Personally, I think it's the only long-term solution to this bilateral matching problem: engineers can't find good work, employers can't find good engineers. The problem is that most software engineers are indeed unskilled (and therefore few companies are willing to start interesting projects that might make use of skill) because high-quality work is so rare. It's a vicious cycle and open-allocation is the only way I can see of breaking it.
If they're intrinsically motivated, engineers are worth millions per year. If companies left them alone and were able to capture even 10% of their true potential, they'd do better than any closed-allocation shop.
There's an indirection that occurs whereby engineers apply to companies-- not specific projects or work sets-- and (until they've been at the company for a while and can change jobs without embarrassment) relinquish their right to choose what they work on. (A certain once-great search giant is notorious for abusing opacity.) They have no control over how they start in the company, and if they're not especially socially skilled, this will determine what kinds of projects and opportunities they get in the future. This indirection gives a lot of power to middle managers, who don't have the vision or inclination to design and allocate work in a way that would enable engineers to have real careers and improvement, and the result is an industry where stagnation, burnout, and middle-age obsolescence are the norm.
I have mixed feelings about open allocation. I think it's very dependent on the context and the company's vision. @ChuckMcM wrote something very interesting explaining two types of engineers, some who are self driven and have an agenda and the ones who don't. I don't think that as engineers we are particularly good at understanding the business value we can bring to a company and how to improve a product. I think that some engineers work better with a lot of freedom, but most don't. I think that actually a good solid layer of "proper" management is way more useful. A good manager is someone who doesn't get in your way but help you realize your full potential while following the company's vision.
Granted, that's not what most managers do, but I do think that most of us need some leadership, some directions to help us focus. To summarize, I think that we need strong leaders and some flexibility at the engineering level dependent on each individual temperament, the company culture and the team's objectives.
This is a good response, and I generally agree with your synopsis of what a good manager is.
One of the misconceptions about open allocation is that it's a "free-for-all". It's not. It's more like an applied research environment where people are free agents to some degree, but still are expected to follow-- to contribute to something sanctioned-- if they aren't ready to lead yet. Open allocation companies still fire people, but it's very rare, because "intrinsic low performers" (as opposed to people who lose motivation due to dysfunctional environments) also turn out to be very rare.
Open allocation doesn't mean, "You show up and do whatever you want." It means, "We expect you to contribute to the business but it's your right to choose how you do so."
My problem with traditional management is that old-style managers aren't really the leaders that the workers themselves would select, but puppet leaders, chosen from above. Also, when companies give managers the right to unilaterally fire people, or to damage their transfer packets through a globally-visible performance review system, they allow managers to turn into extortionists who force people into subordinating to the manager's career goals rather than doing what's best for the employee or the company. Since middle managers also get a lot of pressure from above in these sorts of companies, that's usually exactly what they become, not because they want to be extortionists but because their feet are to the fire as well.
I think that "management" as a cultural institution is archaic. Leadership, communication, and mentorship skills will always be in high demand, but that's not what a manager is. A manager, from the term's roots, is someone hired by an owner of recourses (capital) to maximally exploit them. People often get angry with their bosses when the manager ignores their career goals, but they shouldn't, because that manager is just doing his job. He works for the company, not the employee.
Managed environments lead to localism, by which I mean the extreme trust (e.g. the right to unilaterally fire people) placed in managers allow an environment where the manager's stated goals (which are often neither aligned with the corporate objective or the employee's needs, but the manager's career objectives) take priority. You have lots of indirection and wasted effort and, usually, morale problems as a result.
Open allocation, on the other hand, trusts people to work directly for the company. There are still leaders and followers, but the employees are implicitly trusted in deciding whom to follow.
http://www.quora.com/Software-Development-Methodologies/What...
Personally, I think it's the only long-term solution to this bilateral matching problem: engineers can't find good work, employers can't find good engineers. The problem is that most software engineers are indeed unskilled (and therefore few companies are willing to start interesting projects that might make use of skill) because high-quality work is so rare. It's a vicious cycle and open-allocation is the only way I can see of breaking it.
If they're intrinsically motivated, engineers are worth millions per year. If companies left them alone and were able to capture even 10% of their true potential, they'd do better than any closed-allocation shop.
There's an indirection that occurs whereby engineers apply to companies-- not specific projects or work sets-- and (until they've been at the company for a while and can change jobs without embarrassment) relinquish their right to choose what they work on. (A certain once-great search giant is notorious for abusing opacity.) They have no control over how they start in the company, and if they're not especially socially skilled, this will determine what kinds of projects and opportunities they get in the future. This indirection gives a lot of power to middle managers, who don't have the vision or inclination to design and allocate work in a way that would enable engineers to have real careers and improvement, and the result is an industry where stagnation, burnout, and middle-age obsolescence are the norm.