The iOS Developer Community Survey

How is Swift on the server doing?

This page contains analysis and opinion by Dave Verwer.

Even though using Swift as a language for back-end web development has been possible for several years, it’s still very early days in terms of widespread adoption. That said, the leading server-side Swift framework recently hit version 4, which is a huge milestone for a maturing platform. So how do developers feel about it?

The survey showed very mixed results.

While this result might be disappointing if you were hoping that Swift on the server would be the next big trend in web development, I do think it shows some promise. Yes, the most popular response was at the lowest end, but the average was still slightly above 5.

I’d speculate that one reason for the low responses is that this was a survey of Apple platform developers, not back-end web developers. Would filtering the results to only include responses by developers who have also done back-end development within the last 12 months change things? Let’s see.

Surprisingly, there’s not a massive difference in the results with this filter applied. It does trend upwards very slightly with the average increasing from 5.4 to 5.7, but there’s no significant change in the distribution.

What does this show? While it does seem like frameworks (especially in the JavaScript arena) come and go extremely quickly, the languages that become popular for back-end web development have historically changed very slowly. Swift is still a relatively young language in any capacity, and it’s very new for back-end web development.

It’s also much harder for businesses to change anything as fundamental as their server-side language, so it should be no surprise to see that the level of interest is significantly lower than people’s personal opinions. Even so, it’s impossible to avoid the fact that the overwhelming majority of companies have little to no interest in Swift on the server right now.

After the sad news that IBM ceased development of Kitura last December, it’s no surprise to see that Vapor walks away with the lions share of which framework people are using. Is Vapor going to remain so dominant in next year’s results? It looks likely, although I do hope we see a little more competition before settling on a de-facto framework as I believe that everyone would benefit from more experimentation.

So as we look to the future, are people planning to experiment with Swift on the server this year?

The data clearly shows that there’s more interest in this technology coming, especially for personal projects. That’s a positive sign for sure, and I’ll be interested to see if this translates into an increase in the responses to Q49 in next year’s survey results. Remember that saying you’ll do something is much easier than actually doing it! Watch this space to see if intention turns into action this year.

However, I think there’s more to be learned from that question’s data. I had a slight concern that the people who were interested in seeing Swift on the server succeed might also be the people who think that Swift should be used for every coding task (an attitude which I don’t believe to be particularly healthy). So, let’s filter the responses to developers who already do back-end web development, and by people who are open to using any language on the server.

While the percentage of people who answered “No” increases with this filter applied, I think it’s significant that almost a quarter of developers who would be open to using other server-side languages are planning to use Swift on the server this year. Again, my point about intention being easier than action applies, but I think there still might be the twinkle of a bright future for the language on the server, and I can’t wait to see next year’s results to see if the twinkle turns into a sparkle. ✨

Was there anything that could have been improved in this article? Let us know.