How I made my millions compiled emacs on an M1 running linux

First, you have to clone emacs and checkout your branch of choice (most likely a major version branch like `emacs29`). Go nuts, though, this step is half the fun of building from source.

Once you’ve done that, you have to generate a bunch of c files templatized for different operating systems:

./autogen.sh

twiddles thumbs

Then you have to choose what features to compile in:

./configure --with-pgtk \
            --with-native-compilation \
            --with-json \
            --with-xwidgets \
            --with-imagemagick \
            --without-sound

twiddles thumbs

Then you actually compile the program, build the docs, yada yada yada Is 10 the right number for this computer? Is -j, as I supposed, used for setting the number of cores used to compile? I didn’t check any of that, I just blindly followed the example of some macOS blog post. It was the same sort of hardware, anyway:

make -j10

Congratulations, you’re at the finish line. Nothing left but to install the files on your system:

sudo make install

A question for another day: how can I provide an arbitrary binary name for a given emacs build that won’t conflict with my trusty emacs daily driver command?