Description
Hey there GraalVM team,
To preface: I was able to build a statically compiled native image some months ago, using glibc already, so all that works fine.
At the document here:
https://www.graalvm.org/latest/reference-manual/native-image/guides/build-static-executables/
The steps are explained.
I just re-read it. But after reading it, I wonder about the use cases for a mostly statically compiled native image.
"With GraalVM Native Image you can build a mostly-static native executable that statically links everything except libc. Statically linking all your libraries except libc ensures your application has all the libraries it needs to run on any Linux libc-based distribution."
However had, I am still unsure what the main differences / benefits are. So in my simple brain, I have it now associated that I want a fully statically compiled application, so I am never dependent on libc. I suppose this is not entirely correct or complete.
Could someone perhaps consider adding a new FAQ entry there, e. g. use cases for either these two variants? Or in other words, why we need a mostly statically compiled native executable, as opposed to a fully statically compiled one? Because the latter one appears to be more useful to me, so maybe I got something wrong.
(PS: Also full musl support would be interesting one day, perhaps the resulting binary would be smaller. But I digress here.)