Quick note: Please stop trivializing collaboration!

If we believe our masters at the Googleplex, collaboration is about seeing each other's keypresses.

(I'm only choosing this example because it's the most extreme, and thus most clearly shows the trivialization of collaboration.)

I don't know what collaboration is really about, but I know that it's not about seeing each other's keypresses.

Take scientific collaboration. Some good scientists (like Gödel) work decades and only publish a handful of papers. These papers are their inputs to collaboration. Seeing Gödel's keypresses during all these years would be kinda boring, and probably not really helpful.

Collaboration can only work if people do a lot of great work offline. There, I said it! Collaboration is actually hard. It requires thinking, formulation, etc.

So don't trivialize it. Please.


context-matters said...

As a scientist I must say that being told that my collaborator is working on the previous work section can be useful because then I can avoid making changes to it and then avoid the merge conflict that will arise when my collaborator commits his changes.

In general though, our collaboration is asynchronous and I not interested in seeing keystrokes, I am mostly concerned in what he changed because I have to make sure they didn't cock it up.

The fact is that change introduce errors so knowing about change lets me check the most relevant part of the paper.

So what do we use to collaborate?

If I'm lucky, git. Less lucky, svn or cvs. Even less lucky, I get sent PDFs marked up with annotations or a fax or a scan.

Yet another case of context.

Craig said...

You might enjoy this 2007 OOPSLA podcast of Fred Brooks on Collaboration and Telecollaboration in Design (mp3)