For errors, gently suggest rustup check #784
Conversation
Fixes fermyon#783 Signed-off-by: Neal McBurnett <nealmcb@gmail.com>
There was a problem hiding this comment.
Great catch - thanks! I reckon we should also mention this on the Rust page: https://github.com/fermyon/spin/blob/main/docs/content/rust-components.md - otherwise all grand.
We should have a think about how to ensure the versioning information remains correct. Maybe pin CI to a specific Rust version and add a check that the docs match CI. (That's definitely out of scope for this PR! I'm just thinking out loud.)
question: is version 1.56 still recent enough? Is there a reason for picking that version? Signed-off-by: Neal McBurnett <nealmcb@gmail.com>
|
Thanks, @itowlson. I clarified the rust-components.md page a bit. I note they already mentioned Rust versioning, and they mention 1.56+ which is actually older than what i was running. Given that this project is moving quickly, we could just remove the specific version number recommendations and suggest getting a "recent" version. |
|
We could, though I quite like the specific minimum, as it avoids ambiguity. As the one who got bitten by the issue, do you think "check for a recent version" would have helped enough, or would you have preferred "check for 1.56+"? |
|
I think just mentioning that rust versions are something to check is fine. I doubt that maintainers of the code base will enjoy trying to keep recommendations for specific person numbering up to date, until there's a 1.0 release. |
can confirm grin So on that basis, would you like us to merge this PR as it stands, or to reword, or to close this one and have us pick it up? |
Until 1.0? Signed-off-by: Neal McBurnett <nealmcb@gmail.com>
Signed-off-by: Neal McBurnett <nealmcb@gmail.com>
|
I dare say that this, without the version numbers, is best for now. |
|
Go for it. Thanks! |
Fixes #783
Signed-off-by: Neal McBurnett nealmcb@gmail.com