Input: 2014 retrospective
What is it?
The purpose of Input is to collect actionable feedback from our user base across each channel of our software development process. The application collects feedback and offers a set of analysis methods for looking at the resulting data.
This is my 2014 retrospective for Input.
2014 was a big year for Input. This is the year we hit a point where the project is mature and stable. This is the year we worked on making it easier for new people to start working on Input. This is the year we fixed all the issues that made it difficult for Input to support multiple products. This is the year we created an API that allowed anyone to access the data to let them write their own dashboards. This is the year we implemented a translation infrastructure so that non-English feedback can flow through our tools (previously, it was handled manually). It was a big year.
Bugzilla and git stats
Let's look at some Bugzilla and git stats for the year:
Year 2014 (2014-01-01 -> 2014-12-31) ==================================== Bugzilla ======== Bugs created: 277 Creators: Will Kahn-Greene [:willkg] : 240 [:Cww] : 8 Ricky Rosario [:rrosario, :r1c : 4 Swarnava Sengupta (:Swarnava) : 3 Nicolas Perriault (:NiKo`) — n : 1 vivek :) : 1 Trif Andrei-Alin[:AlinT] : 1 Kohei Yoshino [:kohei] : 1 t.schmittlauch+persona : 1 Mike "Pomax" Kamermans [:pomax : 1 Mark Filipak : 1 Matt Grimes [:Matt_G] : 1 Padraic Harley [:thelodger] : 1 deshrajdry : 1 Matt Brubeck (:mbrubeck) : 1 Christian Corrodi : 1 phaneendra.chiruvella : 1 vivek.kiran : 1 lgp171188 : 1 me : 1 Matthew N. [:MattN] : 1 David Weir (satdav) : 1 Hal Wine [:hwine] (use needinf : 1 Ian Kronquist (:muricula) : 1 EventHorizon : 1 Laura Thomson :laura : 1 Bugs resolved: 245 : 1 WONTFIX : 17 WORKSFORME : 8 DUPLICATE : 13 FIXED : 206 Tracebacks : 9 Research : 5 Tracker : 11 Research bugs: 5 788597: [research] Should we use a stacked graph on the dashboard? 792976: [research] swap jingo-minify for django-compressor and django-jingo-offline-compressor 889370: [research] morelikethis for feedback responses 990774: [research] Investigate database schema changes for easier extensibility 1048459: [research] check if updating from a tarball in a virtual environment with peep removes files Tracker bugs: 11 963275: [tracker] support Metro 964260: [tracker] overhaul mobile and desktop feedback forms 965791: [tracker] implement product table 966425: [tracker] create analyzer search view 985645: [tracker] Upgrade to Django 1.6 988612: [tracker] product dashboards 1040773: [tracker] dashboards for everyone 1042669: [tracker] reduce contributor pain 1052459: [tracker] heartbeat v1 1062429: [tracker] integrate spicedham classifier 1081412: [tracker] add tests for fjord_utils.js functions Resolvers: 14 Will Kahn-Greene [:willkg] : 209 lgp171188 : 7 aokoye : 7 Ian Kronquist (:muricula) : 5 Ricky Rosario [:rrosario, :r1c : 4 Joshua Smith [:joshua-s] : 3 Swarnava Sengupta (:Swarnava) : 2 Rehan Dalal [:rehan, :rdalal] : 2 Ruben Vereecken : 1 Rob : 1 cww : 1 Schalk Neethling [:espressive] : 1 bhargav.kowshik : 1 Anna : 1 Commenters: 65 willkg : 1261 rrosario : 19 mgrimes : 17 cww : 14 MattN+bmo : 11 mcooper : 11 thewanuki : 7 aokoye : 7 swarnavasengupta : 7 padraic : 6 iankronquist : 6 MarkFilipak.mozilla : 5 nicolas.barbulesco : 5 schalk.neethling.bugs : 5 hwine : 4 rdalal : 4 lgp171188 : 4 stephen.donner : 4 educmale : 4 anthony : 4 mbrubeck : 4 ms.annaphilips : 4 nigelbabu : 3 rdaub : 3 me+bugzilla : 3 pomax : 3 deshrajdry : 3 bhargav.kowshik : 3 rubenvereecken : 3 rrayborn : 3 rajul.iitkgp : 2 kbrosnan : 2 christian : 2 glind : 2 peterbe : 2 hcondei : 2 david.weir : 2 dron.rathore : 1 pradeeppaddy : 1 tdowner : 1 margaret.leibovic : 1 aaron.train : 1 scottstensland : 