After almost twelve years of Drupal, I am saying goodbye

Woodcut from Doré. Purely illustrative
Doré Woodcut. Its only function is to make the layout look better. And these images are really nice themselves

Over eleven years ago, I got involved in Drupal, after running two personal blogs on Drupal. A few years later, to my surprise, I found myself to be amongst the first few people to offer commercial and professional Drupal services.

And boy, it was a ride. With no IT-education, other than webmonkey, the fantastic Drupal-community allowed me to piggyback and stand on shoulders. Great, strong shoulders.

I grew. I learned. And I learned how to program somewhat decent. I then, gradually I learned there were books and systems that allowed one to program The Proper Way[tm]. Mostly because I was rolling out Drupal in Real-world projects every day, bumping into issues that, as I found, had been "solved" ages ago in academical books and studies. I learned that "talk is silver, code is gold" is simply not true. Code that is discussed, thought about, refactored, and then discussed again, is of a much higher quality than code that is "just done". I learned that properly architecturing something, turns it into far more than gold. I often complained about "horse-tack-coding" in Drupal. Where working on small, isolated issues was (and is) always preferred over refactoring larger parts. This has led to a lot of repetition, inconsistent APIs and very unpredictable behaviour. I learned about encapsulation, separation of concerns, loose-coupling and more such well-known principals. All of which Drupal lacks, or ignores. I had the feeling I grew faster than Drupal.

I also came across Ruby on Rails and found that there were actually real, technical solutions for several of my gripes with Drupal. We are talking 2005, by now. Remember, I have no educational background. At first, I knew nothing about OOP, other than what the great folks, the great shoulders of the Giants in Drupal, told me about it on IRC and in forums.

I learned a lot of new terms, methods and concepts. They were, and still are, an eye-opener. MVC was something that actually existed! Something actually existed, actually got designed and invented, solely to solve most of the issues I had with Theming in Drupal! And these design-patterns were not just to make technical people happier (or something to flamewar about), they actually solve many management and planning-issue too.

But, I had also grew into something of a local Drupal-expert and goto-guy. Serving most of the Netherlands as freelancer and Drupal consultant payed for my mortgage. I got called in on many failing Drupal-project. Got to help large companies and organisation in their swich to Drupal.

But toggling between Rails and Drupal-work, only made me see all the issues with Drupal more clearly: there was a lot of work for me to do, in order to make Drupal something as elegant and nice to work with as Rails. In an ever growing Drupalcommunity, I decided that my voice and code in this was only noise; especially since that community clearly has a different idea about webdevelopment than I have.

Around that time (beginning of 2009), I co-founded Wizzlern. We developed training and education for Drupal. Training people allows you to meet professionals with lots of different backgrounds. People who have formal training in IT. People with much more in-the-field experience than I will probably ever have. And people who are critical. About things in Drupal.

But developing several training-courses also required me to dive really deep into the what and the why behind things. I suddenly had to paint the big picture around an inconsistent and weird API, answer questions like: Why are modules so hard to find? How come there are so few pretty themes, compared to wordpress? Why is it so much harder to use than Wordpress? Why can't we find a decent workflow to develop in a team and deploy? Feedback from experienced webdevelopers (in Java, Python, .net and PHP-frameworks) made me realize even more that there was something amiss for me.

It became harder for me to defend that, harder to explain my passion for the system, its quircks and its community. They say, that once you have looked in the kitchen of your favorite restaurant, you never want to eat there again. That could be the case here. Or maybe it is because I am a vegetarian.

The realisation came slow. It took years. Drupal actually wants to be what it is now, not what I thought, or hoped, it wanted to be. My idea of a toolkit, developed by webdevelopers, for ourselves, wedevelopers, to create ever better websites for our clients, was a not going to be found in Drupal.

Dries's comment on Copenhagens' keynote made this even more clear for me. He pointed out that Drupal should not focus on developers.

"Drupal made the webmaster redundant. In future it will make the webdeveloper redundant".

Unfortunately, I cannot find the exact quote, this one is from my vague memory and scriiblish note I made back then. I can only find Dries answering my question about that quote. So his exact words are most probably different from what I phrased here!

However, the bigger picture became more clear to me: something we have seen happening in Drupal for a while now: It focuses on the click-and-point development of website, not on the programming of a website. It really wants to be a CMS, albeit a flexible one. Rather than what I consider the future: a developer-platform that allows me to build a CMS.

Development in with click-and-point, offers little challenge for me. Learning, where and how to find, evaluate and configure the umpteenth gallery-slider-view-plugin offers no challenge, nor satisfaction. I also found this approach of clicking together a site, to not satisfy my clients; not being able to deliver 100% what they wish. And I found it inefficient: especially when my programming skills grew. I could churn out a few objects and a hook or two that output the exact JSON I want, much faster than I would ever be able to click together some Services configuration.

In my search to more challenging Drupal-work, I helped several large sites, to solve some of their performance-issues. Helped many project with their problematic Drupal-development and -deployment. I taught many developers how to write themes, modules and how to deploy. Unfortunately the challenges did not revive my love for Drupal, but only took me further from it. I came to realize that most of the problems stem from the way the Drupal community prefers to do stuff. I even wrote some controversial opinions on that (Dutch).

And I decided that it was time to make a shift. Find new projects outside of Drupal, work on some pet-projects and see if I found more challenges and opportunities to grow again, outside of Drupal. After nearly one year of flipping between Drupal and other projects made me realize that I had to cut all ties, in order to progress. That Drupal was never going to be the developer-tool I hoped it to become.

I will put down all my Drupal-work and finish the remaining few of my running Drupal-projects. Both those with clients, and those things like a stable release of tagadelic2 for Drupal 7. I will obviously announce those here.

Moving on to exciting new technologies, tools and development platforms that fit better with my workflow and programming-experience.

Goodbye, it was a fantastic bunch of shoulders to be allowed to stand on!

About the author: Bèr Kessels is an experienced webdeveloper with a great passion for technology and Open Source. A golden combination to implement that technology in a good and efficient way. Follow @berkes on Twitter. Or read more about Bèr.

blog comments powered by Disqus