jkitchin / scimax

An emacs starterkit for scientists and engineers

scimax - Awesome editing for scientists and engineers

Scimax is an Emacs starterkit for scientists and engineers. It provides a comprehensive configuration of Emacs for scientific programming and publishing.

See the manual for more details.

NEWS

[2023-05-30 Tue] scimax 3.0

Another 2+ year update :) This release primarily revolves around making scimax brew-installable. See *homebrew for details. This new method installs a command-line utility scimax that will start scimax for you.

There are some newish features. The org-db code has been expanded and includes full-text and image search now. It also provides an org-db-agenda interface. This code is still “just useful enough”; although I use it almost daily, it is still not perfect and has some issues that make it likely I will never separate it from scimax.

scimax now uses emacs-jupyter. See ./scimax-jupyter.org for some examples.

New files include:

./scimax-org-table.el

some convenience functions for tables

./scimax-projectile.el

projectile integration functions

./scimax-org-attachments.el

org-attachment enhancements for scimax

./scimax-slack.el

integration with Slack

./scimax-ob-flycheck.el

integration of flycheck with src blocks (experimental)

./ox-twee2.el

experimental exporter for org to twee (interactive fiction)

A potentially breaking change needed for homebrew integration is that I moved the user directory out of scimax. Scimax now honors the built in way of providing user files, e.g. in .emacs or .emacs.d.

[2021-01-02 Sat] new features

It has been about two years since I noted the last new features. Here are few new features:

1. org-db - index your org-files into sqlite and then be able to search them
   1. scimax-contacts - an org-db add-on to help you find and use contacts
2. scimax-editmarks - a new markup for editing org-files and more.
3. scimax-@-links - use the @ key to easily insert a link from many different sources
4. scimax-hydra - a leader key like menu to use scimax

[2018-02-28 Wed] scimax 2.0

It is an update of scimax, and the main new features are:

1. Major overhaul on ipython. See ./scimax-ipython.org.
2. A dashboard feature: ./scimax-dashboard.el
3. A new inkscape link: ./scimax-inkscape.el
4. You can put images on links with images in the description: ./scimax-link-thumbnails.el
5. Functional text: ./scimax-functional-text.el

And probably more.

Installation

homebrew

Homebrew is now the preferred way to install scimax. It should work on MacOS, Linux and Windows (via WSL). See https://github.com/jkitchin/homebrew-scimax for all the details.

First you should add the tap where the Formula is:

brew tap jkitchin/scimax

Then to install the default version:

brew install scimax

To get the cutting edge files from the git repository use this command.

brew install --head scimax

This will output some text that tells you how to load scimax in your init file. It will be adding something like this to ~/.emacs.d/init.el, or whatever file you use.

(load "/usr/local/opt/scimax/share/emacs/site-lisp/scimax/init.el")

If you install scimax by homebrew, you want some lines like this in your ~/.emacs.d/init.el file:

;; If you want preload variables
(load  "/Users/jkitchin/Dropbox/emacs/user/preload.el")

;; This is what homebrew installed
(load "/usr/local/opt/scimax/share/emacs/site-lisp/scimax/init.el")

;; This points to your user.el file
(load  "/Users/jkitchin/Dropbox/emacs/user/user.el")

archived methods

These probably still work, and of course you can set up scimax by simply cloning the repo and setting up your init file to load the parts of it you want.

Mac users

Run the next command in your terminal in the location you want to install scimax. The command will make sure you have homebrew, git, and emacs installed, and then will clone scimax and tell you how to use it. It will take some time to install. You need to install your own Python and LaTeX. These days I am using Python3 from Continuum IO and MacTeX from http://www.tug.org/mactex.

bash -c "$(curl -fsSL https://raw.githubusercontent.com/jkitchin/scimax/master/install-scimax-mac.sh)"

Alternative manual installation of scimax for Mac users

Alternatively, you can install homebrew yourself, install git from http://git-scm.com/download/mac, build your emacs like this:

brew install emacs --with-gnutls --with-imagemagick --with-librsvg --with-x11 --use-git-head --HEAD --with-cocoa

Alternatively, lately I have been using emacs-plus which claims to be faster for magit and has all those options enabled by default.

brew tap d12frosted/emacs-plus
brew install emacs-plus

This got me:

(emacs-version)

After that, I clone scimax like this:

git clone https://github.com/jkitchin/scimax.git

and launch emacs with:

emacs -q -l scimax/init.el

Note that scimax no longer loads your user files like it did before. You will need to add some lines to an init file to do this.

