Subtyping bug: equivalent types behave differently when matching with diagonal variables

[abolotina]

In Julia 1.2.0,

julia> (Tuple{Q,Int} where Q<:Int) <: Tuple{T,T} where T
false

julia> (Tuple{Int,Int}) <: Tuple{T,T} where T
true

I think the first check should return true because the types on the left-hand side are equivalent:

julia> (Tuple{Q,Int} where Q<:Int) <: Tuple{Int,Int}
true

julia> Tuple{Int,Int} <: Tuple{Q,Int} where Q<:Int
true

Version info:

Julia Version 1.2.0
Commit c6da87ff4b (2019-08-20 00:03 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
  CPU: Intel(R) Core(TM) i3-7100U CPU @ 2.40GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-6.0.1 (ORCJIT, skylake)

[Moelf]

any update on this? the pattern seems quite common

[adienes]

+1
once encountered when toying with where to learn how subtyping works because I'm too stubborn to RTFM, and this confused the crap out of me before I knew was bug