Stream: git-wasmtime

Topic: wasmtime / PR #4479 Add definitions of tiers-of-support f...


view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 16:40):

alexcrichton opened PR #4479 from tiers to main:

This commit adds documentation of a Tiers-based system for classifying
how supported a component is within Wasmtime. This was somewhat
pioneered in the [Wasmtime 1.0 RFC][rfc] but the documentation here is
expanded to include more than just API stability but additionally other
components. Inspiration for this is drawn from Rust's definition of
[support tiers][rust] as well.

The motivation for this is to help clarify what exactly it means to live
at each tier and what is expected. For example one thing this document
clarifies is the requirements necessary for landing new major changes in
Wasmtime at all. Additionally this helps clarify what it means to have
the highest level of support vs "otherwise well supported".

[rfc]: https://github.com/bytecodealliance/rfcs/blob/main/accepted/wasmtime-one-dot-oh.md#tier-1---api-stable-production-quality
[rust]: https://doc.rust-lang.org/rustc/target-tier-policy.html

<!--

Please ensure that the following steps are all taken care of before submitting
the PR.

Please ensure all communication adheres to the code of conduct.
-->

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 17:05):

abrown submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 17:05):

abrown created PR review comment:

I think wasi-nn could arguably be in tier 2.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 19:22):

jameysharp created PR review comment:

The distinction between "turned off" and "removed" isn't clear in this paragraph. I think you mean that, first, failing CI/fuzz tests will be disabled, and then if the issue still isn't addressed we'll merge a commit removing the implementation entirely. (Although of course that can be reverted later if somebody commits to maintaining it again.) Do I have that right?

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 19:22):

jameysharp submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 19:22):

jameysharp created PR review comment:

For a little more clarity, is this edit accurate?

* Complete implementations for anything that's part of Tier 1. For example
  all Tier 2 targets must implement all of the Tier 1 WebAssembly proposals,
  and all Tier 2 features must be implemented on all Tier 1 targets.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 19:22):

jameysharp created PR review comment:

I think this rewrite means the same thing but it confuses me less:

* Major changes affecting this component may require help from maintainers with
  specialized expertise, but otherwise it should be reasonable to expect most
  Wasmtime developers to be able to maintain Tier 1 features.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 19:22):

jameysharp created PR review comment:

I assume security is part of the requirement for tier 1 targets, not just features?

  in features and targets.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 20 2022 at 19:22):

jameysharp submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 22 2022 at 15:05):

alexcrichton submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 22 2022 at 15:05):

alexcrichton created PR review comment:

Indeed! I'll try to clarify here.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 22 2022 at 15:06):

alexcrichton updated PR #4479 from tiers to main.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 22 2022 at 15:10):

alexcrichton submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 22 2022 at 15:10):

alexcrichton created PR review comment:

Given the current list of requirements sketched out I believe that wasi-nn is still largely missing tests. There's one example that is compiled and run but I think that we'd want more of a unit-test-like-suite otherwise.

At least personally this sort of tracks with my gut feeling about wasi-nn and wasi-crypto where they're present but don't feel all-that-well integrated/first-class. WASI itself isn't really all that first-class by requiring an extra wasmtime-wasi crate which hasn't had much time put into its API design for stability too, though. I don't know how best to think about WASI proposals myself.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 22 2022 at 18:06):

jameysharp submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2022 at 15:38):

alexcrichton updated PR #4479 from tiers to main.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 25 2022 at 17:46):

alexcrichton has marked PR #4479 as ready for review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 29 2022 at 14:17):

alexcrichton updated PR #4479 from tiers to main.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 29 2022 at 14:17):

alexcrichton has enabled auto merge for PR #4479.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 29 2022 at 15:11):

alexcrichton merged PR #4479.


Last updated: Dec 23 2024 at 12:05 UTC