Linux users

Run this command. It checks for a git and emacs, but does not install them. You will have to use your package manager for that. You also need to install your own LaTeX and Python (and other languages you might want).

bash -c "$(curl -fsSL https://raw.githubusercontent.com/jkitchin/scimax/master/install-scimax-linux.sh)"

Windows users

Windows has always been the most difficult installation target, and it is the least maintained.

Install git (http://git-scm.com/download/win). Open a git bash terminal. Run this command.

bash -c "$(curl -fsSL https://raw.githubusercontent.com/jkitchin/scimax/master/install-scimax-win.sh)"

There is an emacs 25.2 binary in this repository for Windows. You should be able to run the scimax.sh command to launch scimax now.

Manual installation

You can use your own emacs if you have one (version 27 or more is recommended). You still have to install Python and LaTeX if you will be using those.

Clone the scimax repo where you want it.

git clone https://github.com/jkitchin/scimax.git

and launch emacs with:

emacs -q -l path/to/scimax/init.el

Alternatively, add scimax to your load path in your init file and just require what you want.

(setq scimax-dir "path/to/scimax")
(add-to-list 'load-path "path/to/scimax")

It is not uncommon to have to restart emacs a few times while all the packages from MELPA are installed. Windows seems to be like that.

Funding for scimax

scimax development has been partially supported by the following grants:

• National Science Foundation (Award 1506770)
• Department of Energy Early Career Award (DESC0004031)
• You can sponsor scimax development at https://github.com/sponsors/jkitchin.

scimax's Issues

Linux: Ispell not working correctly?

Loading scimax on a Linux machine gives the following message

Starting new Ispell process nil with en_US dictionary...
Error enabling Flyspell mode:
(stringp nil)

Selecting a misspelled word gives:

Starting new Ispell process nil with en_US dictionary...
ispell-check-version: Wrong type argument: stringp, nil

I can't verify whether this occurs on Windows at the moment.

Heading levels reset by ipython src blocks when exporting to ipynb

When exporting to org files to ipynb, evaluating ipython src blocks in a subheading causes the heading level to reset. Headings/subheadings that do not have intervening src blocks work as expected:

* Top level heading
This is a top level heading

** Second level heading
A 2nd level heading with a src block will appear as second level, but later 
subheadings at or below will be reset.

#+BEGIN_SRC ipython :session :results output drawer
import sys
print(sys.version)
#+END_SRC

#+RESULTS:
:RESULTS:
3.6.0 |Anaconda custom (64-bit)| (default, Dec 23 2016, 12:22:00) 
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)]
:END:

** Second level heading
This will revert to first level

*** Third level heading
    This will revert to a second level.

Not a big deal, so this can go in the "maybe someday" pile...
emacs26.0.50, org 9.0.5

Two versions of emacs on OSX?

I seem to have two different versions of Emacs on my system. A 26 and 24.5 one. When I open scimax through the terminal, it opens up in the 24.5 version. Is it possible to switch to the newer one?

Add org-edit-latex to scimax

Hi, John
I'm the author of PR #50
I've made that fragment editing feature into a standalone package: https://github.com/et2010/org-edit-latex (it's available on MELPA). A few bugs have been fixed since my last PR. It's more reliable now, along with some enhancements including support for inline math editing, toggle on/off with minor mode, etc.. Since I'm a big fan of scimax, so I think it's better to send a PR here too. But I'm not sure if you want to include another external package. So what's your opinion? Is that OK to add org-edit-latex.el to scimax? Please let me know. Thanks.

Regards.

Please maintain ox-clip in a separate repository

Now that ox-clip is available from Melpa, please consider maintaining it in a separate repository. If you want to maintain its history but have difficulties using git filter-branch to do so, then I can help you with that.

compatibility with emacs 25?

Awesome starter kit!
I wonder if scimax works well with emacs 25. I just installed everything, but encountered a lot of errors--3362 lines. Attached are the error messages.
errors.txt

PS: my emacs version is 25.1.1

Cannot change indent level in python mode

While editing python code, I realized I could not change indent level with tab. I found it very difficult to close loops, since pressing backspace put the contents on the previous line, and shift+tab did not do anything.