1 senicar : 1 phaneendra.chiruvella : 1 trifandreialin : 1 jesse : 1 t.schmittlauch+persona : 1 viveknjadhav19 : 1 mozaakash : 1 me : 1 glob : 1 cturra : 1 nperriault : 1 rmcguigan : 1 vivek.kiran : 1 kdurant35rules : 1 vega.james : 1 fbraun : 1 326374 : 1 vivekb.balakrishnan : 1 mhoye : 1 mluna : 1 feer56 : 1 lorenzo567 : 1 git === Total commits: 653 Will Kahn-Greene : 590 (+244588, -206236, files 2774) Adam Okoye : 15 (+128, -38, files 39) L. Guruprasad : 13 (+203, -39, files 17) Ricky Rosario : 10 (+165, -350, files 30) Ian Kronquist : 7 (+201, -53, files 11) ossreleasefeed : 3 (+197, -42, files 9) Bhargav Kowshik : 3 (+127, -13, files 14) Joshua Smith : 3 (+91, -36, files 10) Swarnava Sengupta : 2 (+2, -2, files 2) Rehan Dalal : 2 (+315, -169, files 13) Anna Philips : 1 (+367, -3, files 12) Ruben Vereecken : 1 (+34, -14, files 6) Gregg Lind : 1 (+9, -8, files 3) Deshraj Yadav : 1 (+1, -1, files 1) aokoye : 1 (+2, -2, files 1) Total lines added: 246430 Total lines deleted: 207006 Total files changed: 2942 Everyone ======== 326374 aaron.train Adam Okoye Anna Anna Philips anthony Bhargav Kowshik christian Christian Corrodi cturra cww David Weir Deshraj Yadav deshrajdry dron.rathore educmale EventHorizon fbraun feer56 glob Gregg Lind Hal Wine hcondei Ian Kronquist jesse Joshua Smith kbrosnan kdurant35rules Kohei Yoshino [:kohei] L. Guruprasad Laura Thomson :laura lorenzo567 margaret.leibovic Mark Filipak Matt Brubeck (:mbrubeck) Matt Grimes Matthew N. [:MattN] Mike Cooper me me+bugzilla mhoye Mike "Pomax" Kamermans [:pomax] mluna mozaakash ms.annaphilips Nicolas Perriault nicolas.barbulesco nigelbabu Padraic Harley [:thelodger] peterbe phaneendra.chiruvella pomax pradeeppaddy rajul.iitkgp Ralph Daub Rehan Dalal Ricky Rosario rmcguigan Rob Robert Rayborn Ruben Vereecken Schalk Neethling scottstensland senicar Stephen Donner Swarnava Sengupta t.schmittlauch+persona Tyler Downer thewanuki Trif Andrei-Alin vega.james vivek.kiran vivekb.balakrishnan viveknjadhav19 Will Kahn-Greene
Some observations:
-
In 2013, we resolved more bugs than we created partially because we closed a bunch of bugs related to the old Input that weren't relevant anymore.
In 2014, we created more bugs than we closed by 10%. I think that's about what we want.
-
15 people had git commits. 26 people created bugs. 14 people resolved bugs. 65 people commented on bugs.
One thing I don't have is counts for who helped translate strings which is a really important part of development. My apologies.
Of all those people, only 1 is a "core developer"--that's me. Everyone else contributed their time and energies towards making Input better. I really appreciate that. Thank you!
That's the stats!
Accomplishments
Things we did in 2014:
Site health dashboard: I wrote a site health dashboard that helps me understand how the site is performing before and after deployments as well as after releases and other events.
Client side smoke tests: I wrote smoke tests for the client side. I based it on the defunct input-tests code that QA was maintaining up until we rewrote Input. The smoke tests have been invaluable for reducing/eliminating data-loss bugs.
Vagrant: I took some inspiration from Erik Rose and DXR and wrote a Vagrant provisioning shell script. This includes a docs overhaul as well.
Automated translation system (human and machine): I wrote an automated translation system. It's generalized so that it isn't model/field specific. It's also generalized so that we can add plugins for other translation systems. It's currently got plugins for Dennis, Gengo machine translation and Gengo human translation. We're machine translating most incoming feedback. We're human translating Firefox OS feedback. This was a HUGE project, but it's been immensely valuable.
