I have started work on building the charts, it's going to take awhile so I'm going to post them here as I finish each to get feedback earlier.
@XAMPPRocky Can we sort the y-axis there so that we have some ordering to them? Otherwise it's sort of all over the place and hard to compare I feel
I would also personally not use a logarithmic scale here, I think it hides the vast disparity too much, instead maybe we should consider throwing the really big ones on a separate chart or so
but that'll be easier to tell once its sorted
I'm not sure the disparity is meaningful in this question. This was a question where people could fill in multiple answers, so for example most people put in "Programmer / Software Engineer" + other titles, so there's multiple entries for a single person. That being said I'm not data scientist or analyst, I'm mostly going off what I remember from college, so if someone feels strongly about it, we can change it. Here's both sorted.
(Also quite surprised architect is second)
@XAMPPRocky As someone who picked
Architect for that question, I'm not too surprised. It's a very common job title for more senior software engineers.
hm, so for multiple entries I'm not sure what the best thing to do is. It does feel a bit odd to not include that information some how
do we have a loose sense of how many people did?
If it's like 100 then I'm not too worried
@simulacrum I can share the sheet with you but just glancing at the raw data it seems like most people filled multiple roles. Screenshot-2020-04-08-at-16.08.59.png
I don't have time to look into this much
:shrug: I'm fine with just sorting I think
@Josh Triplett Yeah I forgot that was a title that programmers gave themselves, I initially thought there was a silent majority using Rust in architecture firms. I think I'll change the name to Systems Architect to make that clearer.
now that I read again those labels: "Computer Administration" equals to a system administrator? e.g. someone that takes care of a local network for a company?
@apiraino In which graph?
the latest you posted (What industries do you work ...)
Ty soooo much for doing this, @XAMPPRocky !
@apiraino Ah, yes I believe that’s like IT administration
that's a really interesting chart
the "job title" one feels a bit less revealing
I wonder @XAMPPRocky if we can get a similar chart of industries for folks who listed "rust is not used at my job" as the reason for not using rust
Yep that's next my on list I should have it finished by the by the end of the day eu time.
... I see my job at work in these numbers
@Nell Shamrell-Harrington That's the last of the graphs from the codebook @Jasun Carr sent.
@XAMPPRocky what would you need to regenerate the graphs that Mark had feedback on in https://github.com/rust-lang/blog.rust-lang.org/pull/555?
(and do you have the time to do so?)
I can do it later but you can download them as SVGs in that sheet.
All charts are in the “X-Clean” sheets
Sweet! Looks like I don't have access to that doc - just requested it
@Nell Shamrell-Harrington Sorry! (I hate GSuite). The link should be public now
Got it! TY so much for doing this work!
BTW, a million thanks to both of you :pray:
@Nell Shamrell-Harrington I've redone all of the charts @Jasun Carr sent over here. https://drive.google.com/file/d/1xX5_RcFLQIASBefTgLqqGuWOnAOFKdRi/view?usp=sharing
ty ty ty! I'll get those in the post today!
The increased interop is pretty surprising, I wasn't expecting "R" to be second.
These two are also pretty interesting to compare. "Compile times" is one of the top things people believe we need to improve, but when looking at "How can we make Rust more appealing to your company?" it's much lower on the list.
Also interesting that error handling is at the bottom of the list when its been one of the more active topics lately.
Yeah, that is all quite interesting I agree
Just finished updating https://github.com/rust-lang/blog.rust-lang.org/pull/555 with the new graphs and a few more edits
Unless there are any major issues with the post in the state that it is...I think we should move forward and post this and, if there are places we want to dig deeper, address them in follow up posts
@Nell Shamrell-Harrington Thank you so much for all the extensive work on this!
Likewise @XAMPPRocky and others.
Anyone able to respond to this? https://twitter.com/tpflug/status/1285851190161551360
@grhmc @steveklabnik @garbas @Argorak @ManishEarth @ag_dubs @jntrnr Can i somewhere find the exact questions that were asked in the last survey? Also i don’t know whom i should tell this but: i have bad eyesight and it’s almost impossible for me to recognize any of the charts on the results page ;(- Tobi (@tpflug)
Hmm I'm not sure how we can make that available
maybe we can reopen the survey and put a banner on it saying "we're not collecting responses but feel free to click through"?
Well it is getting to close to the time to conduct this year's survey, so I'm not sure how much more work on last year's would be worth it. Seems like something to try to make available this go around though.
/me is scared that time flies that fast
Maybe we can try put all the questions on
Pontoon platform this time... So after the localization work is done, the questions are still there, and people can read them from there
We'll also be able to reuse some old translations of questions if some questions are the same.
@nikomatsakis I think making the questions available is not a problem, it's just manual labor of copying them out. Results is a different thing.
I'd love to track the survey translation through pontoon, knowing when the translations were ready was a bit of a pain point last year and using pontoon would address that.
egads, I guess we are getting close aren't we :)
I will see if I can finagle to work with @Jasun Carr again
it'd be good for us to think about who wants to be involved in the effort this time around
@nikomatsakis You rang? Keep me in the loop as things start gearing up for the next survey.
@XAMPPRocky & @Charles Lew - on the subject of translations, some work to standardize the questions/response options as they go through translation (both directions) would be good as well. Small differences in phrasing/translation became a bit of a hurdle last year - and some of the languages may have been either missing questions or had them in the wrong order (problematic when combining multiple spreadsheets and the responses didn't align with the questions or aligned with other questions).