Add-on Developers Survey: Raw Results

Mike Morgan

6

We recently conducted a survey of the add-on developer community. This post summarizes the quantitative results and some key findings only. There were several open-ended questions included on the survey. A special thank you to all those respondents that provided us with extensive feedback and great ideas. A summary of developer comments and suggestions will follow at a later date.

We had a total of 265 surveys completed. Note that some questions were mandatory, others were allowed to be left blank or unanswered and others allowed multiple answers. When multiple answers were permitted, the total number of responses for that question is shown.

The general summary is that most respondents were individual developers with 2 or more years of add-on development experience. Natually, they felt comfortable with JavaScript and XUL. They primarily built add-ons for Firefox and hosted them on their own (possibly, in additional to AMO).

These are simply the raw results, we’ll be applying some correlation analysis to draw out more interesting findings and conclusions.

Q1: Which of these best describes your status?

Q2. When did you write your first add-on? (for Firefox or other applications)

Q3. Select the applications (or devices) for which you have created add-ons, 3rd party applications, or plugins.

Q4. How many distinct add-ons have you created? (for Firefox or other applications)

Q5. Was the functionality of your Firefox add-on similar to your add-on for other applications?

Q6. If you have ever started to develop an add-on and then aborted the project, what were the primary factors in your decision to stop development?

Q7. What tool(s) did you use to develop your add-on(s) (for Firefox or other applications) ?

Q8. What tools would help you develop more efficiently?

Q9. How would you rate the ease of development of Firefox add-ons?

Q10. How would you rate Mozilla’s developer documentation?

Q11. How familiar are you with JavaScript and XUL?

Q12-Q14. What were the top three challenging aspects of building your add-on?

Q15. Where do you host your add-on?

6 responses

  1. Joshua Cranmer wrote on ::

    Just looking at these results, one piece of information I find surprising is the documentation aspect. About 15% say they were challenged by inaccurate documentation, but only 3% said better documentation would help them code better… If you count those wanting “bigger picture” documentation (i.e., tutorials and other source code), that accounts for about 40% of the responders. With that much wanting moderately better documentation, it’s surprising that devmo’s documentation was rated as highly as it was.

  2. Steve wrote on :

    The way you’re graphing the data is rather odd. e.g using a pie chart for Q7 is a bit silly, since the percentages aren’t of any meaningful total.

  3. bhashem wrote on :

    @Steve, if you take a look at how the question was phrased (see survey), several choices were fixed so the percentage is what percentage of people chose an answer. A more in-depth analysis of tools is required, the question was geared at understanding what category of dev tools was being used.

  4. Sébastien Marinier wrote on ::

    @Joshua : In Q8, you see also an IDE might help. In developer’s mind (in mine also), when you hav an “IDE”, you have some useful tools like completion code, searchable (and organized) documentation, browsable class tree… 17% + 3% is greater than 15%

    Moreover, maybe the lack of time (Q6) is due partially to the hours spent to search Code Samples, class documentation, a.s.o… ;)

  5. Maciej wrote on :

    Was this survey only for Firefix Add-on authors or for Thunderbird as well? I have written a Thunderbird add-on only, so I decided not to fill the form. The title clearly indicated you want Firefox add-on authors only. Will there be a similar survey for Thunderbird?

  6. Brett Zamir wrote on :

    “With that much wanting moderately better documentation, it’s surprising that devmo’s documentation was rated as highly as it was.”

    I’d venture that maybe it is due to the vastness of the Mozilla platform. What is documented is nice and helpful but there’s so much more (especially XPCOM) that could be documented with examples, etc.