Angelo Graziosi | 24 Apr 01:26 2014

About 'eager' warnings

Building Emacs trunk (cygw32), in the logs I find several warnings like 

Compiling emacs-lisp/eieio-base.el
Warning: Eager macro-expansion skipped due to cycle:
   … => (load "gv.el") => (macroexpand-all (defalias (quote 
gv--defun-declaration) …)) => (macroexpand (pcase …)) => (load 
"pcase.el") => (macroexpand-all …) => (macroexpand (dolist …)) => (load 
Wrote /work/emacs/lisp/emacs-lisp/eieio-base.elc

Are they to be expected?


Leo Liu | 22 Apr 12:24 2014

Add an optional arg ERROR-IF-EMPTY to register-read-with-preview

My primary emacs is now 24.3.90 and so I am using more of
register-read-with-preview and noticed one deficiency.

There are two types of commands that use register-read-with-preview:
a) those create registers b) those don't

For b), it seems more desirable to just quit then prompt the user. Thus
I would like to add an optional arg to register-read-with-preview and
fix all its callers accordingly. Comments and how possible it is to fix
it in emacs-24?


Leo Liu | 22 Apr 02:42 2014

ELPA screwup

Hi all,

Just now merging easy-kill updates into ELPA somehow all files are
merged into elpa/admin overriding things there.

How could I fix this? (It seems cannot reset HEAD to before my push).

Sorry about this. The process has been smooth in the past so I was
surprised to discover otherwise.


Richard Stallman | 21 Apr 18:28 2014

newline cache

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

I recall that some months ago I encountered a bug in the newline cache
which caused mail-fetch-field to include a extra newline in the value
of the X-RMAIL-ATTRIBUTES field.  ISTR that the bug was fixed.

But it happened again, in Emacs as rebuilt on March 19 from fairly
recent sources.

There was little hope I can find a way to reproduce it from a
standing start -- I could hardly hope to get the newline cache just

Does anyone remember what the fix was for that bug?


Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call.

Alan Mackenzie | 21 Apr 18:17 2014

Scrolling and follow-mode

Hi, Anders, hi, Emacs.

With follow-mode enabled, the functions follow-scroll-up and
follow-scroll-down don't work very well when given a numerical argument.

In follow-scroll-up, the flag follow-internal-force-redisplay gets set,
which causes the post-command-hook function (a) to bypass the code which
selects the window to put point in; (b) to force redisplay to rescroll
the current window.  Both of these are undesirable when point is near a
boundary of one window, and should pass into the adjacent window on the

I have fixed this by causing the two scrolling functions to set a flag
which causes follow-adjust-window to align its windows as its first
action, then allowing it to select the appropriate window for point with
the existing code.  There are one or two minor changes too.

Comments on this fix are welcome.  Here is the patch (based on the
recent trunk revision 116992.

*** emacs/emacs.bzr/trunk/lisp/follow.el	2014-02-15 19:53:57.000000000 +0000
--- follow.el	2014-04-21 16:15:28.000000000 +0000
*** 347,352 ****
--- 347,355 ----
  (defvar follow-windows-start-end-cache nil
    "Cache used by `follow-window-start-end'.")

+ (defvar follow-fixed-window nil
+   "When non-nil, the window which must not be scrolled.
(Continue reading)

Daniel Colascione | 21 Apr 02:57 2014

Tramp test suite and network access

Can we change the tramp test suite so that, by default, it doesn't try
to access the network? If a host isn't connected to the network, tests
can hang for a long time with no user-visible explanation, and even if
network connectivity is available, ssh to the local system isn't
guaranteed to work, or work without asking for input. (The tramp tests
constantly ask me whether I want to connect.)

It'd also be nice to give Emacs /dev/null as its stdin when it's running
the test suite. Is there any reason we want tests to be able to interact
with the user?

