After its first 10 years, interest in Mozilla-bred programming language Rust seems to have run into saturation. Although it continues to be championed by StackOverflow readers, taking the title of most loved programming language for the fourth time in a row, and is still receiving stable ratings in language indices, adoption isn’t exactly soaring.
When in doubt there’s still a tendency to use Go instead of the C/C++ challenger for infra projects, since it seems to be easier to get started with and most of the cloud native crowd are on it anyways.
A few years ago, to tackle this development, the Rust team started the tradition of annual community surveys, collecting feedback and helping them determine the direction of the coming months. The results of the fourth annual survey of the Rust community are now out, pointing to weaknesses when it comes to bringing businesses into the fold and offering help to intermediate users.
For the newly released 2019 edition, the core team managed to motivate 3,997 (compared to 5,991 participants in 2018) Rustaceans to share their joys and woes with the language.
While most saw at least slight improvements in what the Rust team dubs “key” aspects of Rust in the last couple of months, getting into the language was one of the few items on the list which made people compelled to say it had actually gotten worse (~100 participants). This is of course contrasted by almost double that who felt the learning curve had improved a lot. However, being too complicated or hard to learn is also among the top reasons people have stopped using the language altogether or were not looking into it at all, which isn’t the best of signs.
It also plays into the lack of adoption in general, since some users pointed out that a better learning curve, additional documentation and learning resources, better IDE integration, and GUIs would help to convince people to use Rust more. The initial troubles are also reflected in the results to the question, if those using the language would feel productive yet, which was met by 21 per cent with a resounding no.
That 70 per cent felt they could make good use of the language within their first year sets that into perspective, though it probably won’t help devs evaluating it as an option for work if they feel discouraged three months in, if there are other options available.
Since the issue hasn’t been put on the table for the first time, the Rust community had enough opportunity to think about ways to improve the matter. Tips in that regard ranged from more intermediate level tutorials to video content, and holding meetings at more accessible times. Users that aren’t native to the English language also voiced a desire for better documentation in other languages.
Sorting out the latter could be a rather quick job for some languages, though the core team has to follow through as well, since the stagnation of some of the translation efforts is down to the lack of an active maintainer responsible for final approval.
Just recently the Rust core team officially bid farewell to its docs arm, feeling that it wasn’t needed anymore (but mainly because of seeming lack of new blood). This however means that the rest of the Rust ecosystem has to step in now and maybe look at ways to support users beyond their initial trials.
Apart from this, the survey found that most devs use Rust with either Visual Studio Code (34,9 per cent), text editor Vim (23.6 per cent), or JetBrains’ Java IDE IntelliJ IDEA (21.1 per cent), with a bit more than half of all survey participants developing on some version of Linux. The number of programmers making use of nightly build has gone down from 56 per cent to 30,5 per cent, which seems to have surprised the Rust team.
Most targeted platforms for Rust developers are Linux (36.9 per cent), Windows (16.3 per cent) and MacOS (14,7 per cent) – closely followed by WebAssembly at 14.4 per cent, which could be down to Mozilla heavily championing the platform, which makes Rust a good match.