Mariano Viola

Everyday Web Designers have to struggle with several legacy technologies, mostly related to the browsersity and legacy LAMP stacks. However new scenarios might gradually improve this stagnant situation.

The Charm of the Imperfection

JS and PHP share many things, they both first appeared in 1995 and increased their design and reputation over the years without a sudden mirroring of these new features by browser vendors and hosting providers. Being able to manage the span among their several versions is a daily challenge. In fact, the attempt to update the codebase by adopting a new version of the language can easily break a chain of dependent plugins, libraries, and frameworks.

jQuery and Sass are still the most used front-end technologies, WordPress powers 25% of the web and the LAMP stack is quite popular. There are many websites made in this way, mostly because this represents the entry-level option to self-publish something on the web, and frankly, I think that with a bit of discipline, you can build whatever you want and run a profitable business using them.

Face Diversity not Hugeness

Over the years, front-end development and the LAMP stack have been improved dramatically, React is a revolutionary act in front-end development, PHP 7 and HHVM take the “Personal Home Page” to the next level in terms of performance, and MariaDB is a step forward respect to MySQL.

Despite all these improvements, Web Designers have to face the complexity coming from diversity. They don’t need to scale and tweak the performances of a single huge web application. Instead, they have to design, develop, and publish several small-medium websites, and in many cases, they can’t host them on their servers and they don’t have control over the thousands of devices and browsers out there. In these cases, they have to embrace what is at their disposal, considering adopting tools that can safely run on many different browsers/devices and on a baseline/shared LAMP stack.

Beyond Social Mass Media Conflict

This scenario seems to make the role of Web Designers and their products increasingly uncompetitive with respect to the “Social Mass Media” platforms. It is also true that they can take advantage of their progress, coexisting and cooperating with them to keep the web a bright space. Facebook and Google among others, are contributing to making the web pseudo-platform still attractive for many developers, and the new WordPress.com (AKA Calypso) is an example of this informal joint venture. Calypso is also a perfect example of a contradiction that breaks a kind of dogmatic thinking that tries to divide everything into good and bad factions.

These Social Mass Media have inherited the pros and cons of their ancestors, but now, differently from before, we shouldn’t limit ourselves to using them, instead, we should embrace their good parts letting our users add a website that complements their Facebook pages in a way that they can’t perceive a huge experience gap between them. In conclusion, WordPress, once again, can potentially help Web Designers and independent publishers keep the web healthy and balanced.

Business in Front, Party in the Back

Calypso is a nice composition of Node.js modules orchestrated in order to manage each WordPress instance from a central admin interface using a RESTful API. This approach helps to mitigate the pain of Web Designers and the complexity given by the management of several different websites. It is a huge step forward, but we still have to host several WordPress instances and their data somewhere else, pushing and pulling data from them. So what might be the next move?

The first straight proposal should be the reimplementation of each WordPress instance, with an arbitrary stack and the only obligation to expose a compliant API. What about if the data of each website lived in the same place where the admin is hosted? Imagine something like Firebase, an admin interface in which collections of pages can be managed, categorized, easily customized, and exposed with a RESTful API.

A Dynamic Site Generator

It looks like yet another headless CMS. So what’s new here? Well, first of all, this should be an open-source product. Moreover, this new machinery could generate a static representation of the data stored in each website, match these data with a nice template, and push it on any external server (with an FTP/SSH account and a static web server) every time data change.

Just import your data from your old WordPress websites, install missing plugins (rewritten to fit the new platform), install and customize their themes, et voilà, and all your old websites are turned into fast, ubiquitous, and secure static websites. In this way, the main server isn’t overloaded and generated web clients are treated as one of the several kinds of clients that the new WordPress can serve.