sems-server / sems

Sip Express Media Server
161 stars 92 forks source link

Clarification needed on whether make or CMake is authoritative #110

Open scfc opened 5 years ago

scfc commented 5 years ago

README.md advises to use make && make install, .travis.yml uses CMake. Running CMake changes several Git-tracked configuration files, while on the other hand not allowing "[i]n-source builds" as "CMake would overwrite all makefiles distributed with SEMS".

This is all very confusing to a casual observer, and it would be nice if it could be clarified which build system should be authoritative so that any collisions can be addressed.

juha-h commented 5 years ago

Tim Landscheidt writes:

README.md advises to use make && make install, .travis.yml uses CMake. Running CMake changes several Git-tracked configuration files, while on the other hand not allowing "[i]n-source builds" as "CMake would overwrite all makefiles distributed with SEMS".

This is all very confusing to a casual observer, and it would be nice if it could be clarified which build system should be authoritative so that any collisions can be addressed.

README is very much out of date. I personally have not tried to use CMake to build sems and don't know much about it.

lemenkov commented 5 years ago

README.md advises to use make && make install, .travis.yml uses CMake. Running CMake changes several Git-tracked configuration files, while on the other hand not allowing "[i]n-source builds" as "CMake would overwrite all makefiles distributed with SEMS".

This is all very confusing to a casual observer, and it would be nice if it could be clarified which build system should be authoritative so that any collisions can be addressed.

I personally use CMake and my plan is to promote it as a build tool as soon as we ensure it works on all supported systems (I guess FreeBSD, Linux, and probably Windows).

Still it's in a work-in-progress stage. Let's do it together :)

lemenkov commented 5 years ago

@scfc , btw could you please do a PR for -Wreorder warnings you've fixed already? Just strip off "unused variables / functions" patches (do it in a separate PR) and squash the rest into one bigger patch.