On 05.07.2017 19:18, Peter Dimov via Boost wrote:
Stefan Seefeld wrote:
Fair enough. But it's a change, and as such, it comes at a price. (And we haven't even talked about Andrey's comment:
Does it only include the address model without the architecture? If yes, > it doesn't really solve the problem since you'll still have the same > issue when you compile 32-bit x86 and ARM binaries, for example.
In Boost.Build, the architecture (arm mips1 power sparc x86 combined) is a separate property from the address model (32 64).
I know.
In principle, we could also encode the architecture.
My impression though, based on observing earlier discussions, is that tying the address model issue to the architecture issue is a very effective way to stall progress and ultimately do nothing.
How so ? Do you expect to get more opposition if you propose to encode both ? I thought the main argument against the change was the fact that it's an intrusive change. Once we agree to a change, the change itself isn't quite as important. In contrast, the fact that this solves an even larger issue would make the patch more acceptable. (Without wanting to open the door to discussing the separation of "architecture" and "address model" in Boost.Build, I do believe that the two should be coupled, at least in the file naming context: In the real world (i.e., outside Boost), these two are always combined into distinct and unambiguous names.) Stefan -- ...ich hab' noch einen Koffer in Berlin...