Exporting edit marks to pdf fails if there is a % in it

The edit mark below:

[[delete:some 10% edit marks]]

will cause latex compilation failure when export to pdf. The tex source from export looks like:

\textcolor{red}{some 10% edit marks}

It seems the % is not escaped and the rest of line will be treated as comments. The unclosed brace leads to the error.

could not read Username for 'https://github.com'

After successfully loading hunspell a new error is being returned:

fatal: could not read Username for 'https://github.com': No such device or address

It does not seem to interrupt anything, but it may be important.

Cannot find flycheck in melpa upon startup

A recent commit resulted in a new package that could not be downloaded. Upon startup the error given is.

File error: http://melpa.org/packages/flycheck-20160711.205.tar, Not found

Windows error: Eager macro-expansion failure

After attempting a new installation of Scimax on windows I got the following error when attempting to open an org file:

Eager macro-expansion failure: (void-variable path)
(Shell command succeeded with no output)
Eager macro-expansion failure: (void-variable path)
[yas] Reloaded everything (snippets will load just-in-time)....
Eager macro-expansion failure: (void-variable path) [3 times]Eager macro-expansion failure: (void-variable path)
(Shell command succeeded with no output)
Eager macro-expansion failure: (void-variable path)
[yas] Reloaded everything (snippets will load just-in-time)....
Eager macro-expansion failure: (void-variable path) [3 times]

This only prevents loading the file I selected directly. I can still navigate to it after this error occurs, but this is obnoxious.

Could not export scimax.org to LaTeX

I used C-c C-e to export scimax.org to pdf or to LaTeX, but I got this error message: Unknown LaTeX class ‘’.
I think this is a problem with the file scimax.org because I could export other org files to pdf.

New win10 Scimax installation not finding python(?) - Using Anaconda

Windows 10 64b w/ new Anaconda installation, 64b, Python 3.6 placed in C:\Ananconda3.
Tested Jupyter notebook and Spyder. Both execute python properly.
Created c:/Emacs, used git-bash in that directory for Win version. Scimax loaded.
Opened scimax.org, selected python programming example and received this fail message:

executing Ipython code block...
Making python-shell-interpreter local to Python while let-bound!
Making python-shell-interpreter-args local to Python while let-bound!
Sent python-shell-completion-setup-code
comint-send-string: Writing to process: invalid argument, Python
true
(Shell command succeeded with no output)

Does this mean it is not finding python/ipython? Do I have to set explicit lisp commands for a path to these?

Menu: Customize user.el

When I click on Customize user.el under the Scimax menu at the top of the emacs window, I get the following:

let: Symbol's value as variable is void: starter-kit-dir

A similar issue happens with the Help with scimax option:

command-execute: Symbol's function definition is void: scimax-help

when-let not available in Emacs 24

Hi John,

The following files use when-let, which I believe is not available in Emacs 24. Does it make sense to replace them with dash's -when-let, or replace assert_minimal_emacs_version 24 4 by 25 1? Thanks.

ox-ipynb.el:349: (when-let ((md (export-ipynb-markdown-cell (s-trim s))))
ox-ipynb.el:357: (when-let ((md (export-ipynb-markdown-cell (s-trim s))))
ox-ipynb.el:385: (when-let ((md (export-ipynb-markdown-cell (s-trim s))))
ox-ipynb.el:392: (when-let ((md (export-ipynb-markdown-cell (s-trim s))))
scimax-org-babel-ipython.el:221: (when-let (ret (ob-ipython--eval
scimax-org.el:1033: (when-let (f (buffer-file-name b))

Auctex installation solution

I'm not sure if this is the right spot, but I noticed in the packages.el file this:

;; installation is failing
;; (use-package auctex)

You'll have to do following:

(use-package tex
:ensure auctex)

This will install auctex.

Code block image output issues

Hey! thanks for putting together scimax, I stumbled upon an issue while exporting to PDF that I am trying to understand. Here is an example:

I have three R code blocks in which I generate a sample plot from ggplot2. The first one outputs the image correctly, in the second one I use an underscore in the #+name: attribute (for easier complex file naming) and the block does not generate the plot. In the third one I add a line of code which calls to output the plot (bp) which leads to a truncated/weird image in org-mode.

