SpeedUpAmerica - Tickets

This document outlines work that needs to be done and their matching tickets in GitHub Issues. The work outline in this document are mission critical and will receive story points. Each section is a high level objective and will contain work broken down in to bite sized checks of work we can put in GitHub issues.

A high level overview of the work in this document can be found at +SpeedUpAmerica - Moving Forward.

Priorities (2019-07-07)

~~Vector Tilesets and adding Washington State.~~

Improve UX of data exploration

Rewrite Frontend

Narrow result set based on viewport bounding box + buffer

GIS Service & Census Blocks

Maintenance work on v1 (current solution)

Improve logging 1SP - https://github.com/roidrage/lograge https://github.com/Hack4Eugene/SpeedUpAmerica/issues/112

Fix Verizon and XS Media in submissions.provider https://github.com/Hack4Eugene/SpeedUpAmerica/issues/122

Decide what centralized logging system to use

Feed CloudWatch logs into centralized logging system

Ensure we have a dashboard showing Golden Signals

Improve Onboarding document(s) - add migrator info, links to key documents 1SP https://github.com/Hack4Eugene/SpeedUpAmerica/issues/111

Track accuracy of collected geo location https://github.com/Hack4Eugene/SpeedUpAmerica/issues/118

Investigate and fix Safari geolocation bug - https://github.com/Hack4Eugene/SpeedUpAmerica/issues/110

Change Explore the Data, statistical boundaries, to be a drop down menu list instead of the two fields (Zip and Census). To the right of that future drop down menu is the entry field. https://github.com/Hack4Eugene/SpeedUpAmerica/issues/115

Enhance Explore the Data, Internet Service Providers, to auto populate top 3 ISP’s for the defined statistical boundary. Where “top” is defined as median download speed. Include user’s ISP as a 4th ISP https://github.com/Hack4Eugene/SpeedUpAmerica/issues/179

Make sure we are setting client_application to speedupamerica for test run through our site. May requiring updating NDT JS library. (SP1) https://github.com/Hack4Eugene/SpeedUpAmerica/issues/178

Add “Data Source” filter capabilities to both the Map and Explore the Data. Default will be “All”. Drop down filter list options to include M-Lab, SpeedUpAmerica, TestIt et all.

Couldn’t load preview

The file may have been moved or deleted, or is temporarily unavailable.

https://paper-attachments.dropboxusercontent.com/s_2505669AAA2589D9B8E9E39E366078E758D792897038E913E97A5C9B0078AAE5_1563392699421_image.png

Change the word Megabyte to Megabit. (0.5P)

Remove obfuscate_id from repo so we can upgrade Rails (1 SP) https://github.com/Hack4Eugene/SpeedUpAmerica/issues/196

Upgrade Rails to 5.2 (3 SP)

Upgrade Rails to 6.0 (2 SP)

Vector tilesets & additional states

Create vector tileset for Oregon Census Tracts, document process, and store key files in public S3 bucket https://github.com/Hack4Eugene/SpeedUpAmerica/issues/114

Investigate and recommend a JSON data structure that supports easy joining of aggregated submissions on to vector tilesets

Create vector tileset for Oregon ZCTAs, and store key files in public S3 bucket (1SP)

Upgrade v1 maps to Mapbox GL JS (3SP) https://github.com/Hack4Eugene/SpeedUpAmerica/issues/113

Implement endpoint that aggregates submissions using existing filters and returns JSON data in the structure decided https://github.com/Hack4Eugene/SpeedUpAmerica/issues/133

Update v1 Mapbox GL JS map to use vector tilsets and join new group by endpoint data on to it (3SP)

Update vector tilesets (Zip and Census Tracts) to include Washington and Idaho (2SP)

Load Washington and Idaho boundaries into census_boundaries and zip_boundaries tables and update data import process to include the FIPS codes for these states

Implement caching of statistics in new table https://github.com/Hack4Eugene/SpeedUpAmerica/issues/153 @Ryan O

Update maps to handle additional states (0.5SP) https://github.com/Hack4Eugene/SpeedUpAmerica/issues/153 @Ryan O

Delete and Reload M-Lab data for last 13 months so that we have Oregon, Washington, and Idaho cleanly loaded (0.5SP) @Ryan O

Contact improvements

What types of ZenDesk tickets do we want to create and who (groups?) gets an email? Question & Problem

Create contact form that allows users to submitted a ticket for the types we need (2SP) https://github.com/Hack4Eugene/SpeedUpAmerica/issues/137

Create endpoint that takes a POST and create a ticket (2SP)

Service rewrite

Decide what language/platform to use for BFF → Node.js

Decide if we will use traditional pages or SPA → Traditional

Create new service repo and create basic setup

Choose and implement testing framework https://github.com/Hack4Eugene/speedupamerica-frontend/issues/3

Coverage reports https://github.com/Hack4Eugene/speedupamerica-frontend/issues/5

Choose and implement structured logger https://github.com/Hack4Eugene/speedupamerica-frontend/issues/2

Implement database client (must support connection retry and pooling)

Implement HTTP service, router, and request logging (/ should be empty page)

Make sure the dev/test (non-prod) environment have robots.txt denys all and noindex meta tags

Send non-dev env application and request logs to Loggly

Setup CodePipeline for deployments, and ECS services for Test env

Implement root layout file to use for all HTML pages

Decide what endpoints we need to implement and define tickets

Decide what views/pages we need and define tickets

Add pages and endpoints (placeholder for a lot of tickets/work)

Get V2 at feature parity with V1 (placeholder for a lot of tickets/work)

Once endpoints and pages are implemented, setup Producation ECS and and CodePipeline deployment

Make sure that production has allow robot.txt and doesn’t have noindex meta tags (Test must still have deny and noindex otherwise it will show up on search engines)

Switch DNS for site to use V2 application

Start rewrite on all data import and processing related rake tasks (placeholder for a lot of tickets/work)

Endpoints

TODO

Views/pages

TODO

Add “State” filter capabilities to Map, with default being the state the visitor is coming from.

Add “Internet service satisfaction” filter capabilities as seen on speedupsanjose.com

GIS extension & additional boundaries

Add new geometry column and use a migration to convert WKT strings in existing column to the new column. Update data loading scripts to use new column. @Ryan O

Write and deploy a migration that adds a new point column to submissions. Update and deploy change to frontend that writes point data to both lat+long and the new point column. Write and deploy a 2nd migration that that converts existing lat+lng fields into new point column. @Ryan O

Load Census Blocks into GIS service, will likely require additional work optimizing point-in-polygon lookups @Ryan O

~~Create vector tileset of Census Blocks for Oregon, Washington, and Idaho~~ @Ryan O

Update maps to handle the additional boundary and backfill submissions table with Census Blocks @Ryan O

Update Frontend V2 to use point-in-ploygon looks using new columns to set Zip and Census Tracts. @Ryan O

​​Maintenance work on v1 (current solution)

​​Vector tilesets & additional states

​​Contact improvements

​​Service rewrite

​​Endpoints

​​Views/pages

​​GIS extension & additional boundaries