WHY DO DEVELOPERS OF MAJOR PROJECTS UNDER LINUX USE INCONSISTENT CONFIGURE SCRIPT FLAGS !!!

SOME OF THE TIME YOU POINT A PATH WHERE A STANDARD SUBDIRECTORY IS

SOME OF THE TIME YOU POINT TO A WHERE THE FILE CAN BE FOUND PRECISELY

SOME OF THE TIME THEY WANT THE FULL FILEPATH !

AND THE DOCS DONT READ THAT WAY !

'--with-curl=arg path to curl-config'
...

ok /SomePath/bin ? right ? NO
/SomePath/bin/curl-config !!!

When running a script, you are determining the context of the script '.' as the directory you were in when you launched it. So you can use relative paths if you're always going to be in the same directory when you launch it.

My two cents is that absolute paths are almost always better.

@low-bar-hurdler actually found where that was bad because it caused a problem with pkgconfig

I know this, you use relative paths when pointing to files in the "app" or in the config directory, mainly to make life easier.

You use absolute paths when reading or writing to fikes outside of the app.

At least that is what I do :D
But if a file is withing the project, I normally chose that and won't let the user change it, so everything the user faces is absolute paths

Also I hate you fucking bot with generic bot rants

@mundo03 specifically were talking building and linking with the gnu c tool chain

When changing the prefix during the configure script operation I noted something strange post installation when I used the parent directory option when the lib was referenced later.

@MadMadMadMrMim ah miss me with that low lever shit.

@mundo03 heh i mostly understand it from having to do custom builds on linux from other peoples packages so there are alot of nuances missed on me as well lol