You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Do we really need this? I know that occasionally I have thought it would be nice to have, but we have lasted this long without anybody complaining loudly about it, so it doesn't seem vital.
I wonder how it would interact with our various argument modes. Does it effectively turn them all into copies? Is it only compatible with certain modes?
Hm. I assumed that regular modes (non-copy/move) would be able to just create views into the existing value without copying, since such args are immutable locals, but I guess if that value has mutable components, this would mess with the invariant that locals don't get their values changes from under them.
For by-copy and by-move the story gets even more involved. Maybe you're right that we shouldn't bother.
Activity
Aluminium95 commentedon Feb 22, 2012
Would
enumandrecordsupport destructuring in function arguments too ?brson commentedon Feb 23, 2012
Do we really need this? I know that occasionally I have thought it would be nice to have, but we have lasted this long without anybody complaining loudly about it, so it doesn't seem vital.
marijnh commentedon Feb 23, 2012
It'd be consistent. I occasionally find myself wanting to use this -- mostly for tuple arguments.
brson commentedon Feb 23, 2012
I wonder how it would interact with our various argument modes. Does it effectively turn them all into copies? Is it only compatible with certain modes?
marijnh commentedon Feb 23, 2012
Hm. I assumed that regular modes (non-copy/move) would be able to just create views into the existing value without copying, since such args are immutable locals, but I guess if that value has mutable components, this would mess with the invariant that locals don't get their values changes from under them.
For by-copy and by-move the story gets even more involved. Maybe you're right that we shouldn't bother.
marijnh commentedon Feb 28, 2012
I'm closing this. If someone feels strongly about supporting this, discuss.
Remove old action.yml file and update release workflow (rust-lang#1881)
Merge pull request rust-lang#1881 from folkertdev/s390x-project-simd
Merge rust-lang#1881
Correct Windows argument handling
Auto merge of rust-lang#1891 - ChrisDenton:win-args, r=RalfJung