"So, have you written any modules yet?"

This is something I overheard a couple of times while sitting in the hallway between Drupalcon sessions doing email. Despite the danger of extrapolating based on too-few data points, I think it's fair to say that writing some kind of custom module (from scratch) is a "rite of passage" for an aspiring Drupal developer.

Yet the question is: what does "Yes" mean? It certainly means you got something done. But how good was it? Did it follow good Drupal coding practice? Did it do something novel, or implement something old in a new way? Was that way better? (In whose opinion?) What if the person/organization for whom you wrote those modules precludes you from distributing those via drupal.org (assuming that person/entity is being properly GPL-compliant), and you can't actually "show your street cred" by checking in those modules for community observation; can you still prove you "did it?" And is there any objective measure by which anyone can answer those questions?

One way is to have your peers tell you. Peer review is the best. But, if your module is somewhat obscure, you might at best get 3-5 people to comment on it. And either the person or the volume of comments is likely - in most cases (IMHO) - to be insufficient to really say anything about you or your work.

This is why I'm a fan of certification. If a certification program is well-developed, it serves as an objective measure of the answers to those questions.

It is not the only measure. Certainly everybody knows that chx can write modules. But for "mere mortals" who need to have some way to demonstrate their 'cred, we're working now to invest the time, money, and backbone required to create a sustainable program that earns its credibility among Drupalistas.