Page MenuHomePhabricator

[Services] Update commtest protobuf location
ClosedPublic

Authored by jon on Sep 9 2022, 12:11 PM.
Tags
None
Referenced Files
F3387434: D5102.id16560.diff
Fri, Nov 29, 9:29 AM
F3387380: D5102.id16595.diff
Fri, Nov 29, 9:10 AM
Unknown Object (File)
Sat, Nov 16, 9:01 AM
Unknown Object (File)
Sat, Nov 16, 3:17 AM
Unknown Object (File)
Sun, Nov 10, 4:28 PM
Unknown Object (File)
Sat, Nov 9, 6:11 PM
Unknown Object (File)
Sat, Nov 9, 3:18 PM
Unknown Object (File)
Sat, Nov 9, 7:22 AM

Details

Summary

Follow-up to D5046. These paths weren't referenced in any CI gate,
so they were overlooked.

Depends on D5101

Test Plan
nix develop  # optional, if cargo + yarn is already installed
cd services/commtest
cargo clean && cargo build

Diff Detail

Repository
rCOMM Comm
Lint
No Lint Coverage
Unit
No Test Coverage

Event Timeline

These paths weren't referenced in any CI gate, so they were overlooked.

In the future we might want to block new builds/tests until there's a diff up that includes them to CI?

Was able to patch in diff and go through Test Plan successfully!

Accepting to unblock, but please address feedback about comment before landing.

services/commtest/build.rs
18

I think it'd be helpful to explain the "why" here instead of the "what"

This revision is now accepted and ready to land.Sep 12 2022, 8:47 AM

Accepting to unblock, but please address feedback about comment before landing.

This comment was not addressed. @jon, it is critical that you pay attention to comments like this... otherwise, your reviewers simply cannot trust you to handle an accept-with-comments correctly.

In the future we might want to block new builds/tests until there's a diff up that includes them to CI?

As with all things, I think that depends. But I think it's a big "process smell" that we don't have CI set up correctly for a lot of workflows (e.g. unit tests). CI/CD is more than just having passing gates, it's about managing changes and correctness of code; you really need the CI process in place before you start ramming through a bunch of code for it to be effective.

I think it'd be helpful to explain the "why" here instead of the "what"

Agreed, looks like I forgot to follow up on this and then landed it after looking through the rest of the stack.

This comment was not addressed. @jon, it is critical that you pay attention to comments like this... otherwise, your reviewers simply cannot trust you to handle an accept-with-comments correctly.

yea that was my bad. In this case I read this revision first, and then read the others to see if there was any other comments, and forgot to address this.

However, I'm also human, and the diff work queue is there to help me from myself. The revision really should have been blocking until I addressed the comments.

I would really like to get away from "accepted, but not really". It's super error prone, especially with larger stacks. It's also annoying to have to triple check every revision in stack make sure "green is green"

jon added inline comments.
services/commtest/build.rs
18

Passed over this, addressed in https://phab.comm.dev/D5154