Re: [O] Hang on incomplete input
Rafael wrote: In Ubuntu 10.10, emacs 23.2 and recent org, I get an *Org-Babel Error Output* buffer, saying: Error: unexpected '}' in: ddply(x, } Execution halted I'm using: GNU Emacs 23.3 (from http://vgoulet.act.ulaval.ca/en/emacs/windows/) Windows 7 org-mode 7.7 ESS 5.14 What version of ESS do you have? Maybe that's the difference? Any suggestions for how to debug a hang like this? Is there a signal I can send it, to generate a stack trace (if signals are even possible in Windows, I'm not even sure)? -Ken CONFIDENTIALITY NOTICE: This e-mail message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution of any kind is strictly prohibited. If you are not the intended recipient, please contact the sender via reply e-mail and destroy all copies of the original message. Thank you.
Re: [O] Hang on incomplete input
Ken Williams ken.willi...@windlogics.com wrote: Rafael wrote: In Ubuntu 10.10, emacs 23.2 and recent org, I get an *Org-Babel Error Output* buffer, saying: Error: unexpected '}' in: ddply(x, } Execution halted I'm using: GNU Emacs 23.3 (from http://vgoulet.act.ulaval.ca/en/emacs/windows/) Windows 7 org-mode 7.7 ESS 5.14 What version of ESS do you have? Maybe that's the difference? FWIW, I got the same thing as Rafael: Linux 2.6.35-28-generic #50-Ubuntu SMP Fri Mar 18 18:42:20 UTC 2011 x86_64 GNU/Linux GNU Emacs 24.0.90.2 (x86_64-unknown-linux-gnu, GTK+ Version 2.22.0) of 2011-10-27 Org-mode version 7.7 (release_7.7.617.gb1f2) Forgive my ignorance: how do I check the ESS version? And what does ESS have to do with R? I thought this was an R code block, but as you can probably tell, I'm an ignoRamus. When I start R interactively, I get , | R version 2.11.1 (2010-05-31) | Copyright (C) 2010 The R Foundation for Statistical Computing | ISBN 3-900051-07-0 ` Any suggestions for how to debug a hang like this? Is there a signal I can send it, to generate a stack trace (if signals are even possible in Windows, I'm not even sure)? Probably the best thing to do is cut out the middleman: execute the code snippet in the appropriate environment directly - no emacs, no org, no babel - and see if you have the problem. There are differences of course that babel tries to minimize but it can only simulate certain cases. E.g when I enter the incomplete form in the interactive session, it keeps giving me a secondary prompt attempting to convince me to do something sensible: , | | dply(x, | + | + | + ) | Error: could not find function dply | ` babel does not have that luxury. If that does not bear fruit, you can M-x toggle-debug-on-quit, run the code block and press C-g to get a backtrace. Rinse, repeat to see whether you always stop at the same point. It's somewhat hit-or-miss but it can be effective sometimes. Nick
Re: [O] Hang on incomplete input
-Original Message- From: n...@dokosmarshall.org [mailto:n...@dokosmarshall.org] On Behalf Of Nick Dokos Sent: Thursday, December 01, 2011 9:21 AM Forgive my ignorance: how do I check the ESS version? And what does ESS have to do with R? I thought this was an R code block, but as you can probably tell, I'm an ignoRamus. ESS is the mechanism by which org-mode interfaces with a running R process. And I think I left out a crucial piece of information in my first message: I run R in a session buffer, my full .org file looks like this: ### #+TITLE: Test doc #+AUTHOR: Ken Williams #+BABEL: :session *Rt* * My Header Some code: #+begin_src R ddply(x, #+end_src ### With that change, is the hang reproducible? If that does not bear fruit, you can M-x toggle-debug-on-quit, run the code block and press C-g to get a backtrace. Rinse, repeat to see whether you always stop at the same point. It's somewhat hit-or-miss but it can be effective sometimes. I could *SWEAR* I had tried C-g to abort the running R command (and C-c C-c, and C-c C-g, and about a dozen other things), but couldn't get it to react. Now it aborts as desired. I must have not actually tried a simple C-g. I can never remember which keys abort which aspect of operation. When I'm armed with that C-g, the hang is a much smaller issue! -Ken CONFIDENTIALITY NOTICE: This e-mail message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution of any kind is strictly prohibited. If you are not the intended recipient, please contact the sender via reply e-mail and destroy all copies of the original message. Thank you.
