Re: write-region-internal breakage (in conjunction with indirect buffers)
kifer(a)cs.sunysb.edu (Michael Kifer) writes:
It seems that write-region-internal doesn't check what VISIT
actually is
and simply gives an error in all cases.
I tested this on 21.4 (don't have the cvs sources handy).
It seems to work in 21.5.
Also, is it a big problem to add the 'clone (third) argument to
make-indirect-buffer, like in FSF version of emacs (so that both the major
and the minor modes would be inherited)?
It depends a bit on how much state you want to have copied. I see that
unlike XEmacs the FSF version also shares the text-properties between
the two windows, even when not cloning. The XEmacs behavior seems more
useful as a default (so you can do font locking in two different
modes). So what do we want to happen in the clone case? deep-copy of
all associated sructures?, shallow-copy? Hunt for buffer instances in
all specifiers and duplicate those? Should specifier fallback for an
indirect buffer go via the parent?
Copying the minor/major modes/local variable seems the easy part.[1]
Jan
Footnotes:
[1] I noticed that FSF Emacs also copied the "CVS 1.3" version
message, is that correct?