Kristoffer Carlsson
Kristoffer Carlsson
According to @vtjnash this is kind of known in that when compiling, we will not block the other threads from trying to run code, which means they might run a...
I can't repro this on 1.12 or nightly. Could you check again if things are better now.
My guess is that this does not have to be discussed much more.
The reason for this is: ``` julia> @code_lowered foo((1,2), 3) CodeInfo( 1 ─ %1 = Base.indexed_iterate(@_2, 1) 1 1 ─ %1 = Base.indexed_iterate(@_2, 1) 2 │ a = Core.getfield(%1, 1)...
This is the stepping that is done https://github.com/JuliaDebug/Debugger.jl/blob/6a6beab2f559a5a72d3284c1c6482e0620026db8/src/Debugger.jl#L93-L94 and this case is not handled. I'll move this issue to JuliaInterpreter since these things are implemented there.
The current code calls the `VersionSpec` constructor with the string: https://github.com/JuliaLang/Pkg.jl/blob/f4b66f7da6ff40cce9781c548324bdd3680d7cb4/src/Versions.jl#L232 One difference between that and the "semver" version is for example with the current behavior, `version=1.2` gives you a...
> One thing I'm slightly concerned about here is the approach taken to making sure that the executables are on the users's PATH on Linux/BSD systems. With regards to XDG...
> nice work! I'm wondering how apps are shared across Julia versions? e.g. are they isolated by Julia versions like how the global environment are setup? As it is right...
> why the support for creating symlinks in the user's local bin-dir on Linux has been removed (requiring $PATH shenanigans) though I tried to scale off as much as possible...
> I do see making modification of $PATH a last-resort as rather important, given all the potential complications (incidentally there's a conversation that's just gone on in #hpc on Slack...