Better query syntax: We were upgraded to Elasticsearch 0.90.10. I
switched the query syntax for the dashboard search field to use
Elasticsearch simple_query_string
. That allows users to express
search queries they weren't previously able to express.
utm_source and utm_campaign handling: I finished the support for
handling utm_source
and utm_campaign
querystring
parameters. This allows us to differentiate between organic feedback
and non-organic feedback.
More like this: I added a "more like this" section to the response view. This makes it possible for UA analyzers to look at a response and see other responses that are similar.
Dashboards for everyone: We wrote an API and some compelling examples of dashboards you can build using the API. It's being used in a few places now. We'll grow it going forward as needs arise. I'm pretty psyched about this since it makes it possible for people with needs to help themselves and not have to wait for me to get around to their work.
Dashboards for everyone project plan.
Vagrant: We took the work I did last quarter and improved upon it, rewrote the docs and have a decent Vagrant setup now. L. Guruprasad improved on this and the documentation and setting up a Vagrant-based vm for Input development is much easier.
Reduce contributor pain project plan.
Abuse detection: Ian spent his internship working on an abuse classifier so that we can more proactively detect and prevent abusive feedback from littering Input. We gathered some interesting data and the next step is probably to change the approach we used and apply some more complex ML things to the problem. The key here is that we want to detect abuse with confidence and not accidentally catch swaths of non-abuse. Input feedback has some peculiar properties that make this difficult.
Reduce the abuse project plan.
Loop support: Loop is now using Input for user sentiment feedback.
Heartbeat support: User Advocacy is working on a project to give us a better baseline for user sentiment. This project was titled Heartbeat, but I'm not sure whether that'll change or not. Regardless, we added support for the initial prototype.
Data retention policy: We've been talking about a data retention policy for some time. We decided on one, finalized it and codified it in code.
Shed the last vestiges of Playdoh and funfactory: We shed the last bits of Playdoh and funfactory. Input uses the same protections and security decisions those two projects enforced, but without being tied to some of the infrastructure decisions. This made it easier to switch to peep-based requirements management.
Switched to FactoryBoy and overhauled tests: Tests run pretty fast in Fjord now. We switched to FactoryBoy, so writing model-based tests is a lot easier than the stuff we had before.
Python 2.7: Input is now running on Python 2.7. Thank you, Jake!
Remote troubleshooting data capture: The generic feedback form which is hosted on Input now has a section allowing users to opt-in to sending data about their browser along with their feedback. This data is crucial to helping us suss out problems with video playback, graphics cards/drivers and malicious addons.
This code is still "alpha". We'll be finishing it up in 2015q1.
Remote troubleshooting data capture project plan.
Heartbeat v1 and v2: People leave feedback on Input primarily when they're frustrated with something. Because of this, the sentiment numbers we get on Input tilt heavily negative and only represent people who are frustrated and were able to find the feedback form. Heartbeat will give us sentiment data that's more representative of our entire user base.
As a stop-gap to get the project going, Input is the backend collecting all the Heartbeat data. We rewrote the Heartbeat-related code for Heartbeat v2 in 2014q4.
Feedback form overhaul: We rewrote the feedback form to clean up the text, reduce confusion about what data is made public and what data is kept private, reduce the number of steps to leave feedback and improve the form for both desktop and mobile devices.
Feedback form overhaul project plan.
We also fixed the form so it supports multiple products because we're collecting feedback for multiple products on Input now.
Thoughts
Looking at the stats above, it's pretty clear that this is predominantly a one-person project. Ricky, Mike and Rehan do all my code review. Without them, things would be a lot worse.
Having said that, the "baby-factor" stinks on this project. That's something I'm going to work on over 2015:
improving the documentation and making sure all the important things are covered or easily figured out
improving the project planning and making sure most/all project planning is done in public or at least has a publicly available record
reducing the complexity of the application, reducing dependencies and making the parts less tangled
reducing the number of things that make Input a "special snowflake" in regards to other sites my peers maintain
Want to be a part of the team?
Input is a great little Django application that collects feedback for Mozilla products. It's an important part of the Mozilla product ecosystem. Working on it helps millions of people. If you're interested in being a part of the team that develops it, here are some helpful links:
We maintain a Get Involved page.
We hang out on
#input
on irc.mozilla.org.We have an input-dev mailing list.
Update April 21st, 2015
LGuruprasad found a bug in the script that caused commits-by-author information to be wrong. Fixed the script and updated the stats!