I suspect it will always be a largely subjective measure mostly determined by the respect or accolades of one's peers or the satisfaction of one's target audience/customers.
I think StackExchange's "reputation" system is probably the most accurate discrete measure available.
Reputation on SE is very, very dependent on the amount of time you put in and your choice of which questions to answer. It's also somewhat zero-sum in a crowded field, as you need attention to get upvotes.
(I should know, I'm one of the 50 people on electronics.stackexchange with more than 10k rep, and no formal qualifications in electronics)
For me, StackExchange's 'reputation' is inversely correlated with that.
Don't want to attack the contributors, but honestly, the higher levels of reputation just either show me that he either has too much free time outside of work (valid, but doubtful) or they spend work time answering questions, which for me goes against the 10x productivity programmer mantra.
I think StackExchange's "reputation" system is probably the most accurate discrete measure available
I would be very hesitant to rely on any metrics coming out of SE. Consider their great programming survey this year, they say the average age of a programmer is 29. The US Department of Labor says it's 49.
There are a vast number of people out there, well over 90% of the programming industry I'll wager, who come into work at 9am, do good, solid work all day, on applications that run the real world, then go home at 5pm and get on with their lives, who never interact with SE et al at all.
>I would be very hesitant to rely on any metrics coming out of SE. Consider their great programming survey this year, they say the average age of a programmer is 29. The US Department of Labor says it's 49.
That doesn't mean they're both wrong. The two are just comparing different things. The SE programming survey included 157 countries, not just the US. It also doesn't only include full-time professional developers. 13.6% of respondents classified themselves as "Student". Only 66.3% of respondents listed themselves as "Employed full-time"
I think StackExchange's "reputation" system is probably the most accurate discrete measure available.