Bogolisk | 20 Apr 20:20 2014

ediff doesn't work in 24.3.90

running emacs pretest

emacs -Q

use ediff-buffers

type 'n' to go to next hunk

Debugger entered--Lisp error: (wrong-type-argument listp [#<overlay from 656 
to 669 in item.el>])
  ediff-set-fine-diff-properties-in-one-buffer(A 0 nil)
  ediff-make-fine-diffs(0 noforce)
  call-interactively(ediff-next-difference nil nil)

Óscar Fuentes | 20 Apr 16:31 2014

Trunk build broken?

make bootstrap fails on Kubuntu 14.04 64bits:

if test "no" = "yes"; then \
          rm -f bootstrap-emacs; \
          ln temacs bootstrap-emacs; \
        else \
          ./temacs --batch --load loadup bootstrap || exit 1; \
          test "X" = X ||  -zex emacs; \
          mv -f emacs bootstrap-emacs; \
Loading loadup.el (source)...
Using load-path (/home/oscar/dev/emacs/emacs/lisp /home/oscar/dev/emacs/emacs/lisp/emacs-lisp
/home/oscar/dev/emacs/emacs/lisp/language /home/oscar/dev/emacs/emacs/lisp/international
/home/oscar/dev/emacs/emacs/lisp/textmodes /home/oscar/dev/emacs/emacs/lisp/vc)
Loading emacs-lisp/byte-run (source)...
Loading emacs-lisp/backquote (source)...
Loading subr (source)...
Loading version (source)...
Loading widget (source)...
Loading custom (source)...
Loading emacs-lisp/map-ynp (source)...
Loading international/mule (source)...
Loading international/mule-conf (source)...
Loading env (source)...
Loading format (source)...
Loading bindings (source)...
Loading cus-start (source)...
Loading window (source)...
Loading /home/oscar/dev/emacs/emacs/lisp/files.el (source)...
Loading /home/oscar/dev/emacs/emacs/lisp/emacs-lisp/macroexp.el (source)...
(Continue reading)

Sam Fiechter | 20 Apr 01:50 2014

Bug Report -- replace-regexp-in-string from subr.el


I noticed that the lisp function replace-regexp-in-string from emacs does not pass the sub-expression match when the rep is a function call rather than a string.  
For Example:
  (defun samscat (s)
      (concat "[" s "]"))
 (replace-regexp-in-string "\\(foo\\).*\\'" 'samscat " foo foo" nil nil 1))

" [foo foo] foo"
" [foo] foo"

looking at the code, I notice that the sub-expression is hard-coded on line 3642 of subr.el.gz:
    (funcall rep (match-string 0 str)))
rather than:
     (funcall rep (match-string (or subexp 0) str)))

I hope this is helpful!  

Here's my about emacs screen:
This is GNU Emacs, one component of the GNU/Linux operating system.

GNU Emacs (i686-pc-linux-gnu, GTK+ Version 2.24.10)
 of 2014-03-30 on lenovo-laptop
Copyright (C) 2014 Free Software Foundation, Inc.

If you would like any other info please let me know.



Sam Fiechter
(302) 689-3798
Tomohiro Matsuyama | 20 Apr 07:36 2014

cl-loop's "with" destructuring-bind looks broken


I have faced an issue on when using my extension that
includes an expression like:

    (cl-loop with (a b c) = '(1 2 3) return (+ a b c))

This expression causes an error:

    Attempt to set a constant symbol: nil

On a stable release, at least 24.3.1, it works fine and evalutes to 6.

ANSI Common Lisp allows such destructuring-bind with "with" clause, by
the way.


Eric S. Raymond | 19 Apr 22:16 2014

Repo conversion for review, v3

Third review version of the repository conversion is here:

git <at>

This one is current to complete to 2014-04-17 10:59:22.
It has been repacked.

The conversion procedure is in an essentially final form
and will need at most very minor additions and polishing 
before conversion day. 

		<a href="">Eric S. Raymond</a>

The world is filled with violence. Because criminals carry guns, we
decent law-abiding citizens should also have guns. Otherwise they will
win and the decent people will lose.
        -- James Earl Jones