Hi Science team, It's needless to introduce the importance of BLAS/LAPACK to you. Let me directly put forward the question:
Are we satisfied with the performance of default BLAS/LAPACK provider? (which is exactly netlib, without optimization) Should we change the system default? I can think of two kinds of solutions, as listed below (type 1) let's bump the global default -------------------------------------- https://salsa.debian.org/science-team/lapack/-/blob/master/debian/shlibs.local Here is the dependency template used by dpkg-shlibdeps when generating the .deb files. If we change the template, all the reverse dependency will have a updated list of providers. Currently the template is: "libblas3 | libblas.so.3" An updated template may looks like this: "libopenblas0 | libblis3 | <...> | libblas3 | libblas.so.3" (type2) leave the flexibility to individual maintainers ------------------------------------------------------- In this way I can implment a lintian check: When a blas/lapack reverse dependency does not Depends, nor Recommends/Suggests some faster implementations, the lintian check produces a info saying the details about performance. And the package maintainers decide what to use. e.g. Package: foobar Depends: libopenblas0 | libblas3 | libblas.so.3 Recommends: libmkl-rt [amd64] ----- Which way do you prefer? --- Ack: GSoC2020
