Skip to content

Introduce tiers.md #242

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Mar 1, 2023
Merged

Introduce tiers.md #242

Changes from 2 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
f3d76b3
Introduce tiers.md
brunoocasali Feb 14, 2023
f4326ab
Introduce changes from review made by @bidoubiwa
brunoocasali Feb 14, 2023
797572e
Rewrite the guide to remove "we"
brunoocasali Feb 27, 2023
d88a289
Update resources/tiers.md
brunoocasali Feb 28, 2023
f85b912
Update resources/tiers.md
brunoocasali Feb 28, 2023
0173b87
Remove badge & identification section
brunoocasali Mar 1, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
84 changes: 84 additions & 0 deletions resources/tiers.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,84 @@
# SDK tiers

The objective of Meilisearch's Integrations Team is to **ease Meilisearch usage to as many developers as possible**.

Given this statement, **tiers** are nothing more than a __group of SDKs/integrations__ where the Integrations Team can fairly divide their attention. Unfortunately is not possible to keep all the SDKs 100% updated against the new Meilisearch releases all the time. The Integrations Team is a polyglot small team that maintains 30+ different integrations and tools, so it is humanly impossible to consistently meet the Meilisearch quality standards.

The ultimate goal of introducing tiers is to allow the team to have more time to invest in new projects/integrations to impact more users.

To measure the impact of an SDK, the data came from different places like [telemetry](https://docs.meilisearch.com/learn/what_is_meilisearch/telemetry.html) and GitHub's stars, forks, and watchers.

It is worth remembering that this priority is just a concept introduced by the team that aims to give a decision weight before investing time in some issues of the SDKs. So it does not mean we will never work on feature requests from the bottom tiers anymore, but it says we will prioritize the top tiers most of the time transparently.

:warning: The team will still keeping all officially maintained SDKs **working**. :warning:

## Tier #1

The SDKs in group #1 have total priority over the team's time. For example, issues are visited before issues are present in tier #2.
It is worth remembering that this does not mean open issues are addressed and fixed promptly. Every issue needs time to be understood/confirmed/worked on, and consequently closed.
All new Meilisearch features that make sense for those SDKs will be available on the release day.

## Tier #2

Tier #2 SDKs are the emergent ones, we strongly believe they can move further, and we are betting on them.
In any case, since they bring less value compared to tier #1 we only spend the amount of time required to keep it working and just do essential maintenance code and code cleanup.

Only in urgent cases where, for example, the SDK is no longer working, issues from the SDKs in group #2 are prioritized over tier #1.


## Tier #3

The SDKs in group #3 have relatively low user adoption compared to groups #2 and #1. They are probably new or have a small community behind the tech stack in the first place.
As with tier #2, SDKs in this group are prioritized only in critical cases, and they are not going to receive active improvements or maintance from the integrations team.

This is a very special group because we really need help from the community here.


## Tiers list

Integration | Tier |
-------------|------|
Javascript | #1 |
PHP | #1 |
Instant Meilisearch | #1 |
Python | #1 |
Go | #1 |
Strapi | #2 |
Ruby | #2 |
Rails | #2 |
.NET | #2 |
Rust | #2 |
Symfony | #3 |
docs-searchbar.js | #3 |
Copy link
Member

@gmourier gmourier Feb 28, 2023 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is there a reason not to have doc-scrapper in the list?

That's potentially something we will want to maintain / replace actively at some point because It's an excellent opener to get many projects moving from Algolia (but not only) to Meilisearch for their documentation.

Copy link
Member Author

@brunoocasali brunoocasali Feb 28, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not really. It is just because we don't have the User Agent there!

Firebase | #3 |
Dart | #3 |
Java | #3 |
Copy link
Contributor

@alallema alallema Feb 14, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why Java is only on tier #3?

Copy link
Member

@curquiza curquiza Feb 14, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think you indeed should keep an eye on Java usage, because the community is huge and already came to a lot of time (they are currently a little bit frustrated because our SDK is not up to date). Can quickly become tier #2 or (even tier #1 😱)

Copy link
Member Author

@brunoocasali brunoocasali Feb 14, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I hope they will! But for now, since we don't have data coming from the Java users, Java is #3. I'll share what is my rationale behind the list and the order after merging this first version :)

Swift | #3 |
Vuepress | #3 |
Gatsby | #3 |

## FAQ
### How are the tiers organized?

The method used is mostly manual and subjective, where we use the data previously collected from the [telemetry](https://docs.meilisearch.com/learn/what_is_meilisearch/telemetry.html) and Github to generate groups.

To begin we only have 3 tiers, in the future, this number may change.

### Is it possible for an SDK to move up a level?

Yes, fully possible! Especially if the team detects that more users are interested in and using it.
Whenever a anomaly is detected, a change can be submitted to the `integration-guides` and the groups will be modified to better manage the team's time. The reassessment of the tiers will be carried out every release cycle (8 weeks).

The other way around is also valid. A SDK can move to less important tiers. :shipit:

### How to identify which tier an SDK belongs to?

It is possible by accessing the README.md of each repository. A tier badge is present below the Meilisearch logo. The badge can have the following values: "tiers #1", "tiers #2" and "tiers #3". For example:
Copy link
Contributor

@bidoubiwa bidoubiwa Feb 28, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not convinced that adding the badges is a good idea. SDK's will still work with the latest versions of Meilisearch and most users will be satisfied with its current state.

By warning them that this is a low priority SDK it might result in:

  • User fearing deprecation
  • User complaining because they don't want their SDK in that tier list even if they would never user the missing features
  • Give the impression that we are going backward in activity like a "cut in budget"

If something is missing, an issue exist in any case with the missing feature which reassure the user that we did not forget about the SDK, provide them a place to tell us they need it, and help them implement it if they want it sooner than later.

Copy link
Member Author

@brunoocasali brunoocasali Feb 28, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't want to give up of the badges 😢. But I have to agree with you on that.


[![tiers - #1](https://img.shields.io/static/v1?label=tiers&message=%231&color=blue)](https://github.com/meilisearch/integration-guides/blob/main/resources/tiers.md)
[![tiers - #2](https://img.shields.io/static/v1?label=tiers&message=%231&color=2ea44f)](https://github.com/meilisearch/integration-guides/blob/main/resources/tiers.md)
[![tiers - #3](https://img.shields.io/static/v1?label=tiers&message=%231&color=critical)](https://github.com/meilisearch/integration-guides/blob/main/resources/tiers.md)

#### Why SDKs are present in a particular tier?

Because those SDKs have the most active users. Not only from users using the Meilisearch Cloud. They have a larger community in general. Other principles could be used to move SDKs around if we think it is relevant.