Update README + BUILDING

.circleci/config.yml

@@ -13,7 +13,7 @@ jobs:
       - image: reiniscirpons/hpcombi-env-arm64v8:v1
     resource_class: arm.medium
     steps:
-      - run: 
+      - run:
           name: "Set up compiler"
           environment:
             COMPILER_NAME: << parameters.compiler >>

.github/workflows/runtests.yml

@@ -49,7 +49,6 @@ jobs:
          - name: "Install dependencies . . ."
            run: |
              sudo apt-get install --yes ccache
-             sudo apt-get install --yes libbenchmark-dev
          - name: "Configure the HPCombi build . . ."
            env:
              CC: ${{ env.CC }}

BUILDING.md

@@ -1,23 +1,24 @@
 # Building HPCombi
 
-## Build Prerequisites:
+Note that HPCombi is a C++17 header-only library, and as such does not need to
+be built. The instructions below are only for those who wish to either run the
+tests, experiments, examples, or benchmarks.
 
-- CMake 2.8 or later
+## Build prerequisites:
 
-- A recent c++ compiler. I have tested the code on
-  * g++ 5.3.1, 6.2.1 and 7.1.1.
-  * clang 5.0.0
-  * g++ 4.8 and 4.9 are known to be broken (I can fix it if needed at the price
-  of some uglification of the code).
+- CMake 3.8 or later
 
-- [optional] : Google sparsehash/dense_hash_map, sparsehash/dense_hash_set.
-  if not the less efficient standard containers will be used.
+- A recent C++ compiler implementing the C++17 standard. We routinely test
+  HPCombi using:
+  * gcc 9 to 12; and
+  * clang 11 to 15
+  on both x86 and arm processors.
 
-- BOOST.test (shared library version) : needed for testing.
+- Your machine must support a small number of builtin functions such as `__builtin_popcnt`.
 
-- Your machine must support AVX instructions.
+- [optional] : Google `sparsehash/dense_hash_map` and/or `sparsehash/dense_hash_set`.
 
-- Doxygen for generating the API documentation (in progress).
+- [optional] Doxygen for generating the API documentation (in progress).
 
 ## Building
 

README.md

@@ -1,10 +1,20 @@
 # HPCombi
 High Performance Combinatorics in C++ using vector instructions v0.0.8
 
-SSE and AVX instruction sets allows for very fast manipulation of
-combinatorial objects such as transformations, permutations, boolean matrices
-of small size. The goal of this project is to implement various new algorithms
-and benchmark them on various compiler and architecture.
+HPCombi is a C++17 header-only library using the SSE and AVX instruction sets,
+and some equivalents, for very fast manipulation of combinatorial
+objects such as transformations, permutations, and boolean matrices of small
+size. The goal of this project is to implement various new algorithms and
+benchmark them on various compiler and architectures.
+
+HPCombi was initially designed using the SSE and AVX instruction sets, and did
+not work on machines without these instructions (such as ARM). From v1.0.0
+HPCombi supports processors with other instruction sets also, via
+[simd-everywhere](https://github.com/simd-everywhere/simde). It might be the
+case that the greatest performance gains are achieved on processors supporting
+the SSE and AVX instruction sets, but the HPCombi benchmarks indicate that
+there are also still signficant gains on other processors too.
+<!-- TODO add link to HPCombi wiki with benchmark graphs -->
 
 ## Authors
 
@@ -13,12 +23,16 @@ and benchmark them on various compiler and architecture.
 
 ## Contributors
 
-- Finn Smith : discussions + BMat8 reference code
-- Viviane Pons : algorithms discussions
-- Reinis Cirpons : CI
+- Reinis Cirpons : CI + benchmark graphs
+- Viviane Pons : discussions about algorithms
+- Finn Smith : discussions + `BMat8` reference code
 
+## Thanks
 
-## Acknowledgments
-
-- This development is funded by the [OpenDreamKit](http://opendreamkit.org/)
-  Horizon 2020 European Research Infrastructure project (#676541).
+- The development of HPCombi was partly funded by the
+  [OpenDreamKit](http://opendreamkit.org/) Horizon 2020 European Research
+  Infrastructure project (#676541), which the authors acknowledge with thanks.
+- Thanks also to the
+  [simd-everywhere](https://github.com/simd-everywhere/simde) and
+  [catch2](https://github.com/catchorg/Catch2) authors and contributors for
+  their excellent libraries!