Results from MDN Developer Survey

17 Sep 2019


tantek, jorydotcom, CharlesHall, dom, zcorpan, cbiesinger, Boaz, heejin, tobie, karl, JohnJansen, Dongwoo
Kadir Topal


<scribe> ScribeNick: jorydotcom

kadir: Introduces the MDN product advisory board
... they helped put this together as an effort of the MDN not just Mozilla
... recap is that we published this in July to reach a broad audience of developere and designers to understand their needs
... the goal is to do this annually
... we want to understand and prioritize painpoints over time
... 29 queations, localized into 8 languages. fielded for 4 weeks. 76k responses
... more than 28k people completed the survey from 173 countries

<dom> [thinking aloud - since survey will be run in northern Spring next year, we can expect an even broader set of participants compared to mid-Summer (northern)]

kadir: 10k hours of developer time to complete

<JohnJansen> please paste a link to the report

kadir: the goal of the session is to walk through the survey and show you some results that we have
... and understand what you want to see for segmentation
... what can we do differently for the next version of the survey

Dom_for_JohnJansen Can we share the report?

kadir: not now, hope to have full public report published by end of October beginning of November

<zghadyali> I have a question. Will the published report include the raw data of the responses to the survey?

kadir: we had 29 questions, 2 were free form

<dom> qq+ zghadyali

kadir: they were aboutu paintpoints and what they would like to be able to do
... we got more than 12k responses in 9 languages so we need more time to analyze, taht will be part of the full report

zghadyali: will the report be published with teh raw data?

kadir: we will only be able to do a pdf report for budget reasons, hope to have budget for next time
... this is the first iteration, we really just wanted to get it out so we can take into account

tobie: was it a legal problem?

kadir: no it was just technical. We didn't have the budget. so we didn't even ask legally. hope to have that next year

tobie: I think it would be valuable to consider next and last year's data if you can

kadir: makes sense!

karl: I read the survey

<zcorpan> https://docs.google.com/spreadsheets/d/11jMfXWxw9bBZeCG8PYSoWsQZnRd2OBGSKSEwSF8nd0Q/edit#gid=0 - please add your IRC nicks here to help the scribe

karl: I don't remember if there was a distinction between front and backend devs

<Zakim> karl, you wanted to ask about backend development survey (http, server-side framework).

karl: did we focus enough on the questions/needs of backend devs?

kadir: maybe to answer this lets look at who we surveyd
... we wanted to talk to people who actually code
... that was the first disqualifying question. a pure vis designer wouldn't qualify
... we do ask people about this later

Boaz: a format could be that we go through all of these now and have a group discussion after?
... or the 5-10 key ones

<Travis> +1 let's see the data!!

jgraham: keep talking and save questions for the end

kadir: I wanted to walk quickly through this and do some live segmentations, we might be able to do that for some of the segments that I have pre-prepped
... the first one is the qualifying question: what do people do when they say they code/ code and design
... how many hours to do you spend writing code? 80% spend more than 20 hours a week
... next one is the type of dev you are. How do you ask whether you are a fe dev, backend, fullstack? there aren't shared established terms
... it was a difficult disctintion to make, pilot testers got it but it may not have been clear in other langs
... 10% said they were backend developers. the main channel we sent this out in is MDN, so it makes sense we may not have reached a backend community

iank_: is there a difference between vis design and code design

kadir: we hadn't thought about that
... we saw roughly equally proportions JS/HTML/CSS
... 50% had less than 2 years experience
... we think because of MDN's reach and the reach of the PAB we think this is pretty representative
... this is something to keep in mind

<dom> [there may be a bias in that MDN attracts beginners more / more often than experienced people]

kadir: sometimes people don't know that browsers have versions,

<CharlesHall> so there is zero excuse for that experience demographic to not be aware of accessibility. 20+ years of a spec for it.

kadir: keep this data in mind, about half have less than 2 years and 2/3rds have less than 5
... something went wrong on this one (gender demo slide)
... Gender question, we did not get a good representation 86% of respondents identified as male. We think in the Us it should be 20% at least
... for next year we want to do better and to compensate for this bias.
... the needs and painpoints might look different. We hope to have budget to address this next year
... what country do you live in? This is roughly what we expected with the expecption of china
... it's hard to parse out to know if we got more participation because we localized into those languages.
... this is not just US or EU, we really wanted diverse representation

dom: can people talk/tweet about this?

kadir: yes, recall that this is raw data and is just an indication of where things are going
... so when you segment this you partially get different results
... we already talked about the first 5 and how interop is a big problem

iank_: what are the plans for next year to understand the interop problems more?

kadir: once we have analyzed the free form results we will see if there are additional items
... interop is at the top of the list, let's look at the bottom
... making sites accessible. This doesn't mean it's not a problem, it may mean that they don't focus on it or aware of the issues
... they may not know what they don't know
... that's why we do the interviews
... keeping up with changes to the web platform, that was also a surprise. It's evidently not a big isssue
... but compared to keeping up with the tools and frameworks, that was a bigger issue

iank_: do you find the interop quesiton when you look at other countries?

