Ember's growth among developers remains steady.
Over half of Ember developers self-identify as being very familiar with
both JavaScript and with Ember itself (over 7 on a scale of 1-10). The vast
majority (86%) live in North American or Europe, prefer Sublime or VIM
(85%) and commit their code to a Git repository (93%).
The vast majority (88%) of participants self-identified as a developer. 90%
of participants used Ember at work. Most (77%) were on small teams of
one to ten developers, although many (62%) claimed their entire
development staff was one to ten individuals. This leaves
over a third of Ember developers working for a company employing
more than ten developers. 13% work for a company with over one
hundred developers. 84% have an Ember application in production.
During the survey window, Ember.js 1.10 was announced. By these numbers
75% of Ember developers are working with a very recent stable release (most
recent or previous).
25% of developers still support a version previous to 1.7, released August
23rd 2014 (about 6 months prior to this survey).
Most Ember applications move forward and adopt new versions
of the framework. Especially
when compared the adoption of Ember-Data versions, the story
of sematic versioning and committing to API stability
reads like a success.
Ember-Data adoption tells a very different tale. 60% of users are on
an older version of the library, only 45% on the most recent release.
Ember-Data does not adhere to semantic versioning (yet), and the difficulty
of upgrades is visible in this data.
Nearly half of all respondents reported that Ruby is the primary language of their server-side stack,
and a little over a fifth reported JavaScript-based backends.
Together, Ruby and JavaScript account for over two-thirds of the server-side stacks in use by survey respondents.
Server-Side Language preference follows a similar curve as the languages in use,
but with JavaScript having the biggest difference between usage (22%) and preference (45%).
Rails is, unsurprisingly, the most commonly used framework, with slightly over half of respondents reporting it as one of the server-side
frameworks in use. Express, at 25%, is the second most-common reported framework. Ember allows for many potential server-side options, however, as borne out by the survey results:
The third-most likely answer, at 14% of survey respondents, was "other", indicating a framework that wasn't included in any of the 11 enumerated choices.
Ember-CLI has quickly become the preferred build tool for Ember
apps as 83% of respondents indicated using it to build their apps. This
level of adoption is astounding, and speaks to the community's focus on
and willingness to adopt new conventions.
The number of developers supporting mobile platforms will continue
to grow over the next year (56% to 67%), while the number supporting
legacy platforms will continue to drop. The percentage of developers
supporting IE8 (the last
commonly supported browser without ES5, and for which Microsoft will drop support in January
of 2016) is expected to drop from 15% today to 5% next year.
Less than 1% of developers try to support IE6 or IE7.
Microsoft's inability to launch an evergreen browser continues to be a
burden on the development community. More developers support IE9 than
legacy versions of evergreen browsers. It has been suggested that the
Windows 10 browser, Spartan, will be an evergreen browser.
The addition of addons through Ember CLI has opened the door to extensibility,
and nearly half of survey respondents report having authored at least
one addon. Less that 20% of respondents, however, report publishing an
addon publicly, indicating a high level of internal addon usage.
The testing story for addons remains under-developed as less than 15% of
respondents report implementing either unit or acceptance tests in their
addons.
SEO is undoubtedly an important aspect of web development. Despite this
the large majority of Ember developers don't see SEO as a challenge
they need to solve in their applications, or don't take pro-active
steps to improve SEO. There is definitely
an element of self-selection in these numbers: It is unlikely
that Ember is being chosen for SEO-sensitive applications.
Revisiting these numbers next year, in light of FastBoot, will
be revealing, as 46% of developers report that their applications
target public consumers.
We would like to thank everyone who took the time to participate
in the 2015 Ember Community survey! We hope this information can
provide a platform for discussion and ideas around the entire
Ember ecosystem as it moves forward.
You can view a summary of the responses to all the questions from the
survey by clicking
here, and you can
view the raw survey data
here.
Share the results:
For Next Year...
Hindsight is 20/20, and there are a few topics we regret not covering
in the 2015 survey:
“Do you self-identify as an under-represented demographic in
technology?”
Ember-RESTless should have been included in the data-layer options.
“What is your preferred learning resource? Videos, books, ebooks,
open source code, formal training, etc.”
“What hosted data service do you use? Firebase, Meteor, Parse, etc.”
“What level of education do you have? High-school, College degree, CS degree, etc.”
Question to better understand accessibility requirements and implementation should have
been included.
You are encouraged to discuss these and other issues at EmberConf
this year. If you have feedback to share, or think we missed a question
not mentioned here, feel free to email survey@201-created.com