#+name: test
#+BEGIN_SRC R :results output graphics :exports both :file-ext png :res 180 :width 16000 :height 900
library(ggplot2)


bp <- ggplot(data=PlantGrowth, aes(x=group, y=weight, fill=group)) + geom_boxplot()
#+END_SRC

#+RESULTS: test
[[file:test.png]]

#+name: test_1
#+BEGIN_SRC R :results output graphics  :exports both :file-ext png :res 180 :width 16000 :height 900
library(ggplot2)

bp <- ggplot(data=PlantGrowth, aes(x=group, y=weight, fill=group)) + geom_boxplot()
#+END_SRC

#+RESULTS: test_1
[[file:test_1.png]]

#+name: test_2
#+BEGIN_SRC R :results output graphics :exports both :file-ext png :res 180 :width 16000 :height 900
library(ggplot2)

bp <- ggplot(data=PlantGrowth, aes(x=group, y=weight, fill=group)) + geom_boxplot()
bp
#+END_SRC

#+RESULTS: test_2
[[file:test_2.png]]

Now, when I try to export this file to either pdf or html, I get different results as you can see here wit HTML or here for PDF.

Any idea what is causing this?

Some abbrev characters (') are not word constituents

When I start Emacs with scimax I get the following error:

Error (use-package): scimax-org :init: Some abbrev characters (') are not word constituents in the standard syntax

These abbreviation have been introduced by:

ad9b0ca [ad9b0ca]

I'm using Emacs 25.1 on MacOS X with Org 9.0.5-435-g106ecb

For now I have disabled the contractions.
They aren't that useful for me since I normally write in german.

scimax-org-babel-ipython.el error: mapconcat: Symbol’s function definition is void: loop

Hi,

Great work!

I am trying to use only your version of ob-ipython, via:

(add-to-list 'load-path "~/.emacs.d/elisp/scimax")
(require 'scimax-org-babel-ipython)

and I get the error mentioned in the subject line above when running either of the following in a .org file:

#+BEGIN_SRC ipython :session :results output drawer
4 + 5
#+END_SRC

#+BEGIN_SRC ipython :session :results output drawer :async t
4 + 5
#+END_SRC

Here is *Messages* output:

Contacting host: localhost:9988
Code block evaluation complete.
error in process filter: mapconcat: Symbol’s function definition is void: loop
error in process filter: Symbol’s function definition is void: loop
executing Ipython code block...
Contacting host: localhost:9988
mapconcat: Symbol’s function definition is void: loop

If I remove the scimax config lines above and simply use ob-ipython via

(require 'ob-ipython)

then both src blocks above work fine.

Any thoughts as to what might be happening with the loop calls in (I think)
org-babel-execute:ipython?

I am using Emacs 25 and have tried two versions:

GNU Emacs 25.1.1 (x86_64-apple-darwin13.4.0, NS appkit-1265.21 Version 10.9.5 (Build 13F1911)) of 2016-09-17

and

GNU Emacs 25.2.1 (x86_64-apple-darwin13.4.0, NS appkit-1265.21 Version 10.9.5 (Build 13F1911)) of 2017-02-21

I am using org-version
Org mode version 9.0.5 (9.0.5-elpa @ ~/.emacs.d/elpa/org-20170210/)

I think the following is unrelated, but I include it anyway:

In both cases (ob-ipython or scimax version), I get the following warnings, but with no effect:

Warning (python): Python shell prompts cannot be detected.
If your emacs session hangs when starting python shells
recover with ‘keyboard-quit’ and then try fixing the
interactive flag for your interpreter by adjusting the
‘python-shell-interpreter-interactive-arg’ or add regexps
matching shell prompts in the directory-local friendly vars:
  + ‘python-shell-prompt-regexp’
  + ‘python-shell-prompt-block-regexp’
  + ‘python-shell-prompt-output-regexp’
Or alternatively in:
  + ‘python-shell-prompt-input-regexps’
  + ‘python-shell-prompt-output-regexps’
Warning (python): Your ‘python-shell-interpreter’ doesn’t seem to support readline, yet ‘python-shell-completion-native’ was t and "jupyter" is not part of the ‘python-shell-completion-native-disabled-interpreters’ list.  Native completions have been disabled locally. 

Thank you.