joe: I wonder if people not keeping up with the platforms is related to not knowing that browsers have versions

zcorpan: most of the respondents are pretty new. If you've been a web dev for 1-5 years you may not id it as a problem

kadir: experience level is something that we segmented by
... (shares slides)
... it's the same thing. Interop is the biggest issue whether you just started or have been on for 10 years
... also keeping up with the frameworks and tools

jgraham: arguably there's a degneracy between these because a lot of devs use frameworks to leverage the web platform

kadir: how many people use these vs. abstract that away? the platform underneath may not be of concern to you?
... we asked them how much do you think this represents your painpoints. 72% agree or strongly agree that this was a good representation
... it makes us happy that this was a good representation for most people. that's why we had the free form to get the 8% who didn't agree
... a lot of it was related to non-tech issues like office politics, which may well top the list if we had added it
... but this was not the purpose on the survey
... how would you rate your overall satisfaction? 77% are satisfid or very satisfied. we are surprised because people like complaining!
... most people are actually quite happy
... we asked abotu whether they were doing this for personal or professional. most of them are doing it for their job
... almost half the participants had a cs or engineering degree
... almost everyone also added that they were self-taught
... we didn't expect that
... what team structure best describes your situation? d you work alone or with others?
... 66% said they work in an org with other devs, 10% said they are the sole dev in an org
... which platforms do you target? this was another surprise
... 95% target the desktop, only 62% target the mobile web?

iank_ is this skewed for the US?

kadir: no

cbiesinger: did you ask about enterprise vs. the consumer web?

kadir: no
... what browsers do you support?
... almost everyone came out with I hate IE 11
... but there's more to it. the surprising thing is that safari is supported mroe than safarie IOS which is dispropotionate to marketshare
... follow up to that was which browswers give you the biggest issues? Chrome is at the top of the list, but keep in mind most people aren't supporting the other browsers
... the chart is difficult to read,
... when IE was ranked, it was the biggest issue. whereas with chrome it was ranked but gave the least problems
... if people have to support safari IOS it is a bigger issue
... the browser people primarily use for development is chrome, but we were surprised by chrome for andriod so we think there may be an issue with this question
... so we look into whether this is understood

iank_ this doesn't surprise us

Kadir: we were looking to understand what dev tools people have, this may have been unclear in terms of how we phrased the question

<CharlesHall> ? will there be opportunity to provide feedback on the wording of the questions and research process for 2020?

Kadir: which browsers do you test in?
... chrome, firefox, edge
... Installable web applications. this is about PWAs. when we asked about PWAs people didn't really know what it means
... so we asked a question with a definitaion about PWAs, and with that definition, 29% said yes, and 10% said IDK
... when a new tech becomes available what are your barriers for adoption. It's usually documentation but this was interop by far

gsnedders could people be interpretting outdated browsers / legacy browsers into this question?

kadir: maybe, we should look into that

gsnedders: is tehre any support of implenation differences?

kadir: we should look into that
... we asked people which languages they use, this is not a surprise. WASM was at 3%. this queastion was asked so we could segment other questions by it

Travis: what's the answer?

kadir: i don't know. this is a very early preview of thedataset
... we asked where are you using JS? almost no one is using it on the server exclusively?
... what are the biggest painpoints in JS development? feature adoption and performance
... people don't really differentiate between the core lang and the APIS, and it changes over time
... next issue there are two things taht stand out. performance is not surprising.
... with HTML, there were no pain points! second one is lack of browser adoption and inability to customize components. THere was a difference in China
... in CSS it wasn't interop, it was challenges with layout.

<tantek> also surprised to see no pain points here

<karl> [probably the webdevs having difficulties to implement what the designers ask them to do.]

<tantek> [yeah]

kadir: WASM it was lack of debugging/tooling

<dontcallmeDOM> zcorpan, close the queue

kadir: do you get to choose the tech you use? 2/3rds to get to choose what they use. Most people do get to choose. Devs are the right people to target

Boaz conversation of use cases. There's active work weher these insights are going to be useful

scribe: one of our action items was to be able to give efeedback on this next year so we can help compute priority
... excited to collaborate with you on this

<Zakim> tobie, you wanted to ask whether data about framework usage was collected?

kadir: offer stands if you need a cut of this data so you can get if you need

tobie: it would be interesting to have data about specific JS and CSS frameworks? what frameworks and whether they are or not

kadir: shows the next sli ewith some of that data
... for next year we want to understand the buckets better so we can ask better questions

JohnJansen: the PWA data that you saw is verified by our crawler. It seems to match what you do. SW debugging is a bad experience. A future question could be more nuanced

kadir: abso;luutely

jgraham: you were surprosed about the testing results. I think it depends what youmean by testing. Pulling apart those things would be better. Then you can pull out which tools you are using
... I think for this kind of question it might be more interesting, if you say you're using CI which tool are you using

kadir: very true
... this is very helpful, i fyou think of additional questions, go to your PAB person or come talk to me or email me at atopal@mozilla.org
... any recommendations for the next iterations are welcome, and your feedback as well

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2019/09/18 03:03:11 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.154  of Date: 2018/09/25 16:35:56  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/??/Dom_for_JohnJansen/
Succeeded: s/be/backend/
Succeeded: s/???/iank_/
Succeeded: s/Chinca/china/
Succeeded: s/iank_/iank_:/
Succeeded: s/actua;ly/actually/
Succeeded: s/???/cbiesinger/
Succeeded: s/Chinca/China/
Present: tantek jorydotcom CharlesHall dom zcorpan cbiesinger Boaz heejin tobie karl JohnJansen Dongwoo
Found ScribeNick: jorydotcom
Inferring Scribes: jorydotcom

WARNING: No "Topic:" lines found.

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

People with action items: 

WARNING: No "Topic: ..." lines found!  
Resulting HTML may have an empty (invalid) <ol>...</ol>.

Explanation: "Topic: ..." lines are used to indicate the start of 
new discussion topics or agenda items, such as:
<dbooth> Topic: Review of Amy's report

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)

[End of scribe.perl diagnostic output]