lock-up requiring C-g in R chunk with org-mumamo-mode

[Dan Davison]

Steps to recreate:

1. emacs -Q

2. load nxhtml and ESS
i.e. (require 'ess-site)

3. create Org-mode buffer containing:

--------------------------
* R block
#+begin_src R
  aa <- 1
#+end_src
--------------------------

4. M-x org-mumamo-mode

5. Move point in and out of R chunk and along the line inside the R chunk

This causes a lock-up in which no further movement of point is possible. C-g does exit the lock-up, but afterwards org-mumamo-mode is not properly functioning (e.g. the variable org-mumamo-mode has the value t, but chunks are not recognised, and fontification of chunks in the buffer is not correct).

The following is written to *Messages* when the lock-up occurs:

--------------------------
Making change-major-mode-hook buffer-local while locally let-bound!
OVERVIEW
Making change-major-mode-hook buffer-local while locally let-bound! [27 times]
Error in pre-command-hook: (quit)
--------------------------

GNU Emacs 24.0.50.1 (i686-pc-linux-gnu, GTK+ Version 2.20.1) of 2010-08-27 on Luscinia
nXhtml mode version 2.08
Org-mode version 7.01 (release_7.01h.227.g2f93)

ess-version is a variable defined in `ess-custom.el'.
Its value is "5.12"

Thanks,

Dan

[lborgman (lennart-borgman)]

Thanks for the bug report, Dan.

Where do I find ess-site? Which version of it are you using?

[lborgman (lennart-borgman)]

I found ess-site (I actually had it, sigh).

I can reproduce the problem with the latest nXhtml beta 2.09. I will look into it.

[lborgman (lennart-borgman)]

I have sent a mail to the ess maintainers and asked for help with this.

[lborgman (lennart-borgman)]

I have got no reply from the ESS maintainers so I do not know what to do. If anyone is willing to try to contact them again please do!

[Erik Iverson (3y9m2vc-erik-fkzsxrq)]

Hello, using the latest devel version of mumamo through github. (2.09beta) and emacs 24.1.1.

I think I made some progress on this. I was trying to get mumamo get work with markdown-mode + R through ESS. I experienced the same lockups when editing R code chunks. I could fix it by specifying:

(setq mumamo-set-major-mode-delay -1)

 I believe this comes down to the fact that while there is a function in ESS called "R-mode", the actual major-mode variable is called "ess-mode". So mumumo sees the chunk is in "ess-mode"and thinks it *should-be* in R-mode, so tries to switch it. In mumamo.el, this line in mumamo-set-major-post-command

(set-it-now (not (or in-pre-hook (mumamo-fun-eq major major-mode)))))

will assign t to set-it-now even when it does not have to, because of this R-mode vs. ess-mode incongruence. I believe this is related to the infinite loop that mumamo ends up in without setting the mumamo-set-major-mode-delay as above.

I also currently get things like the following in my *Messages* buffer, even with setting the delay to -1:

Switched to ess-mode [3 times]

so I'm just going to remove the message function call that is generating those lines for now. I realize this does not fix the problem per se, but it makes it usable for me, and hopefully this helps track down the larger issue.

Thanks, mumamo is great!