Re: [O] Hang on incomplete input
Ken Williams ken.willi...@windlogics.com wrote:
-Original Message-
From: n...@dokosmarshall.org [mailto:n...@dokosmarshall.org] On Behalf
Of Nick Dokos
Sent: Thursday, December 01, 2011 9:21 AM
Forgive my ignorance: how do I check the ESS version? And what does ESS
have to do with R? I thought this was an R code block, but as you can
probably tell, I'm an ignoRamus.
ESS is the mechanism by which org-mode interfaces with a running R process.
Ok, thanks! I must have set it up at some point, but I had completely forgotten
about it. The README says it's 5.11.
And I think I left out a crucial piece of information in my first message: =
I run R in a session buffer, my full .org file looks like this:
###
#+TITLE: Test doc
#+AUTHOR: Ken Williams
#+BABEL: :session *Rt*
* My Header
Some code:
#+begin_src R
ddply(x,
#+end_src
###
With that change, is the hang reproducible?
Yup - I did it twice and got the backtrace at the bottom of this mail
both times. One funky thing is that it only happens every second C-c C-c
for me.
Note also that #+BABEL: is obsolete for versions of org more recent than
Oct. 20 or so. You seem to still be running 7.7 from late July or early
August, so you should still be OK, but when you next upgrade, you will
need to make changes to your files. I had to change it to #+PROPERTIES:
... for my org version.
A good starting point on this change is Eric Schulte's message entitled
Standardized code block syntax and Property Accumulation merged into Master
Nick
If that does not bear fruit, you can M-x toggle-debug-on-quit, run the co=
de
block and press C-g to get a backtrace. Rinse, repeat to see whether you
always stop at the same point. It's somewhat hit-or-miss but it can be
effective sometimes.
I could *SWEAR* I had tried C-g to abort the running R command (and C-c C-c=
, and C-c C-g, and about a dozen other things), but couldn't get it to reac=
t. Now it aborts as desired. I must have not actually tried a simple C-g.=
I can never remember which keys abort which aspect of operation.
When I'm armed with that C-g, the hang is a much smaller issue!
-Ken
CONFIDENTIALITY NOTICE: This e-mail message is for the sole use of the inte=
nded recipient(s) and may contain confidential and privileged information. =
Any unauthorized review, use, disclosure or distribution of any kind is str=
ictly prohibited. If you are not the intended recipient, please contact the=
sender via reply e-mail and destroy all copies of the original message. Th=
ank you.
Backtrace:
Debugger entered--Lisp error: (quit)
file-exists-p(/tmp/babel-5986bN8/R-5986N_O)
(not (file-exists-p file))
(while (not (file-exists-p file)) (sit-for (or period 0.25)))
org-babel-comint-eval-invisibly-and-wait-for-file(*Rt*
/tmp/babel-5986bN8/R-5986N_O
{function(object,transfer.file){object;invisible(if(inherits(try({tfile-tempfile();write.table(object,file=tfile,sep=\\\t\,na=\nil\,row.names=FALSE,col.names=FALSE,quote=FALSE);file.rename(tfile,transfer.file)},silent=TRUE),\try-error\)){if(!file.exists(transfer.file))file.create(transfer.file)})}}(object=.Last.value,transfer.file=\/tmp/babel-5986bN8/R-5986N_O\))
(let ((tmp-file (org-babel-temp-file R-)))
(org-babel-comint-eval-invisibly-and-wait-for-file session tmp-file (format
org-babel-R-write-object-command (if row-names-p TRUE FALSE) (if
column-names-p (if row-names-p NA TRUE) FALSE) .Last.value
(org-babel-process-file-name tmp-file (quote noquote
(org-babel-R-process-value-result (org-babel-import-elisp-from-file tmp-file
(quote (16))) column-names-p))
(cond ((eql result-type (quote value)) (with-temp-buffer (insert
(org-babel-chomp body)) (let ((ess-local-process-name (process-name
(get-buffer-process session (ess-eval-buffer nil))) (let ((tmp-file
(org-babel-temp-file R-))) (org-babel-comint-eval-invisibly-and-wait-for-file
session tmp-file (format org-babel-R-write-object-command (if row-names-p
TRUE FALSE) (if column-names-p (if row-names-p NA TRUE) FALSE)
.Last.value (org-babel-process-file-name tmp-file (quote noquote
(org-babel-R-process-value-result (org-babel-import-elisp-from-file tmp-file
(quote (16))) column-names-p))) ((eql result-type (quote output)) (mapconcat
(function org-babel-chomp) (butlast (delq nil (mapcar (lambda (line) (when ...
line)) (mapcar (lambda ... ...) (org-babel-comint-with-output ... ... ...)
