Re: [cjk] Fwd: HELP with emacs 25.1 and cjk-enc.el

[Hin-Tak Leung]

I think there are at least two problems. I can get emacs 25.1 to crash by 
adding just these two lines:

--- a/emacs/cjk-enc.el
+++ b/emacs/cjk-enc.el
@@ -881,7 +881,8 @@
               (setq last-pos (point))
               (message "Converting: %2d%%"
                        (/ (* 100 (point)) (point-max)))))
-
+        (message "temp-buf is %s..." (buffer-name temp-buf))
+        (message "work-buf is %s..." (buffer-name work-buf))
         ;; Advance to the next character and loop.
         (forward-char 1))
 
24.5 does not crash but don't produce the right answer. Both against 
cjk/examples/Big5.tex .

I got emacs 22 and 21 to work correctly; emacs 23 depends on which file and 
LANG/LC* and which example (cjk/examples/* and mine).

This is all running older fedora binaries on current fedora 25 - I did not 
build any new binaries.

The crash obviously shouldn't happen - 
They seems to have done a fair number of changes to the message buffer in 24.4, 
then again in 25.1 :
https://www.gnu.org/software/emacs/manual/html_node/elisp/Logging-Messages.html

I also suspect cjk-enc.el is relying on the current buffer is restored after a 
set-buffer goes out of scope: 
https://www.gnu.org/software/emacs/manual/html_node/elisp/Current-Buffer.html#Current-Buffer
This probably isn't the case, and especially as the emacs people rework'ed the 
buffer-list code (another emacs 25 feature).

At the moment, the most obvious thing is that 
"\ifx\CJKpreproc\undefined\def\CJKpreproc{cjk-enc}\RequirePackage{MULEenc}\fi" 
getting prepended to stdout
(the message buffer), because it should be prepended to the work buffer; so it 
seems that some buffer setting/swapping code is wrong.

If some of you have a non-fedora emacs 25, can you give the two additional line 
above a try and run it against 
cjk/examples/Big5.tex ? I wonder if it is fedora's emacs 25 being buggy. (I 
have also some firefox crashes which are specific to fedora-compiled firefox 
also, from its gcc 6).

It certainly looks like there is more than 1 problems.

--------------------------------------------
On Mon, 12/12/16, Werner LEMBERG <address@hidden> wrote:

 > I
 think that after I retrofitted my Mac book using Mac ports
 with emacs
 > 24.5 and no other change ...
 everything worked as before.
 > 
 >  did I clear up anything or confuse the
 issue even further?
 
 The
 missing ability to debug is a very old problem.  However,
 the
 problem with the cjk-coding itself might
 be a newer bug, so there is
 no contradiction
 that it works with 24.5 and fails with 25.x.
 
 
     Werner