Closed gnail closed 5 years ago
I wonder if we should name it microsoft_gsl
as there is another https://en.wikipedia.org/wiki/GNU_Scientific_Library. @koponomarenko another interesting example triggering policy inconsistency. The problem is, there are clear cases when the project is well known and unambiguous, sometimes the opposite (like json), but we will see something in the middle pretty often as well, e.g. this one.
A rule of thumb used in the beginning was to name things after their pkg-config names. I would imagine gsl
to be taken by GNU Scientific Library so we should probably reserve that for it.
For reference there's also https://github.com/martinmoene/gsl-lite which provides another implementation so microsoft_gsl
does make sense. I can see there could be disagreement on whether it should be ms-gsl
, msft-gsl
, gsl-ms
, gsl-cpp
etc. but I guess that's more about naming policy in general which is outside the scope of this issue.
Okay, I am not sure if it matters, but what I found in our wrapbase is that name-project
is used for ambigous names (e.g. https://github.com/mesonbuild/hinnant-date) or https://github.com/mesonbuild/nlohmann_json which is the full name itself. So I created https://github.com/mesonbuild/microsoft-gsl.
upstream url: https://github.com/Microsoft/GSL/releases version: 2.0.0