Ghosts of the dead tree

A little over twenty years ago, Edsger Dijkstra delivered an address titled On the cruelty of really teaching computing science [PDF]. In the first screen's worth of text, he raises the danger of trying to understand "radical novelties" in existing frameworks, something that we would today call "disruptive technologies."

When it comes to thinking about disruptive technology, I will be forever handicapped by my inability to divest myself from old thinking. And so I'm going to engage in the stagnant practice of "reasoning by analogy," and in doing so maybe put forth another perspective on collaborative sites.

Collaborative sites of the active kind

Around these parts, there are three major newspapers, excepting the tabloid-y one. All of them have some kind of political leaning through their coverage and editorial bias, but that has been going on for ages. Nowadays that kind of bias is part of the brand. All of the papers report the news, but the decision comes down to which one-stop shop has the best combination of news, bias, layout.

And maybe comics.

In that sense, these collaborative sites are like newspapers. There's a lot of overlap in coverage, so they have to compete on other things. And readers, hypothetically (and anecdotally), like the convenience of a one-stop shop. Readers and writers have ported the newspaper concept to the web, because it's what we're familiar with.

That's it, we're done. The concept can be advanced no further because it takes a leap to envision what life beyond a newspaper-like or magazine-like site could look like. The next step is out there already, though, and layout isn't really a part of it.

Collaborative sites of the passive kind

I've kicked around the term "vertical integration" enough so let's use another one: monolithic. Isn't that what a collaborative site aspires to be? Selecting article writers and columnists, unifying presentation and distribution, the web address demands that readers make frequent pilgrimages.

But there is nothing stopping a reader from building his or her own monolith. It's called dropping feeds into a folder, or collecting a bunch of feeds with a Yahoo Pipe, these feeds being those of individual writers. In this kind of world, everyone is a potential columnist, and delivery is on demand.

The idea that everyone is an unwitting contributor is a strange one, but it's only a step up from everyone being a knowing contributor, which is fairly common, having great success in the world of peer-reviewed publication.

Collaborative sites of the semi-active kind

The concept of the peer-reviewed site has been proposed but hasn't, at least outwardly, gained much traction. I think it's indicative of the challenges, and none of them are technological.

Here's my understanding of the process, this being derived from a conference. A conference consists of two parts: paper submissions, and paper presentations. A call for papers goes out to industry and academia, outlining topics of interest and the date and venue for the conference. Paper submissions come in from all over, where they are evaluated, primarily on the basis of novelty. Papers that are accepted are not given a strictly free pass; they may be accepted with some reservations, which the paper authors will have to address and then resubmit.

There may be some back and forth, as the authors make corrections and/or clarifications until the reviewer is satisfied. The paper can then be published in the conference proceedings, and the authors present their idea and findings during the conference, receiving additional feedback from those in attendance.

In trying to port this kind of model, the biggest challenge is the problem of reviewers. Who has the time? What's in it for them? To be quite blunt, if you might make a great reviewer, your time is better spent editing or QC-ing fansubs than it is poring over someone else's entry.

Other challenges — defining a list of topics, setting a deadline, optionally holding a podcast in place of a conference — are minor in comparison to getting reviewers. And submitters, I suppose, but I think that there are more willing authors than there are reviewers.

When all is said and done, will anyone read it? It's not something that should deter anyone from giving it a go, because until someone does it will continue to be an open question. Conference proceedings, transactions on selected topics, societal journals, and the like are in demand because it is the primary means for which the research community vets and disseminates discoveries, as well as provides a body of work with which to build on. Could other writers build upon what others publish? Possibly, but do we need the kind of vetting required for academic integrity?

More to the point, are we researchers? Hardly.

In hindsight, it's more productive to view such a venture as a creative anthology, which should provide some clues as to what kind of subject matter might garner the most attention.

Collaborative sites of the semi-passive, totally public kind

Whereas totally passive is just feeds in a folder or feeds in a Yahoo Pipe, semi-passive implies some form of gatekeeping, and I think manual gatekeeping is getting more attention at the moment.

You can already share what you find interesting through your social networking site of choice, or FriendFeed, but that's not what we're talking about. Those get the job done, yes, but I think we want the ability to present an article in its entirety.

Oh but we can do that, too! Google Reader Shared Items makes it easy to generate a page and feed containing complete entries if drawn from a full feed, but even that isn't quite good enough, because it lacks reach and you can't market it like a stand-alone site.

And it's here that we enter a minefield, because what do you call a site that copies whole entries from the feeds of others? Historically, it was called theft. And so to do it right means an upfront cost for each site that you wish to share entries from, because you have to get their permission, or inspect their licensing terms. By the way, you can view my own licensing terms in the footer.

So it turns out that trying to exercise a little self-hosted selectivity could potentially result in a lot of trouble and burned bridges if you're not careful. This dicey proposition is made more palatable if you don't try to make any money off it, ever. That way, others will be less suspicious, and you also avoid ambiguities with Noncommercial clauses.

Experiments on the cheap

A lot of what I've written here is decidedly negative, but that's because my mind is stuck in the frameworks of old, and when confronted with disruptive technology the pitfalls are much more apparent than the gains.

However, it should be emphasized that implementation is what gives an idea shape, and the monetary costs associated with the implementation of any of these ideas can be nil, at least on your end. The only sure cost is time.

Build it and they will come. Maybe.