tl;dr: Video from CMS Philly presentation, Slides from the presentation, and scroll down to see graphs of particular interest to Drupal developers.
On May 1st, Chris Urban and I presented 2020 Developer Tool Survey Results at CMS Philly. For the past few years, we've run an annual Drupal Dev Tool Survey (2019, 2018) and presented the results at DrupalCon and some local Drupal Camps.
Since DrupalCon went virtual this year, and Chris was helping make CMS Philly virtual, he suggested I join him at that conference and reveal the results at this session.
This blog post summarizes the main results relevant to Drupal developers like me, who might want to know what other tools and techniques are currently popular for Drupal development.
Note that these 2019 numbers, as well as in the rest of the numbers from the rest of the comparison graphs below, are adjusted to scale to the same number of responses given this year. We saw a 44% decrease in total responses this year (from ~1100 to ~600), likely in part due to the survey's timing—it was released right before a certain 2020 global event started taking over the airwaves.
In this blog post, I limit graphs to the top 6-10 results in the interest of more readable graphs. You can find more data in the presentation and slides linked earlier in this post.
First of all, it's important to know the kind of geographical diversity represented in the survey respondents:
This year we had a bit of a larger representation from North America, at the expense of responses from Europe.
One of the graphs I'm most interested in compares the popularity of different local development environments for Drupal:
One quick note: Acquia BLT can use one of Drupal VM, Lando, or Dev Desktop as its backend, so the absolute numbers here may not reflect reality exactly, but they are a good representation of relative popularity.
What's most interesting to me is the year-over-year differences; Lando and Docksal have accelerated their growth, DDEV has jumped up four spots, and Drupal VM, Dev Desktop, Vagrant, and Homebrew (most of the 'non-Docker' solutions) have put on the brakes.
It seems that by next year, we may see the overwhelming majority of Drupal developers using Docker-based local development environments.
The next graph shows the relative popularity of different local IDEs or code editors. As in prior years, PHPStorm rules the roost, and other tools fall off in popularity pretty quickly. If you use PHPStorm, VS Code, Sublime Text, or Vim, you're with the majority of developers:
There are some other graphs Chris and I covered comparing code editors to region, OS type, and Drupal project type, which can provide some interesting data giving clues to why some people may choose one editor over another (e.g. more of the Linux users use Vim, Emacs, or other free and open source editors). Again, see the presentation and slides linked from earlier in this post for all the details!
Chris and I discussed adding a new question to the survey this year:
Where are you hosting your production Drupal sites? (Select up to 3)
The results are interesting:
And as with the other results, some of the relationships between team size and region can reveal interesting facts about the Drupal ecosystem. Since this was the first year we asked this question, we can't offer any historical comparison or trends, but one thing is clear: outside of having your customer manage their own hosting environment (which could also be one of the other listed providers—we don't know), the overwhelming majority of Drupal sites are hosted either on Acquia or Pantheon.
(Again, please note that there is a long tail of data including providers with < 17 respondents, but those are not included in this post for brevity.)
You can view the entire video of the presentation on Drupaldelphia's YouTube Channel:
Chris and I will hopefully be at it again next year. Until then:
Great to see the results! Look forward to it every year. Nice to see Sublime Text still hanging in strong.
Acquia BLT is an interesting one since it's really just a PHP task runner, it can run in virtually any environment that satisfies its dependencies. I myself am a Docksal addict and I've been happily using BLT with it for several years now.
Now if Linux could just get a decent Git GUI I would be so happy...
Yeah, the only reason we included BLT is a lot of people think they use 'BLT' and don't really know/care what is behind it, running their Drupal site locally. It used to be almost exclusively Drupal VM, but now there are many people using BLT with other backends to run the sites.
Also: Sublime Text rocks! Still use it day to day, or vim/nano in the terminal.
Thanks, great article! In the past I used AcquaDesktop, and it worked without internet connection. Recently I started using Lando and it requires WiFi for me to work on local instance. I am wondering if there is way to configure Lando to work without internet connection?
Yeah that's one thing I don't particularly like about the 'magic DNS' used by many of the docker-based solutions. For Lando, it looks like there's a workaround; see https://docs.lando.dev/guides/offline-dev.html#mac-specific-instructions.
Thank you for this great overview and for DrupalVM, I've used it for years
I would like recomend to look on this tool https://github.com/ironstar-io/tokaido
Personaly I'm developing similar docker-based tool called Nox.
The reason to build this tool is willing to spend less time for students to write their first Drupal module.
It is written in Node.js, designed to run code sandboxes for example on workshops, training days, local sandboxes.
It is similar to wodby.com in terms of services configuration using GUI but designed as CLI application.
Let me know if you found something interesting from the above.
For the Customer's Own hosting option, I'm wondering if it makes sense to split that into two, Customer's Own Standalone and Customer's Own with Aegir. The other way to do it would be to split this into its own section, e.g. "If you're hosting it yourself, which hosting system are you using?". Then there could be Aegir, Lagoon, etc.
Thanks for this! Looking forward to see this year’s results. It would be very interesting to know WHY people who used one tool in the last survey now prefer another tool, or WHY some people tried a new tool but decided to stay with their d one. I would very much appreciate follow-up questions regarding that.