At work I’m very fortunate to be able to try new stuff and be asked to try new stuff. And something I had to learn about recently was HubSpot (HS), more specifically their CMS.
In this blogpost I will talk about what my thoughts on the whole CMS is from a developer’s point of view.
# What is HubSpot?
Well, you could describe it as a corporate tool for marketers for inbound marketing. The reason I call it corporate is because it’s not really made with editors and everyday people.
# My thoughts start here
Alright, lets get started.
# How to setup a project
The first step is clunky. There is no way to create a local development environment, you have to work in their cloud. So their recommended first step for a developer creating a new website is to create a developer sandbox account. OK, fair enough. And to make things simple they have a boilerplate starter on GitHub.
After cloning the boilerplate you need to install their CLI in order to work with your newly created developer sandbox. And after you’ve connected you have to use the CLI again to actively watch your local filesystem for changes so that they can be “streamed” (one-way) to your sandbox.
How do you do version control? Well, you have to have seperate accounts for production and staging, and then you can use proper VCS. So remember the boilerplate I mentioned earlier? It has a GitHub Actions workflow that lets you deploy changes to your cloud environments.
To me this feels cumbersome, it might be because I’m so used to Craft CMS, but nonetheless I want to be able to have all my stuff offline so I don’t have to rely on anything other than my computer.
HS has its own templating language called HubL. It’s very similiar to other templating languages like Twig, Liquid, Tera etc. But it has its quirks 😅
One thing I haven’t seen in any of the other languages is the ability to apply the templating to CSS-files to make them dynamic. I’ve always looked at that as design decision so that you don’t overcomplicate your code (example from HubL).
For the most part it’s OK, but the documentation is confusing at times. I remember when I nedded to filter specific values from an array. I tried to find a
filter-filter, but none where to be found. But after a bit of time I discovered two filters:
selectattr. Not a big deal, but it was not expected to find two entirely different named filters for such a common thing to do.
I would really like for docs to learn from Twig’s docs about how to structure and name everything.
# Serverless functions
This was an adventure. First there’s no mention in their official docs for running the functions locally. So after running the functions online in their cloud for a few days I found it by randomly checking their CLI with
# Outdated Node
Back in November 2022 when I started working on the functions the latest version of Node available was
v14.x, which had its EOL on 30th of April 2023. I can now confirm that they have upgraded to
# Pre-installed packages
As of today, May 11th 2023, there are currently four pre-installed packages that are available:
Even their own first-party package isn’t updated, the latest version is
v8.9.0 as of May 11th 2023. All the other packages are also super outdated, and “request” has been depreacted since February 11th 2020.
Since their own package is outdated I have to use their own REST-API when working with data coming from the CMS.
I don’t think I need to say anything else.
# Shared code
There is no offical way to share code between different functions other than a officially recommended community post about using Webpack. However, it’s entirely possible to have shared code when running the functions locally 🤷.
# Secrets / Environment variables
You have to either store the secrets in the
serverless.json-file or use their CLI to add secrets to you cloud accounts. Be sure to remember to switch accounts using their CLI to update the correct account’s secrets.
Also, when you’re working locally you can use dotenv, but the functions ignore comments like
#VAR_NAME="secret". It’s still available with
One really cool thing is that you can use diffent dotenv-files with the CLI when running the functions locally with
hs functions -c, and this commands actually checks if the dotenv-file is ignored by git which can come in super handy if you’re not careful.
You’re limited to 4 kB logs unless you run stuff locally, then you can see everything. And 4 kB may sound like a bit, but it really isn’t - it’s basically for writing
Not sure if there’s anything to conclude, I’m not very impressed with HS after working with it for a longer period of time. I haven’t even talked about the editors UI for creating pages etc. I might also have forgotten a couple of things, but this post is mostly for reference and in case some people stumple apon it thinking about trying out HubSpot CMS.