Assisting Drupal Community Growth

Both Dries and I have agreed on one key thing with each other since our first conversations about Acquia: 18 months after we start our business, we truly hope the community at large thinks Acquia has been a great thing for the Drupal project.

Issues presented by Drupal's expansion

One of the challenges the Drupal community must to be concerned about is growing the number and quality of Drupal developers. Drupal downloads and drupal.org content are both doubling annually. This rapid growth leads to a key bottleneck: there are (on average) more people looking for Drupal-qualified developers than there are people who are available or qualified to do it. As long as Drupal's sustains its current growth rate, this problem will continually worsen.

This has several potential "bad" outcomes:

• New developers who stumble could give Drupal a black eye from which it may be difficult to recover. A failed project may not have been Drupal's "fault," but it's easier for a novice Drupalista to blame the software, since the software can't defend itself.
• Drupal's reputation could suffer from weak functional implementations as it is tried in new usage scenarios. Insufficient growth in the quality or quantity of developers will prevent it from excelling in these new uses, which could (again) result in an opinion-backlash about Drupal's capabilities and greatness.
• Competing alternatives will leap on any perceived "slowdown in growth" if Drupal doesn't keep up it's doubling-every-year momentum. Maintaining this momentum was easy when Drupal doubled from 1,000 to 2,000 Drupal sites. But it takes a lot more people out there who know Drupal in order to double from 250,000 Drupal sites to 500,000.
• Drupal's quality may not grow in lock-step with its quantity. Among Drupal insiders, there are more than a few under-the-breath admissions that some of the code contributed modules in Drupal are of dubious quality. If this trend continues, it will become harder and harder for each new release of Drupal to come out on time, and for new users to find modules that work reliably and meet their needs.

Train and build

Therefore, a key priority for us at Acquia for helping the project / community is to build a rigorous training and certification path for new Drupalistas to follow. During the code sprint day at Drupalcon, Robert Douglass was interviewed about his role in helping create this plan. This is why we hired him: Of all people in the community, we think Robert is best qualified to do this. He was co-author of the first book on using Drupal (and is currently contributing to another), and his contribution to community education & growth is plainly seen from his extensive library of tutorial articles he wrote during his work at Lullabot. Watch the video here. Please note that our program is targeted at building both training and certification. I want a lot of our attention focused on the training part; this will have the biggest impact on the issues listed above. And we know that with Drupal's rate of change, we'll have to develop a process that results in a fairly "living curriculum." This isn't an easy thing. But Drupal isn't the first technology that has needed training while the tech is rapidly moving. It's been done successfully before, and we'll try to learn from the best to replicate it for Drupal.

The case for certification

But certification needs to be a component, too. I will acknowledge that I've read, and considered the discussion about the certification topic in the Drupal mailing lists. I'll acknowledge that there exists a wide variety of opinions on the value of certifications.

I'll also agree that certifications themselves are not the only answer to addressing the possible outcomes listed above. I'll also agree that many good coders aren't certified / don't need it. Internally to Acquia, we believe that good coders must have more than mere domain experience or book knowledge. We also think development process is important; we're an Agile/scrum development shop. But we also think that people with CS degrees (on average) have fewer gaps in software design patterns or coding strategies than those who don't have them. Would we hire great coders who don't have the CS degree? Yes. But looking for BS/MS in CS constitutes one of the tools we use to try to decode the capabilities of an applicant.

In the same way, the primary value of certifications is in helping people who are hiring Drupal consulting organizations make a first-pass cut at who they might interview and not interview when they are in a state of complete lack of knowledge about "Who's who?" The post by Evan Leibovitch said it best: A certification of reasonable quality is a useful tool in the hiring process; however, it is only one tool and it is no substitute for interviews, references and even (as appropriate) code samples. It is important to understand certification use and limitations -- anyone who hires someone only based on being certified (as used to happen in the Novell and Microsoft worlds) deserves what they get.

Certification is a little bit like tests for code: Tests (certifications) don't make good code (programmers), but tests can help weed out weak code (programmers). The Drupal community isn't (yet) as comfortable with tests as it needs to be. Dries wants that to change. In the same way, I hope the perception of the value of certification will change.

And we believe that if we succeed at this, we'll help the community of quality Drupalistas grow - which can only help Drupal.

The Yellow Jersey project

Robert and Jeff have already started working on identifying the best practices, and best programs in the software industry, and gleaning from them what we think will apply to the Drupalsphere. (Fortunately, Jeff comes with extensive experience in this from his executive experience at Adobe.)

Succeeding at this will take investment. Fortunately, we have some capital to invest. But we want to combine industry best-practices with community goals. So we'd love to have your thoughts about what you need, or want. Please tell us on our Project page for Yellow Jersey.