Most CS papers are theoretical, and this is another reason why HN is wrong about requiring papers to publish code.