| This is a crazy interesting concept. In part because I just never considered it before, but also because its kind of a midway: Let the elitists feel elite and set the standards, and the rebels and riff-raff fight amongst themselves for status. Honestly, having worked in media (film) and with those guilds (SAG, DGA, etc) its very difficult for me to say this is a good idea. The guilds were constructed the way they are because of the in-and-out nature of working in production. Without going into detail, the model just doesn't fit unless all devs turn freelance. Now a union, as the states has for many different labor professions etc: this is a concept that could work, but without an effective standardization system (something VERY hard to do in a world of open source, fast moving technology development, and new coding languages) it would be for show and simply an opportunity for the 5% who shook the right hands to sit on their golden thrones. Sounds very American though. ;-) The thing is, neither guilds or unions were really built to enforce standards of work-quality. They were built to protect the workers against potential abuse from corporations/foundations/government firms/private sector contractors. So this would be inverse to the problem we're trying to solve, which is qualification of good devs versus "bad" or under qualified. So what would work? Well if we all started by doing Oracle IT and Microsoft certificates, the furiously patient devs of the world could at least recognize each other for it. Actually... they already sort of do. :-p I think ultimately the next form of standards is going to have to be written by the academic world OR some other government sponsored party of incredibly savvy individuals who's entire "business" revolves around standardization testing for devs. Not that there aren't companies kind of doing this, but well? And on a government level of respect and reach? Nope. I suppose the only back draw is that it risks removing the creativity and how problem solving abilities are associated to personality type. I'm sure you could eventually formulate that as well, but sheesh, what data mayhem. Would be awesome.
I guess one point for it is hell, if google and amazon can do it for themselves, who's to say evaluations couldn't be built across the board? That was really long, sorry, but super interesting question! |
Kinda surprised that you've not heard it before, it's knocked around the professional programming circles for a while now. Anyway.
You're quite right that it's not a trivial problem. IEEE promulgated the SWEBOK, which the ACM publically refused to support (I read through the SWEBOK, it's, imo, generally irrelevant).
It's also not a question of creativity and problem solving, it's a question of quality assurance - i.e., the downside of being in such a group would be the culpability for being sued for malpractice. :-)