Please log in to watch this conference skillscast.
Most web applications and sites require a back end component running on an always-on server and executing code (or even querying a database) on every request. This is a very wasteful approach both in terms of environmental and running costs. But it also has a big impact on the website visitor experience, making things slower, less reliable and less secure. Caching can go a long way in mitigating these impacts, but it’s difficult to make work well, especially if it’s spread onto multiple layers of the stack requiring a coordinated effort from multiple disciplines.
A much better approach is emerging at the confluence of static site generators, “Serverless" (or “Serviceful") back ends, microservices, browser APIs for working offline, scriptable CDNs and build tools, which are getting smarter and more integrated every day. The core of the idea is to preprocess and serve up as static files as much as possible, while letting the dynamic parts loaded runtime by directly connecting to APIs. This is very similar to how native apps on mobile platforms work (and what makes them popular), by bundling the application shell as a separate download and fetching data on demand, but the web offers a much more granular control and with that a possibility for an even better experience.
Join Daniel as he compares the traditional application server based approach with this new method, offering some practical insights and the pros and cons of each, as of course the applicability depends on the type of project at hand and the structure of the team delivering it.
YOU MAY ALSO LIKE:
- Gojko Adzic's Impact Mapping Workshop (in London on 27th September 2017)
- Uncle Bob's Advanced TDD (in London on 30th - 31st October 2017)
- Uncle Bob's Clean Code: Agile Software Craftsmanship (in London on 1st - 3rd November 2017)
- Agile Testing & BDD eXchange 2017 (in London on 9th - 10th November 2017)