Mention autoconf and location of repository

diff --git a/INSTALL b/INSTALL
index 9c690ff..1999a7a 100644 (file)
--- a/INSTALL
+++ b/INSTALL
@@ -1,8 +1,38 @@
-============
+Getting libaec
+==============
+
+The source code of libaec is hosted at DKRZ GitLab.
+
+## Source code and binary releases
+
+The latest releases of libaec can be downloaded at the following
+location:
+
+  https://gitlab.dkrz.de/k202009/libaec/tags
+
+## Developer snapshot
+
+  git clone https://gitlab.dkrz.de/k202009/libaec.git
+
+
 Installation
 ============
 
-The installation procedure uses CMake, as follows:
+## Installation from source code release with configure
+
+The most common installation procedure on Unix-like systems looks as
+follows:
+
+Unpack the tar archive and change into the unpacked directory.
+
+  mkdir build
+  cd build
+  ../configure
+  make check install
+
+## Installation from source code release with CMake
+
+As an alternative, you can use CMake to install libaec.
 
 Unpack the tar archive and change into the unpacked directory.
 
@@ -22,7 +52,27 @@ or by setting the options manually, e.g.
 
 in order to set the install prefix to ~/local
 
-=======================
+CMake can also generate project files for Microsoft Visual Studio when
+used in Windows.
+
+## Installation from cloned repository
+
+The configure script is not included in the repository. You can
+generate it with autotools:
+
+  cd libaec
+  autoreconv -iv
+  mkdir build
+  cd build
+  ../configure
+
+Also not included are CCSDS sample data which are needed for
+testing. They have to be downloaded prior to running 'make check':
+
+  make update-sampledata
+  make check install
+
+
 Intel compiler settings
 =======================
 
@@ -30,7 +80,7 @@ The Intel compiler can improve performance by vectorizing certain
 parts of the code on x86 architectures. Assuming your CPU supports
 AVX2, the following options will increase encoding speed.
 
-  ./configure CC=icc
+  ../configure CC=icc
   make CFLAGS="-O3 -xCORE-AVX2" bench
 
 On a 3.4 GHz E3-1240 v3 we see more than 400 MiB/s for encoding