A new bit of analyst data crossed my radar this weekend, this time from the annual Evans Data survey of over 400 Linux developers, "OSS/Linux Development Survey 2010." While I haven't read the entire report, the one bit of information that Evans is pushing from the report is, naturally, the most controversial.
It turns out, according to the report, that nearly two-thirds of open source developers do some of their non-work related open source project work on their employers' time.
Whether that figure is accurate or not, I expect we'll see a whole new line of FUD coming soon from those software companies which are less-than-enthused about open source software: open source projects steals your employees' time and reduces their efficiency.
Evans itself seems to be spinning it that way: their CEO, Janel Garvin, was "shocked" to see these results:
"I'm not sure if this is shocking to anyone else, but it sure is to me. Yes, developers get wrapped up in their projects and it's hard for them to leave them alone to concentrate on their employer's projects, and yes it's true that it's hard to police developers. If they're sitting there writing code it's not that easy to tell what that code actually goes to. But doesn't the fact that the employer is paying them count for something? Perhaps not in the new world order where believers think all code should be free," Garvin wrote.
Perhaps she should read her own report. In the excerpt available online, the report itself deemphasizes the impact of this mixing of "personal" project participation while developers are punched in to the metaphorical time clock.
"Not that it matters really, because sometimes the developer is merely taking extra steps to fix a bug or enhance a module that ultimately will aid the employer’s goal as much as the developer’s private goal," the report states.
And that is certainly the first thing an open source developer should likely say if they were working on an outside project while at work. Many projects they work on can have direct effects on their employer's code, and even indirect: figure out how to solve a nasty problem in an non-work project, then they have the skills to solve a similar problem later on for the employer.
Evans seems to be going for the headlines, though, since the other piece of data they released for public consumption late month seems to enforce the meme that open source developers are a bunch of freeloaders. According to their data, 65 percent of open source developers make less than US$100 annually for their open source work. The implication is that since these coders aren't generating their own revenue, they must be leeching off their employers' dime.
Looking at the data, it is easy to draw these kinds of conclusions. It is also equally easy to come up with the notion that developers are coding for their employers to earn their paychecks and are working on open source stuff just for the heck of it. And while it is troubling that many coders are doing personal stuff on company time, I wonder if the report makes allowances for the opposite situation: how much personal time is used working on employer projects? Given the state of the economy, and how much work is piled on employees of all types these days, what percentage of these developers can actually finish their work projects in 40 hours/week? Or 50? Or 60 hours?
Given that kind of scenario, it's little wonder there's a blur between personal and professional projects. If my employer is asking me to work so much, I wouldn't feel too bad taking a break during work hours to get something of my own done. As long as deadlines are being met and my efficiency is on the mark, then is this something the employer is really worried about? If they're that worried about how I spend my time, then I should just give them my 40 or 50 hours a week (depending on my contract) and just stop.
The other question I would propose is how much do proprietary vendors spend working on professional versus personal projects? I would be very surprised if the percentages of work vs. non-work effort in the workplace were nearly the same for that group, too. This is not, I suspect, an open source developer problem, but a larger time management problem between creative developers and employers who are trying to squeeze more effort out of them.
Of course, my scenario doesn't sell as many reports.