Follow Us on Twitter
by hjanssen on January 25, 2007 11:50am
Whether deploying Windows in a Linux or UNIX environment, or vice versa, identity management can be a challenge. In order to explore this topic, Hank spent some time talking with Dustin Puryear author, consultant, featured speaker at TechX World and owner of Puryear Information Technology, LLC. Dustin specializes in solving integration challenges related to mixed environments and has penned a few books on the topic. Two of his books: Best Practices for Managing Linux and UNIX Servers and "Spam Fighting and Email Security for the 21st Century" are available as free e-books on his site.
In this podcast (Part 1 of 2) Hank and Dustin explore issues related to implementing Identity Management solutions and focus primarily on the importance of policies. While the act of deployment is by default important, without proper policies even the best solution can fall short. In the next podcast the conversation will focus on implementation strategies. If you have particular questions you'd like answered in the next podcast, please feel free to comment below.
by Bryan Kirschner on January 26, 2007 12:45pm
I started this chain of blogs about the law-and-open-source–analogy based on something Matt Asay had written that struck me as interesting—but didn’t sit with me as quite right. So it seems appropriate to tie up this set of blogs with something he wrote that seems to me to be entirely right, and helped frame a comparison about law-and-open-source that makes a lot of sense to me.
I previously blogged about the fact that legal knowledge actually seemed more “open” than “closed.” I also kicked around the idea that the supply of lawyers was especially “artificially” restricted. Without rehashing those discussions, I didn’t find those to be compelling as singular ways the legal domain was different from the software development domain.
Then I read a recent blog from Matt. Blogging about companies for whom it is an “open question” if they “are abandoning some of the community benefits of open source by having some of its technology proprietary,” Matt wrote that such companies are “trying to balance being a good community citizen with getting paid. It's not an easy problem to resolve” (my emphasis).
I read this statement as characterizing what a lot of companies and individuals—open source or not, in the software business or not—are trying to do. Specifically thinking about the legal analogy, here’s what struck me.
Among friends, family, and colleagues I happen to know a lot of lawyers (probably a result of both my wife and brother being lawyers rather than a character flaw on my part (lawyer joke)…). Thinking of this balancing act Matt describes, I realized that:
One lawyer I know, a litigator at a firm, recently made a change to a corporate (non-legal) job, in which he’s making about the same as he had been practicing law. In both cases, a pretty respectable overall salary.
Another lawyer makes literally half what he did as a litigator, working in a public (government) agency.
And one more lawyer I know makes literally a third of what he did, working for a non-profit doing public interest law.
And of course, lawyers always have the option of providing their services at no charge (pro bono)—the American Bar Association recommends 50 hours per lawyer per year.
(Qualitatively I would have to say intrinsic job satisfaction among the lawyers I know is inversely related to total compensation against these four cases. The assumption here is that, for most people, being a “good community citizen” is in itself at least somewhat rewarding.)
Additionally, individuals have broad latitude in the US to represent themselves (pro se), generating a “do it yourself” supply of legal services.
For the sake of plotting everything on two axes, you could plot the total units of legal service delivered (y-axis) against the payment received by the “deliverer” for that unit (x-axis)—in this view paying someone else would be a negative payment while DIY would be a 0 payment. (I don’t know exactly the shape of the curve here…but it actually doesn’t matter much for the sake of the thought experiment.)
Now let’s think about open source code with a similar plot in mind. Lakhani and Wolf’s open source developer study found 40% of developers were being paid while developing code. So on the chart above they somewhere in quadrant 1.
Cutting the data a different way using cluster analysis, the authors show that 80% of the sample is essentially covered by work need/payment (cluster 1) or two clusters where “intellectually stimulating” and “improves skill” lead by a long shot as motivations among predominantly unpaid developers (see page 13 of the linked document). This is code in quadrant 2 of the chart. “Obligation from use”—that is, giving back, something that seems like a commitment to giving back through pro-bono work—is the far-and-away leading motivator in the fourth cluster.
If we plotted all the units of code—open and on top of it not-open—in the world, would the shape and motivational dependencies look much different from all the shape of all the units of legal service?
In quadrant 2 (no payment received by deliverer) in either domain would be a lot of things that “scratch your own itch” or are themselves stimulating and rewarding. In quadrant 1 would be whole lot of things for which these characteristics apply less or not at all (or perhaps inversely—they are painful or tedious or otherwise unrewarding…).
This brings me all the way back to the question of law, software, and “openness.”
I took issue with the idea the primary “problem” with the legal domain was it wasn’t “open” enough. I think the process of thinking through comparisons between law and software illuminates the fact there is no single magic bullet for increasing the supply of something—whether code or legal services—at lower cost. (Note I’m taking a macroeconomic and long-term view here: although I work for a company that sells software, in the grand scheme of things more-for-less is probably good for commercial interests and consumer interests alike; neither I, nor, I hazard a guess, whoever defined that product strategy for Microsoft feels “bad” that tens of millions of end-user developers can DIY hundreds of millions of lines of code using the Excel macro engine and Office object model. That product substantially “shifted the curve” and was a win-win overall.)
Making more knowledge—through exposure of raw artifacts like source code or judicial decisions—or pedagogical materials can shift the curve. So can increasing the supply of people who are trained to be developers or lawyers. So can decreasing the difficulty of delivering a unit of service or code (XNA is really interesting me in this respect….). So can make making delivering some unit of service or code more intrinsically rewarding (fun, tear-jerking, intellectually stimulating, guilt-assuaging….). So can making it more financially rewarding to deliver some unit of service or code.
Balancing being a good community citizen with being paid is a hard problem to solve because there is no one single formula for striking the right balance. But I find that to be a good thing: that translates, to be sure, into many opportunities to try to strike the right balance and fail—but also many opportunities for diverse individual, communal, and corporate co-exist, adapt, prosper—and surprise. A world where one size does not fit all may be messy but it is a lot more interesting.