So, if I try to put GTK resource specs in faces.el like so:
(set-face-background 'secondary-selection
'(((x default color) . "paleturquoise")
((x default color) . "green")
((x default grayscale) . "gray53")
((gtk default color) . "paleturquoise")
((gtk default color) . "green")
((gtk default grayscale) . "gray53")
((mswindows default color) . "paleturquoise")
((mswindows default color) . "green")
((mswindows default grayscale) . "gray53"))
'global)
XEmacs compiled for X or TTY fails to dump, citing the error below. It
works fine if the code is built with GTK support, or if I comment out the
(gtk default color) specs in the above list.
Any clues? This would be one of those !@#%!@ we can't put GTK in type of
bugs.
-Bill P.
Loading /usr/local/src/gtk-xemacs-21.2/i686-pc-linux/debug/tty/mule/nounion/lisp
/faces.elc...*** Error in XEmacs initialization
(error "Invalid color instantiator" global)
*** Backtrace
really-early-error-handler((error "Invalid color instantiator" global))
check-valid-instantiator(global color)
# bind (result noerror specifier-type spec)
canonicalize-spec(global color nil)
# bind (rest result)
byte-code("..." [specifier-type res2 noerror spec-list result rest nil throw
c
ann-spec-list t signal error "Invalid list format" canonicalize-spec] 5)
# (catch cann-spec-list ...)
# bind (result noerror specifier-type spec-list)
canonicalize-spec-list((global ((x default color) . "paleturquoise") ((x
defau
lt color) . "green") ((x default grayscale) . "gray53") ((gtk default
color) . "
paleturquoise") ((gtk default color) . "green") ((gtk default grayscale) .
"gray
53") ((mswindows default color) . "paleturquoise") ((mswindows default
color) .
"green") ((mswindows default grayscale) . "gray53")) color)
# bind (is-valid nval how-to-add tag-set locale value specifier)
set-specifier(#<color-specifier global=<unspecified>
fallback=#<color-specifie
r global=<unspecified> fallback=(...) 0x186> 0x79a> (((x default color) .
"palet
urquoise") ((x default color) . "green") ((x default grayscale) .
"gray53") ((gt
k default color) . "paleturquoise") ((gtk default color) . "green")
((gtk defaul
t grayscale) . "gray53") ((mswindows default color) . "paleturquoise")
((mswindo
ws default color) . "green") ((mswindows default grayscale) .
"gray53")) global
nil nil)
# bind (how-to-add tag-set locale value property face)
set-face-property(secondary-selection background (((x default color) . "paletu
rquoise") ((x default color) . "green") ((x default grayscale) .
"gray53") ((gtk
default color) . "paleturquoise") ((gtk default color) . "green")
((gtk default
grayscale) . "gray53") ((mswindows default color) . "paleturquoise")
((mswindow
s default color) . "green") ((mswindows default grayscale) .
"gray53")) global n
il nil)
# bind (how-to-add tag-set locale color face)
set-face-background(secondary-selection (((x default color) .
"paleturquoise")
((x default color) . "green") ((x default grayscale) . "gray53")
((gtk default
color) . "paleturquoise") ((gtk default color) . "green") ((gtk
default grayscal
Show replies by date
(when (not (member 'gtk (specifier-tag-list))
(define-specifier-tag 'gtk)))
??
Of course you can't do that becuase it needs to not match things that
aren't gtk, but maybe there's some little thunk of code in the
gtk-specific code defineing the device tag (and its predicate?) that
actually belongs in generic file.
Or maybe it's just lacking the predicate; if the tag were undefined it
should barf with an intelligible message. But if it swallows the tag
and then applies it willy-nilly to X or tty resources.... (I don't
see why that would be a problem here, though.)
--
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091
_________________ _________________ _________________ _________________
What are those straight lines for? "XEmacs rules."