I'm not sure where to log this complaint, its not a real bug. I've had about 6 or 7 emerges like emerge -v -u -D world break in the middle somewhere because emacs has been called to byte compile or whatever, however whoever is writing this stuff is calling emacs WITHOUT taking the normal precations used by savvy emacs users.
That is, to call emacs with `emacs -q --no-site-file CMDS'. This is starndard practice to use when calling emacs during builds. But on a 6-7 occassions, the last in the middle of emerge -v -u -D --newuse world' Not something you are likeley to set and watch for something like this. So you come back in hour and half to find its errored out in the middle because emacs wasn't called with simple protective flags like -q --no-site-file. -q (don't use an .emacs init file) --no-site-file (don't load any site-start.el files) People might have code in either of those places that will cause emacs to give unexpected output during build. site-start.el is a tradition older than gentoo. Gentoo installs don't have it but many long term emacs users use this file to set code for system wide stuff. That is what it is for. Emacs will load it before anything else (except `default.el') if it is found in the path. That is why it is wise to use --no-site-file when calling emacs to bytecompile. In my site-start.el I have code that causes emacs to tell me what it is loading anytime a library or file is loaded. That output can be quite handy when debugging things emacs, but it breaks an emerge. So unless I realize this particular emerge will call emacs and move site-start.el out of the path it will be loaded and break the emerge becasue someone didn't have sense enough to call emacs the standard way: emacs -q --no-site-file WHATEVER Where do I need to take this complaint? -- [email protected] mailing list
