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
typeFoowithmemberx.Bar()=()typeFoo2=memberx.Bar2()=()typeFoo3=abstract Bar3: unit -> unit
leads to
typeFoowithmemberx.Bar()=()typeFoo2=memberx.Bar2()=()typeFoo3=abstract Bar3: unit -> unit
Does the new line in type Foo with make sense? Is it desired by the teams?
My gut feeling is that we never discussed this case in great detail.
And the result we have today just happens to be how the implementation works.
Eugene advocated for having it only when there's a representation or non-member bindings/fields in the default code style.
typeA=leta=1memberthis.P=1typeA=valfield:int // not sure whether it's correct syntaxmemberthis.P=1typeA=do()memberthis.P=1typeA=inherit T()memberthis.P=1typeA=| Case
memberthis.P=1typeB=memberthis.P=1typeB()=memberthis.P=1typeBwithmemberthis.P=1
I think this sounds reasonable, would be interested to know what the GR folks think about it.
The text was updated successfully, but these errors were encountered:
(This is @Smaug123 commenting from my GR-internal GitHub account.)
I think the final snippet looks reasonable; the type Foo with\n\nmember x.Bar() = () does look a bit ugly. Omitting the first newline looks right to me. I'll just send this round in case anyone disagrees.
Hi @Smaug123,
@auduchinok and I were talking about fsharp_newline_between_type_definition_and_members yesterday and we may have found a questionable case.
The setting is on by default in Fantomas:
leads to
Does the new line in
type Foo with
make sense? Is it desired by the teams?My gut feeling is that we never discussed this case in great detail.
And the result we have today just happens to be how the implementation works.
Eugene advocated for having it only when there's a representation or non-member bindings/fields in the default code style.
I think this sounds reasonable, would be interested to know what the GR folks think about it.
The text was updated successfully, but these errors were encountered: