Re: [boost] clang-win, again

Subject: Re: [boost] clang-win, again
From: degski (degski_at_[hidden])
Date: 2018-10-26 14:28:14

On Fri, 26 Oct 2018 at 15:44, Edward Diener via Boost <boost_at_[hidden]>
wrote:

> Why can't you get it in your head that when you use clang targeting vc++
> you have clang as the compiler using the vc++ headers and libraries
> creating a Windows executable, and when you use clang targeting
> mingw-64/gcc it uses the mingw-64/gcc headers and libraries and also
> produces a Windows executable.

 I have no problem having that in my head. I'm only interested in one of
the two, targeting VC++. I'm also convinced that targeting MinGW, involves
compiling Clang/LLVM yourself, against those MinGW/GCC headers. That
compiler should then be used only with MinGW.

If both are not testing clang on Windows
> I do not know what is.
>

I'm not interested in testing Clang in conjunction with MinGW, mostly
because MinGW is not a stable target and by definition lags.

If I, and obviously the OP of this thread, could get clang-cl using
> clang-win to work reliably I would suggest using that.

To do that you should use/do what you suggest IMO, go the linux route. With
this choice you'll be fully relying on whatever the MinGW devs do [or don't
do] and just hope that that will work. Your own experience seems to
demonstrate that this is the wrong path to follow. Going forward, MinGW
will become, IMHO, a thing of the past.

But it is not my experience that the latter works very well. If that setup
> works for you, that is fine.
>

No, I'm not trying to do what you are trying to do. I don't want to be
dealing with MinGW!!! And trying to do that fails. More so nowadays, CMake
plays an ever more important role and makes building stuff that used to be
Linux/GNU only, often trivial. Most of the reasons to use MinGW [in many
cases] have gone away [also due to and helped by better compliance of VC].

What I take away from this is that at least it has more or less become
clear what you are discussing. So in the future we should actively try to
separate these different issues/configs. Making this clear on the cfe-dev
list [when you address an issue] might also help you, although from what
I'm reading supporting what you are doing/trying to do is back-burner stuff.

The question is, what is clang-win.jam supposed to support, in your terms,
"targeting gcc" or "targeting msvc"?

degski
-- 
*“If something cannot go on forever, it will stop" - Herbert Stein*