Peter Dimov

Interesting. I use it unconditionally: https://github.com/boostorg/container_hash/blob/87c9eefe6ea8ed3cc0a7381e2580c1ce562e6acd/.appveyor.yml#L72 and the `msvc-9.0,msvc-10.0` jobs pass: https://ci.appveyor.com/project/pdimov/container-hash/build/job/tbqpop4ap9qveye6

I believe Nikita means https://github.com/bfgroup/b2/blob/0c499449ae4e271a5eb4dcd81daf97422f9bf05e/src/tools/msvc.jam#L602-L617. Not sure what the intended meaning of "saliently" here was. :-) The above does seem to set linker as the default though. Is this a...

Recent fixes to clang-win.jam may have made this unnecessary, but I think embed-manifest-via=linker should still be the default.

--abbreviate-paths should probably split on both '.' and '-' in order to handle this case. (It's common for library-local features to have such composite names separated with dots instead of...

> I can't think of any. Well... https://github.com/bfgroup/b2/issues/325.

> I honestly consider `address-model` a mistake. What do you mean by that? That 32 bit should be a separate `architecture`? How would that solve our problem?

Another option is to special-case `architecture` and `address-model` and instead of adding `address-model-64/architecture-x86/` to the path, just add `x86-64/` to it.

Yet another option is to have a max property set path length threshold in b2.exe and switch to a hash automatically once that is hit. Unfortunately we'll have to set...

We're looking at implementing the `x86_64/` suggestion above on the Boost side, via https://github.com/boostorg/boost/pull/898.

Hiding the paths isn't ideal; it's useful to see what's being built. At minimum we should keep the paths that don't exceed the hash length (32), but the budget can...