The vast majority (87%) of participants self-identified as a developer. 92% of participants used Ember at work. Last year 62% of participants had a development staff under 10, this year that dropped to 56% of respondents. The number of participants at a company having over 50 developers rose from 18% to 22%. This, along with the two graphs above showing how development roles have changed and Ember staff size has changed, points to a small shift toward larger companies. 85% of participants have an Ember application in production, 8% of participants work on an Ember application used by millions of users.
Ember 2.4 was the release version of Ember during this survey. 76% of participants use Ember 2.x in an app, which demonstrates incredible adoption for a major release. However 36% of developers are still managing a codebase on 1.13. Major semver releases include breaking changes, and a spike on the last 1.x release is expected to some degree. Anecdotal evidence suggests many large companies are just now crossing the threshold into 2.x.
Ember-Data, which did not follow semver until its June release of 1.13, shows a similar trend after that release. Despite a bump at 1.13 (27%) most participants upgrade seamlessly through minor versions. This is a great improvement over last year.
Participants reported that their applications trended older this year. More developers are working with long-lived Ember applications than ever before, moving through release versions over time.
Ruby retains the top spot in desired and actual server-side languages, however it also dropped 6% as production platform and 8% as a preferred platform. The only notable mover further down the list is Elixir/Erlang, which lept to surpass the popularity of Go in both use and preference.
Participants overwhelmingly support modern browsers. In a shocking twist, the number of Ember developer who maintain support for mobile browsers dropped from 57% to 49%, and is only projected to return to 57% next year.
In the past year Microsoft launched its evergreen browser Edge, however Edge was not made available on older platforms like Windows 7. Older IE versions will be widely supported until those platforms are all retired, and their last available version of IE with them.
Current app support for IE9 is lower than support for IE8 was last year, providing a strong justification for Ember 3.0 (when it arrives) to retire IE9 support. IE10 support is popular despite the relative rarity of that browser (it is less common that IE9).
The number of community members who have written an addon grew in the past year. An astonishing 52% of participants have written an addon, and the percentage of participants writing addon unit tests rose by 12% to 64%. 25% of participants report publishing an addon publicly, up from 20% last year.
Concerns about SEO grew this year, if only by 5% of the participants. FastBoot usage remains nascent. Similar to last year, we suspect an element of self-selection in these numbers as it is unlikely that Ember is being chosen for SEO-sensitive applications. The number of Ember apps targeting public consumers rose slightly from 46% to 49%.
We would like to thank everyone who took the time to participate in the 2016 Ember Community Survey! We hope this information can provide a platform for discussion and ideas around the entire Ember ecosystem as it moves forward.
Questions? Feedback? Please email firstname.lastname@example.org