The 2020 Julia language survey is live now

Announcements
1

The 2020 User & Developer Survey closes Sunday June 28. Complete it today!

Take the survey now in English, Spanish, Japanese or Chinese.

Findings will be presented at JuliaCon, just like last year.

36 Likes
5

For Questions 1 and 2, it’d be really helpful to default to “none”. Otherwise you have to click a LOT of radials…

13 Likes
6

Not sure if it’s too late to offer suggestions on the survey format, but for

  1. Thinking only about the TECHNICAL aspects or features of Julia, what are the TECHNICAL aspects or features you like LEAST about Julia? Please select all that apply.

I think there are a few missing options that I think at least some users express often. e.g.

  • Writing performant code is difficult or requires too much specialized knowledge
  • Lack of support for formal interfaces
  • Too much special, irregular syntax

I don’t necessarily agree with these points, but I see them all raised somewhat often and I think are all at least valid technical gripes.

Regarding

  1. Thinking only about the NON-TECHNICAL aspects or features of Julia, what are the NON-TECHNICAL aspects or features you like LEAST about Julia? Please select all that apply.

I don’t see how these entries are relevant as they don’t have anything to do with julia:

  • I am more comfortable in another language
  • I have more experience with another language
  • I enjoy exploring other languages

There are quite few non-technical gripes I could imagine people commonly holding that aren’t listed here such as

  • I feel that the developer team does not communicate it’s plans or intentions clearly enough
  • The community is too conservative with regards to language changes.
  • The community is too liberal with regards to language changes.
  • I feel uncomfortable sharing my opinions with the community

For

  1. Which of the following best describe your use of Julia? Please select all that apply.

I would list

  • I use it for play and hobby programming

For

  1. Do you identify as underrepresented in science or computing because of your:

I would list

  • Political or ideological beliefs
7 Likes
7

Looking forward to seeing the survey results! Julia is a great language and getting better by the day.

I’d echo Mason’s comment, but that there could be a text box under “Other” in certain categories for people to write in answers.

My write-in under least-favorite technical features would be “No in-language or built-in support for mutable, fixed size arrays.” I have used such arrays to great effect in C++ for optimizations. Glad to have StaticArrays.jl in Julia but hope to see it become part of the language in the future.

1 Like
8

We provided text boxes in various questions last year. But we got 100 pages of comments, and eventually, only the specific answers were valuable in analysis. That’s why we dropped the free-form comments this year.

Over such a large survey, we pretty much end up getting the union of every possible view.

-viral

6 Likes
9

3 posts were split to a new topic: Introductory textbooks

10

Hello !

I don’t know if this is the right place to ask this, but any particular reason the survey is not available in French ? If it’s because of lack of translator I’d be happy to provide French translation for future surveys ! :slight_smile:

4 Likes
11

I am sure that it is not available only because in the original discourse no one suggested or provided a French translation (I know it a little, but without confidence enough to provide one), many of us love the French language. In future surveys, I am sure that @viralbshah would be more than glad to have a French translation.

1 Like
12

In previous years I think there was a separate question that was basically “What non-technical reasons stop you from using Julia more?”, and these options were under that question instead. Explains why they now seem out of place when folded into the question about disliking the language.

14

This was the first year we experimented with translations. We basically picked languages where we believed there would be little overlap with English, and we wanted to keep the set small.

We didn’t do German and French mainly because we figured that those speakers can still take the English survey. With the experience gained this year, we definitely will include more languages next year!

-viral

6 Likes
16

That’s definitely valid for German, not necessarily for French.

(Actually I don’t understand anyway how is it possible to be a “Julia speaker” without being able to read documentation in English)

17

I am not sure about this. According to EuroStat, 80%+ of working age people with a tertirary education spoke at least one foreign language in France. For the majority this is English (for this you need to look at the raw data, it is not tabulated on that page). The same is true for all EU countries, except for the UK and Ireland, where English would be their native language.

I would assume that for languages spoken mostly in OECD countries, a translated survey is nice if someone wants to do it, but by no means essential for the typical audience Julia targets.

2 Likes
19

I found the survey to be extremely confusing to the point where I didn’t finish the questions nor submitted my response.

In particular I feel like the response alternatives are not fitting the questions. Here are some concrete points about some of the more confusing (to me) things:

  • For Q1 (“How frequently do you use each of the following languages?”) I think it would be better with alteratives “Daily”, “Weekly”, “Monthly”, “More seldom” instead of “Great deal”, “Some” and “None” which seems to be very much up for interpretation.

  • For Q4 (“Thinking only about the NON-TECHNICAL aspects or features of Julia, what are the aspects or features you like MOST about Julia?”) the response alternative “Learning a new language - I like learning new languages” should be “Julia is a young language” or something instead. “I like learning new language” is not a feature of Julia (although perhaps it is a feature of the person responding??).

  • The response alternatives to Q5 (“Thinking only about the TECHNICAL aspects or features of Julia, what are the aspects or features you like LEAST about Julia?”) seems very “suggestive”. All of the alternatives feels like responses/comments you would get if you ask “What do you dislike about Julia?” with a free-form response field instead of “features or aspects” that you can like/dislike. Just to comment on some examples from the response alternatives:

    • “Not stable enough”: Maybe better phrased as “Stability” to make it a “aspect that you can dislike”.
    • “Doesn’t have all the packages I need”: Maybe better stated as just “Package ecosystem”. If “Doesn’t have all the packages I need” is an aspect about the language, then if you DO think it has the packages you need, should you select this one since it is double negative?
    • “Package manager is confusing, difficult or doesn’t do what I expect it to do”: Is this really a feature or established aspect about Julia…
    • etc…

  • For Q6 it is essentially the same as the previous points: most of the response alteratives are not “features or aspects” that you can like/dislike which makes it really hard to connect the response alternatives to the question. They all seem more like statements you can agree/disagree with instead of things you can like/dislike. For example: “I don’t like learning a new language” doesn’t seem like a feature of the language, but more like a respons to “Why are you not using Julia?”.

  • For the question pairs about things you like/dislike (Q3+Q5 and Q4+Q6) I think it would be much better to just have a single list of (actual) “features and aspects” to be used in both of them. For example why can I like “Speed and performance” but not dislike it? It seems very “guided” somehow as if “Speed and performance” is an established truth that you can only agree with.

Hopefully this feedback is useful and the survey can be improved for next year and maybe I can answer then!

9 Likes
20

It’s nice that we totally agree here, aren’t we :stuck_out_tongue_winking_eye:

And thank you for the reference. Now, approx. 83% of working age people with a tertirary education “know” at least one foreign language. Of these, some (esp. in the border regions) would speak only Spanisch, Italian, or German. Then, by the level of command in France on the average 50% know this foreign language only to the “basic” level. So we may assume that like 30…50% of professionals in France do not know English sufficently good - which is my personal experience, too.

21

That may be, but I would be very surprised about people

  1. with a tertiary education and
  2. doing the kind of programming that Julia is suited for

not knowing English (of course I don’t have data on this).

Most of the French people I know are economists, and of course all of them speak English — maybe I have a biased sample.

In any case, I think that the bottom line is that if someone provides a translation, it will be included.

22

May be erlang / elixir should be taken into consideration too