Decision for #175: C++ interoperability goals

proposals/p0175_decision.md

@@ -0,0 +1,47 @@
+# Decision for: C++ interoperability goals
+
+<!--
+Part of the Carbon Language project, under the Apache License v2.0 with LLVM
+Exceptions. See /LICENSE for license information.
+SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+-->
+
+Proposal accepted on 2020-11-24
+
+Affirming:
+
+-   [austern](https://github.com/austern)
+-   [chandlerc](https://github.com/chandlerc)
+-   [geoffromer](https://github.com/geoffromer)
+-   [gribozavr](https://github.com/gribozavr)
+-   [josh11b](https://github.com/josh11b)
+-   [zygoloid](https://github.com/zygoloid)
+
+Abstaining:
+
+-   [noncombatant](https://github.com/noncombatant)
+-   [tituswinters](https://github.com/tituswinters)
+
+## Open questions
+
+All open questions in the document are intended to be answered later after we
+have more experience.
+
+## Rationale
+
+The core team agrees with the mapping of the proposed interoperability goals to
+the overall Carbon goals as rationale.
+
+Embedding of Carbon in C++ code is inspired by experience with CUDA, where users
+provided feedback that CUDA was preferred over alternatives such as OpenCL due
+to the ability to embed it in the same source file as C++. It also matches the
+approach of bridging between C++ and C where bridge code is often embedded as
+`extern "C"` within C++ source files.
+
+The non-goals seem appropriately motivated by the non-interoperability
+priorities, without undermining the effectiveness of the interoperability as a
+whole.
+
+We agree with ensuring some level of C interoperability in order to effectively
+interoperate with both some parts of the C++ ecosystem that rely on the C ABI as
+well as all other languages which target C-based interoperability.