just because a browser supports those protocols, doesn't mean they are part of the web, which uses HTTP.

You can't really like email documents or accounts to each other, so I don't think they're part of the "web" of connected pages.

You can link nntp news articles.

Even the social web dates back further than most people think. For instance, Livejournal's been around since back in 1999 complete with friends lists and communities and privacy settings for posts and all sorts of social functionality, some of which the mainstream social networking sites didn't even adopt until recently.

He may have focused too much on icq and email, but the point is still valid. Forums existed in the 90s. They were websites.

Forums did more than exist-- they were massively popular. Sure, I don't remember a forum with 1 billion members, but I'm still boggled at the colossal output of some of them. I remember pretty niche topics still with large, competing boards and real-life meetups... all before Zuckerberg even hit puberty. I think the author's point is that the web didn't just magically come alive in 2004, and this is more than true.

I was just saying to someone that we had IM on IBM mainframes (the chat facility) or the (confusingly named) Ethernet on DEC VAXes back in the mid 80s.

What we used them for was 'are you going for tea?' and other social chit-chat between the various terminal rooms up the Uni...

I just discovered the 'kibitz' command on linux, which is basically remote desktop assistance for linux (well, remote terminal). Apparently, this is used for chatting quite often too (i.e open up a text editor).

EDIT: screen can do this too! screen -x can attach to a non-detached screen session.

Can't you just use talk(1)?

In a quaint bit of online archaeology, Douglas Hofstadters' Metamagical Themas describes his experiences using some of the early Unix communications tools, particularly talk (predecessor to IRC) and the dynamics it triggered. They're very, very similar to what emerged later as IM and SMS became prevalent.

Modulo the rather smaller share of users in possession of PhDs in mathematics, computer science, or other technical disciplines.

yep- just installed ytalk on my Debian dev box for this. Works great!

This is one of those topics where you have to be careful not to slip into meaningless semantics where you're just arguing about how we should use some word and what things should be named.

By their logic the "dark social" is ANY inbound link not coming from "Facebook, Twitter, Reddit, YComb, Digg, Stumble" or do I get that wrong? If so, well, that is called the "Internet". It uses hyperlinks to send people between pages on it.

That's the point.

I think we can all agree that site owners are interested in "social networking" as the mechanism by which their content is shared and discovered by people passing links around. Essentially, the propagation of "word of mouth".

But they've been tracking data based on a definition of "social networking" that only includes the big social networking sites.

And that when they broaden the definition to "this traffic is the likely result of social networking" to "someone has directly requested a deep link without referrer data", they see that even the sum of traffic from all those huge sites is just the tip of the iceberg.

Which comports what we all already know. Which is that Twitter/Facebook/Reddit/et al didn't invent any new social behaviors. They just aggregated large chunks of them.

No, any link with no set referrer. That exactly not stuff from other pages.

No, it is any inbound link to a page unlikely to be typed manually "blah.com/articles/random/crap/article-title-goes-here/35487623", that has no referrer header at all. Traffic coming from other websites will still have referrer headers, just traffic from email, IM, etc don't.

I believe any referral traffic from an https:// site will NOT send the referrer header. It does not seem like the article is taking that into account.

He did mention that, but of course there's no way to figure out how much of that traffic is from https referrals. Of course, I wouldn't be shocked if 90%+ of https referrals are from gmail, which would put it right back to what he is assuming: email, IM, etc.

You make a pretty good point, but not just gmail. Pretty much all -mail providers use https when users access their web interface.

Oops, someone (the author) forgot about AOL, which was founded in 1985 and began mailing floppy disks to people in 1993 (http://techcrunch.com/2010/12/28/aol-floppy-disk/). That is when and where I had my first social web experience (it was in an AOL chat room, specifically). The company and the "social web" boomed from there, so much so, in fact, that eventually AOL bought Time Warner. It was the largest merger ever, at the time (http://money.cnn.com/2000/01/10/deals/aol_warner/). And it flopped, but that beside the point :)

+1 for tldr inserts on The Atlantic articles. Always have found their articles meaningful but verbose.

+1 for immediately thinking the same thing

The title could be better. What he's describing with the term "dark social" is not the "Web", it's the internet. That's a distinction that is important to understand. There is much you can do over the internet that is difficult to do over the "web". Moreover, the "web" is still quite centralized in the sense that there are a disproportionate number of servers to clients. It has been called the "calf-cow" model. Some are calling for an end to that.

I vividly remember working on SunOS (later Solaris) and SGI machines and using ytalk, irc and gopher all the time before WWW. WWW made us all want to occupy graphics workstations more instead of vt320s, ha! Even though most of us were in graphics programming at the time, we were working on vt320s, imagine that.

I think he's confusing the web and the internet.

AIM, email, USENET etc. are not really web. Unless you have web clients.

I think, most people confuse the internet with the web, so maybe the essence of the article should have been that the internet is "social" even without Facebook, Twitter etc. wich would be still a very similar statement.

I use few "social" web links, maybe 1%. He fails to mention bookmarking. Most of the traffic I generate comes from bookmarked or search engine URLs. (I may or may not be unusual in that I surf with referrer turned off.)

This piece reads so much like the Salon.com piece in 2004: http://www.salon.com/2004/03/09/deep_web/

I'd say this is a flaw of the "web" and relying on hyperlinking and "HTML forms" as a means to access all content. That's what web crawlers do.

Better would be putting databases on certain known ports, with no "hostname" requirements and let search engines scan IP ranges for openings on those ports. You send a ping to the port and if a "Welcome" response comes back, then you have a public database to explore. The response might even describe the contents of the database. (Yikes, I'm having flashbacks to gopher.)

What's funny is that "port scanning" is like the term "hacker" it has a negative connotation even though there's nothing inherently devious about it. Pinging a port to detect if there's a public database open on it is actually a more efficient less resource intensive means of service discovery than crawling hyperlinks in HTML generated by some web developer (maybe even one employed by a governemnt agency with a tight budget) hoping to stumble upon every possible data resource.

This is why app developers need really well-thought out sharing features - preferably based on URIs.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact