Decoupled Architectures: State of the Art, Not Bleeding Edge

Watch Pantheon's Introduction to our Front-End Sites Blog Series - brought to you by Steve Persch, Pantheon's Director of Technical Marketing.

At a recent conference, I took everyone back to what was once the future. I wanted to give the audience a bird’s eye view of how we got to decoupled architectures and where we are going next in the remainder of the 2020s: the Fourth Decade of Website Deployments. 

I used a folding two-meter measuring stick and labeled significant dates in computer science, from 1970 – the beginning of the Unix epoch – to 2070 to make a somewhat comically large physical timeline. I find that with a prop like this one it is easier to see where our present moment sits in a larger context. 

The last decade especially has been a noisy time in website operations where the pace of change can feel overwhelming, even if that change often spins in circles to wind up in the same place again. The noise fades out when looking at the longer-term view.

That longer historical context also makes it easier to see the value in a guiding phrase that steered Pantheon to the arrival of Front-End Sites in the Pantheon dashboard: we need to be state of the art, not bleeding edge. Not coincidentally, that phrase came from a web leader within a centuries-old university running more than a thousand sites on Pantheon. Through conversations with him, I first thought of the yardstick visualization. Our multi-year relationship with this university may feel long measured in tech-startup terms but on a timeline where each decade only spans an inch, our years feel pretty small.

Zooming out to the scale of decades elevates the importance of website operations questions such as

• Are we reaching our goals?

• Are the stakeholders working together effectively?

• Are we staying on track? Or, is there a track at all?

At Pantheon, we see it as our responsibility to maintain the track that web teams move along. It is easier for our community to move efficiently along the track when we set reasonable guardrails, when we make the state-of-the-art options be the default options. Or, if we can go a step deeper and adhere to the WordPress philosophy of "Decisions, Not Options," which calls on developers to avoid putting the pressure of complex technical decisions on end users.

With our Front-End Sites product we want to apply the same mentality to an ecosystem overloaded with options. The endless options and effectively infinite permutations of npm packages pull teams into the weeds. These are questions that drag you mentally from the yardstick view of decades down to a nanometer scale of feeling like everything is riding on your ability to nail the perfect technical architecture right now with fast-evolving technical components.

That's a bad place to be. We have heard too many horror stories of teams needing too many (figurative) band-aids on their decoupled sites because there was too much bleeding on the bleeding edge.

As a company, we have kept our customers clear of band-aid solutions by taking years of experience running CMS sites and baking best practices into an opinionated platform. The right time to start Pantheon was roughly a decade after the emergence of PHP CMSs like WordPress and Drupal. Now that the wild world of the World Wild Web’s front-end frameworks has evolved for a similar amount of time, we see an opportunity to set new guardrails.

We expect that in the rest of the 2020s, more web teams will be able to spend more time delivering better user experiences for their site visitors if they keep their focus out of the minutia and on the big picture.

Back to my yardstick. We can see more clearly that now in 2022 we will soon have more years of web development in the iPhone era (debuting in 2007) than we had years of pre-iPhone websites. In the moments of that post-iPhone era has often felt like a turbulent and ever-changing storm for teams on the bleeding edge.

Right now, while Front-End Sites is in Early Access, we want to engage more with teams who have been to the bleeding edge and back and can point to where they think the guardrails need to go so that more teams can follow safely. If that sounds like you, please let us know and you can start trying out Front-End Sites.