HN Explorer

Ask HN: Do you have examples of larger companies using static site generators?

6 points by lwhi ↗ HN
I'm currently looking into the viability of using a static site generator (GatsybyJS) to provide a PWA-type experience for a multi-locale CMS driven site.

The content will be maintained using a CMS, and the site will be regenerated and re-deployed on each publish. Any transactional requests will be serviced by a custom API-driven backend system.

Aside from performance and scalability advantages, I like the extra layer of security provided by removing dynamic-access to the CMS. I also like the level of control provided by combining code and content within each discrete build artifact.

In order to help make my case, it would be really useful to reference examples of larger companies that have successfully used this approach. Conversely, any stories about why this approach could fail would be equally useful.

Best

Luke

4 comments

[ 2.9 ms ] story [ 24.4 ms ] thread
jppope
I'm moving all my WP/Static stuff to a SSG and am incredibly interested in knowing the same... I know the Vue docs are generated using a SSG, and all of Segment's stuff too... but I would be very interested to know other companies
smt88
This is effectively how most large orgs use WordPress. There's aggressive caching that stores static versions of each page.

It's unusual for high-traffic sites to dynamically generate every page. For WordPress, the trigger is sometimes "on first visit" instead of "on publish", but otherwise it's a good analogy for what you're doing.

sgallant
Good job! We see many orgs adopting this approach for the same reasons you mentioned (disclaimer: we build a CMS for these sites https://forestry.io).

The healthcare.gov site might be a good comparison for your team. It was a large legacy-CMS site when it launched. It had a lot of performance problems so they migrated to Jekyll which solved everything. Here’s a quick video from one of the guys describing the migration: https://www.msnbc.com/all-in/watch/the-nerds-who-saved-obama...

Also, Kyle Rush worked on the original Obama site (Jekyll) and shares insights here in this video. He attributes their success to the fact that their front-end devs were able to A/B test the site multiple times per day to increase the donations. It’s loaded with good arguments to move to a static platform. https://www.youtube.com/watch?v=ia8vsuiXiL0

If you're curious to see some build-time benchmarks for Jekyll/Hugo, check out this (from our team) https://forestry.io/blog/hugo-vs-jekyll-benchmark/