Re: valid forms of face property: documentation bug, font-lock-{prepend, append}-text-property bug

[Richard Stallman]

    NOTE: I'm briefly changing topic to something I noticed while
    investigating this.  Supposedly, facep returns non-nil for internal
    face objects, which are special vectors.  Do these internal face
    objects work as face property values?

I am not sure the question is important.

                                           Also, it appears that (facep
    'bold) in fact returns one of these internal face objects, but (facep
    (facep 'bold)) returns nil, so facep seems to violate its own
    documentation.  Is this what is happening?  Or is the vector returned
    by (facep 'bold) not one of the internal face objects?  Can anyone
    answer these questions?

It sounds like a documentation bug to me.  I will fix that.

    By the way, there are two bad implementations of the same concept in
    other places in Emacs.  They should both be replaced by calls to
    font-lock-prepend-text-property or should be changed to use
    font-lock-prepend-text-property.  First, rcirc-add-face function in
    rcirc.el is a bad implementation of the idea of
    font-lock-prepend-text-property which appears not to properly handle
    the case where the value of the face property is a symbol.  Second,
    erc-button-add-face in erc-button.el is yet another bad implementation
    of the same idea that does not handle the case where the face property
    is a list but is not a list of faces.

Could you write a fix for that?