Scala Center and VirtusLab would love to hear from you!
Let us know how you use Scala, what parts of the ecosystem are vital for you, and what we can do to make Scala even better.
The survey is open until October 15th.
5 Likes
I think the survey was a lost & wasted opportunity to collect open-ended, written feedback.
It presented question after question that presented detailed, long lists of fine-grained libraries and technical options and asked my rating on each of them. But why were particular libraries chosen for inclusion and not others? Why no ability to add other options than the pre-selected ones?
I found myself surprised when having churned through pages of these ratings, the survey abruptly ended without a single opportunity to give an open-ended response.
Scala is facing some big challenges currently. Disappointingly, the narrowly constrained responses in this survey won’t really shed light on how best to respond to them.
2 Likes
Seconded. I feel it was really important to gauge the outcome of the Akka license change with respect to scala adoption. I do know of companies that decided to drop scala entirely due to this, but without an actual survey it’s impossible to know if it’s just outliers.
That was never the goal of the survey. If the survey has more open-ended questions, we couldn’t count on having a thousand answers (this is how much we have right now), but probably a quarter of that. Moreover, gathering helpful information from those answers would be much harder. People only give longer answers when they have strong emotions about a topic. That could lead to skewing the results towards one non-representative group.
From my experience, I can say that open-ended questions have a high user drop-off rate. Some people close the tab when they see an open-ended question instead of simply skipping it. I don’t understand the psychology behind this behavior, but it exists.
This doesn’t mean that we do not value the longer feedback. A big chunk of my work is talking with developers and team leaders from different companies about their experience and difficulties with Scala, especially about prospects for adopting Scala 3. We also spend much time reading posts with feedback on social media, such as this site or reddit.
The focused meetings with teams give a much better opportunity to gather more nuanced information. On the other hand, the survey is targeted at the widest possible group of scala users and is great for assessing trends in time (this is why this survey was so similar to the last one) and correlations.
We based this on perceived popularity. We used exactly the same methodology last year and we only got messages about three missing libraries. We added them. Is there something you find missing this year?
3 Likes
This is the kind of nuanced feedback that I mentioned in my previous comment. I personally believe it is much better to obtain it by talking directly to the teams and companies, and we are doing it.
Sadly, the Akka license change had a big negative impact on the entire community. However, when talking with teams that are considering dropping Scala, this is hardly ever the only reason or the deciding reason. The truth is usually complex and I don’t think that a survey like this is a good way to know more about such complicated matters.
1 Like
I didn’t mean to imply that. In both cases I saw the general eng-teams at the company were at odds with scala to begin with and would’ve chosen something different if they could. They were using Scala because Akka was chosen for its features a long time ago and stuck with it because alternatives (in other languages) weren’t really worht it. A typical case of framework justifying the language 
.
In general, the argument against open-ended questions in surveys like this is that, more often than not, they’re essentially a lie. They make people feel good because they get to say whatever they want, but in many cases they simply get ignored, because it’s way too labor-intensive to code them properly.
I think there are one or two places here where it could have been helpful to offer more tightly-constrained “Other” fields (eg, to let people specify additional libraries / frameworks), but beyond that, unless the people processing the survey were prepared to do a great deal more work, it probably would have been misleading to offer open-ended questions.
bummer. missed it but probably wouldn’t have had a entry for the Eclipse Scala IDE again anyway
1 Like