# 2021-10-11 - Sports Hands
###### tags: `work` `comcast` `farside`

## Background

* [x] what products did Legacy serve
* [x] what products was Farside intended to serve
* [x] what products does Farside serve

## Consumers

### Distribution

| | current | in development | on the horizon |
|:------------------------------ |:-------:|:--------------:|:--------------:|
| DVR Auto-Extend | **X** | | |
| ISEP | | **X** | |
| Merlin (`entityDS`) | **X** | | |
| Merlin (`linearDS`) | **X** | | |
| Peacock Sports Exp | | **X** | **X** |
| Sidekick | **X** | | **X** |
| Sky Mobile* | | | **X** |
| Sky Q | **X** | **X** | **X** |
| Sky Sports Exp | **X** | **X** | **X** |
| Sky Sports Hub Max | **X** | **X** | **X** |
| Sports App (~temporary) | **X** | | |
| Sports Voice ([sunsetting][1]) | **X** | | |
| Watch With ("Sport Radar") | | | **X** |

> (*) actual funcationality TBD

### Data Use Cases

| | Static | Semi-Live | Live | | Pull | Push |
|:-------------------- |:------:|:---------:|:-----:| --- |:-----:|:-----:|
| DVR Auto-Extend | | **X** | | | | **X** |
| ISEP | | **X** | | | **X** | **X** |
| Merlin (`entityDS`) | **X** | | | | | **X** |
| Merlin (`linearDS`) | **X** | **X** | | | | **X** |
| Peacock Sports Exp | | | **X** | | | **X** |
| Sidekick | **X** | **X** | | | **X** | |
| Sky Mobile* | | **X** | **X** | | **X** | **X** |
| Sky Q | **X** | **X** | | | **X** | |
| Sky Sports Exp | | | **X** | | | **X** |
| Sky Sports Hub Max | | | **X** | | | **X** |
| Sports App | **X** | **X** | | | **X** | **X** |
| Sports Voice | **X** | **X** | | | **X** | |
| Watch With | | | **X** | | | **X** |

> **Static** (pull) - schedules, rosters, pre-game stats, post-game stats
> **Semi-Live** (pull, push) - scores, statuses
> **Live** (pull, push) - live game stats, plays, lineups
> (*) actual funcationality TBD

## Projects & Technology

* [ ] is our team expertise being properly leveraged
* [ ] maintenance load
* [ ] onboarding load
* [ ] scalability

### Neo4j
* [ ] why did we choose this?
* [ ] is it delivering on our expectations?
* [ ] drawbacks?
* [ ] accumulated expertise

### GraphQL
* [ ] why did we choose this?
* [ ] is it delivering on our expectations?
* [ ] drawbacks?
* [ ] accumulated expertise

### Room for Improvement (aka _Matt's Complaints_)

#### `farside-api`
* [ ] caching? hardly :shrug: - re: team expertise
* [ ] resolvers...
* [ ] why don't we use them?
* [ ] where can we use them?
* [ ] `listingsConnection` and `programsConnection`...
* [ ] why do we own these?
* [ ] why do we source these from our db

#### `graph-parser` / `graph-types-cypher`
* [ ] incoming data is built and validated at push time, which makes sense; this data can be unreliabile from push to push
* [ ] outgoing cypher statements are built and validated at push time, which doesn't make sense; they don't change from push to push

#### `clara`
* [ ] potentially too many inputs - keeping `Program`'s and `Listing`'s up to date at the hourly (for events X hours out) or daily (for events Y hours out) scale is possibly sufficient and might end up being more easily triaged
* [ ] who owns `Program`'s and `Listing`'s? why do we write them to our database?

[1]: https://etwiki.sys.comcast.net/pages/viewpage.action?spaceKey=VI&title=*WIP*+Insights+Sunset+Plan+v1