Conversation
These new names are pithier and match up with the rest of our terminology: - `--emit=static-files` matches the default name of the directory that it actually emits, which is `static.files` (the hyphen is used for emit because every other emit option uses hyphens, but the directory uses a dot because we don't want its name to conflict with a crate). - `--emit=data-files` matches the convention that emit is a noun, not an adjective. This commit changes the docs, but leaves in support for the old names, to break the cycle with cargo and docs.rs. This commit needs merged, then cargo and docs.rs will be updated to use the new names, then, finally, the old names will be removed.
|
r? @fmease rustbot has assigned @fmease. Use Why was this reviewer chosen?The reviewer was selected based on:
|
|
Should we poll these names, discuss these names with the other team members on Zulip or something like that? |
|
@rfcbot ask rustdoc Do you like @rfcbot ask rustdoc Do you like This is for resolving concern naming raised in pFCP #146220 (comment). See also #146220 (comment). At the latest, I'm going to close both polls in one week from now (2026-03-12, 12pm UTC). I might merge the PR before that. This is just a vibe check & non-binding. |
|
Team member @fmease has asked teams: T-rustdoc, for consensus on:
|
|
Team member @fmease has asked teams: T-rustdoc, for consensus on:
|
|
I'm fine with both. |
|
Added to meeting notes for 2026-03-09 |
|
the new names also seem a lot more convenient to type out that the old ones, which i like. |
|
I'm fine with What about re-using Rust terminology with |
|
Re. |
| "toolchain-shared-resources" => Ok(Self::Toolchain), | ||
| "invocation-specific" => Ok(Self::InvocationSpecific), | ||
| // old nightly-only choices that are going away soon | ||
| "toolchain-shared-resources" => Ok(Self::StaticFiles), |
There was a problem hiding this comment.
Is it worth emitting a warning here for now?
There was a problem hiding this comment.
This won't work IINM since Cargo wouldn't have a way to suppress it so its tests would fail. Moreover, we can't fix Cargo from within this repository since it's a git submodule.
See also: #139869 (comment), #139869 (comment).
That's why we have to temporarily keep the old names in the first place.
|
According to #t-rustdoc/meetings > 2026-03-09 @ 💬 it looks like we've settled on:
Since this is non-binding, I'd say let's go for it (ultimately, only the pFCP #146220 (comment) is binding ofc). @rustbot author |
|
Reminder, once the PR becomes ready for a review, use |
These new names are pithier and match up with the rest of our terminology:
--emit=static-filesmatches the default name of the directory that it actually emits, which isstatic.files(the hyphen is used for emit because every other emit option uses hyphens, but the directory uses a dot because we don't want its name to conflict with a crate).--emit=data-filesmatches the convention that emit is a noun, not an adjective.This commit changes the docs, but leaves in support for the old names, to break the cycle with cargo and docs.rs. This commit needs merged, then cargo and docs.rs will be updated to use the new names, then, finally, the old names will be removed.
CC #146220 (comment)