\n)))
(case result-type (value (with-temp-buffer (insert (org-babel-chomp body))
(let ((ess-local-process-name (process-name (get-buffer-process session
(ess-eval-buffer nil))) (let ((tmp-file (org-babel-temp-file R-)))
(org-babel-comint-eval-invisibly-and-wait-for-file session tmp-file (format
org-babel-R-write-object-command (if row-names-p TRUE FALSE) (if
column-names-p (if row-names-p NA TRUE) FALSE) .Last.value
Re: [O] Hang on incomplete input
-Original Message- From: nicholas.dokos Sent: Thursday, December 01, 2011 12:19 PM Note also that #+BABEL: is obsolete for versions of org more recent than Oct. 20 or so. You seem to still be running 7.7 from late July or early August, so you should still be OK, but when you next upgrade, you will need to make changes to your files. I had to change it to #+PROPERTIES: ... for my org version. A good starting point on this change is Eric Schulte's message entitled Standardized code block syntax and Property Accumulation merged into Master Thanks for the heads up, I'll make the change. But it seems like the thread you referenced might only be tangentially related to a BABEL-PROPERTIES change: http://thread.gmane.org/gmane.emacs.orgmode/49247/focus=49309 Or if it's there, I'm not seeing it. =) -Ken CONFIDENTIALITY NOTICE: This e-mail message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution of any kind is strictly prohibited. If you are not the intended recipient, please contact the sender via reply e-mail and destroy all copies of the original message. Thank you.
Re: [O] Hang on incomplete input
Ken Williams ken.willi...@windlogics.com wrote: A good starting point on this change is Eric Schulte's message entitled Standardized code block syntax and Property Accumulation merged into Master Thanks for the heads up, I'll make the change. But it seems like the thread you referenced might only be tangentially related to a BABEL-PROPERTIES change: http://thread.gmane.org/gmane.emacs.orgmode/49247/focus=49309 Or if it's there, I'm not seeing it. =) You are probably right - I haven't had the time to keep up with the discussion, This change is slightly older and it was discussed a lot during October. And I got it wrong too: it's #+PROPERTY: , not #+PROPERTIES: ... Sorry about that. In a recent version of the info version of the manual, check out (info (org) Buffer-wide header arguments) Nick
[O] Hang on incomplete input
Apparently, org-mode Babel (or maybe Emacs, or ESS) will hang when given incomplete input, like the following: #+begin_src R ddply(x, #+end_src I was just tearing my hair out trying to figure out why my code was hanging, because of course it's much longer and I didn't notice the missing paren. Is this replicable on other people's platforms too? I'm using GNU Emacs on Windows, with org-mode 7.7 and ESS 5.14. There are other situations where I really wish org-mode (or Emacs, or ESS) didn't freeze up while evaluating R code - for example if I accidentally did 'debug()' on a function that's now being called, or even just on long-running calculations. Any chance evaluation could be done in a separate thread or something? Thanks. -- Ken Williams, Senior Research Scientist WindLogics http://windlogics.com CONFIDENTIALITY NOTICE: This e-mail message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution of any kind is strictly prohibited. If you are not the intended recipient, please contact the sender via reply e-mail and destroy all copies of the original message. Thank you.
Re: [O] Hang on incomplete input
Ken Williams ken.willi...@windlogics.com writes: Apparently, org-mode Babel (or maybe Emacs, or ESS) will hang when given incomplete input, like the following: #+begin_src R ddply(x, #+end_src I was just tearing my hair out trying to figure out why my code was hanging, because of course it's much longer and I didn't notice the missing paren. Is this replicable on other people's platforms too? I'm using GNU Emacs on Windows, with org-mode 7.7 and ESS 5.14. There are other situations where I really wish org-mode (or Emacs, or ESS) didn't freeze up while evaluating R code - for example if I accidentally did 'debug()' on a function that's now being called, or even just on long-running calculations. Any chance evaluation could be done in a separate thread or something? In Ubuntu 10.10, emacs 23.2 and recent org, I get an *Org-Babel Error Output* buffer, saying: Error: unexpected '}' in: ddply(x, } Execution halted