Inserting citation issue

Hi professor,

I met a problem of citation insertion.
There's only a message of "No PDF(s) Found" and no citation inserted. Would you please check this?
Thanks,

Tianyu

html-clip-w32.py created on Linux

My understanding is that html-clip-w32.py is for Windows systems. However, it is also created on my Linux system. As far as I can tell, the logic at

scimax/ox-clip.el

Line 349 in fcae1c3

(unless (and (eq system-type 'windows-nt)

is to blame (?)

Auto pairing?

Hi John,

While the autopairing of parentheses, quotes, etc is nice, I've noticed that if you delete the closing parentheses or quote, it seems to delete everything within the quote or the parentheses. Can you tell me which function/mode controls the pairing in the init files so I can learn more about how it works?

Make #+attr_org :width 300 work on Windows

There doesn't seem to be a Windows emacs with imagemagick support, so the rescaling doesn't work.

Design question for export-ipynb-keyword-cell in ox-ipynb.el

export-ipynb-keyword-cell removes the keywords in a hard-coded list when generating the keyword cell. Could this approach be inverted? That is, build the keyword cell based on a user-configurable "keyword include" list rather than a "keyword remove" list?

Motivation: the generic org-mode export templates (C-c C-e #) add many (often empty) keywords that get added to the ipynb keyword cell unless ox-ipynb.el or the export template is manually edited.

Thanks,
Steve

ox-ipynb.el is incompatible with Org-mode version 8.2.10

Looks like org-org-export-as-org no longer accepts body-only and info as parameters and expects ext-plist with export options.

So I had to change

(defun ox-ipynb-export-to-ipynb-file (&optional async subtreep visible-only body-only info)
  (let ((ipynb (concat (file-name-base (buffer-file-name)) ".ipynb")))
    (org-org-export-as-org async subtreep visible-only body-only info)
    (with-current-buffer "*Org ORG Export*"
      (setq-local export-file-name (or
				    (and (boundp '*export-file-name*)
					 *export-file-name*)
				    ipynb))
      (ox-ipynb-export-to-file))))

to

(defun ox-ipynb-export-to-ipynb-file (&optional async subtreep visible-only ext-plist)
  (let ((ipynb (concat (file-name-base (buffer-file-name)) ".ipynb")))
    (org-org-export-as-org async subtreep visible-only ext-plist)
    (with-current-buffer "*Org ORG Export*"
      (setq-local export-file-name (or
				    (and (boundp '*export-file-name*)
					 *export-file-name*)
				    ipynb))
      (ox-ipynb-export-to-file))))

ox-ipynb does not always preserve org subtrees using Markdown headings

Hi John, I just discovered that there are cases when level of nesting is not being preserved with Markdown headings - often a block that was a subtree in org file, becomes a H1 heading in Markdown.

The simplest example I could come up with would be content below, if you were to export it, are not preserved will become H1 heading instead of H2.

* Markdown levels
Some comment
#+BEGIN_SRC ipython
import numpy as np
#+END_SRC

** are not preserved
#+BEGIN_SRC ipython
import pandas as pd
#+END_SRC

The end.

sympy results are always converted to png

Pounding on scimax's ob-ipython tweaks, I find that all results from sympy are getting converted into large pngs. This makes viewing output in emacs difficult and confuses the html exporter, which inserts the pngs instead of sympy's latex for mathjax rendering. (sympy's init_printing() allows the html exporter to work effectively with sympy latex output and mathjax.)

Using ob-ipython without scimax tweaks produces standard sympy output in results, so may be related to the mods in the ob-ipython-async-callback / mime-types?

The issue is reproducible on linux, emacs 26.0.50 using scimax or spacemacs:

Trying sympy in org.
#+BEGIN_SRC ipython :session :results output drawer
from sympy import *
# commenting out init_printing() results in no output
init_printing()
#+END_SRC

#+BEGIN_SRC ipython :session :results output drawer
3+2
#+END_SRC

#+BEGIN_SRC ipython :session :results output drawer
var('x y')
x**2 + y
#+END_SRC

(Found this simple problem after some wt? moments: I kept getting BIG result blocks with "nothing" in them. Turns out my dark emacs theme was dutifully showing a large png with black text and a transparent (black) background...oof :)

Side question: I'm also getting a #+BEGIN_EXPORT latex block regardless of exporter; it's benign, but should this happen for only latex exports?

Thanks,
Steve

ox-ipynb: underscores always exported as subscripts in markdown cells

Trying to include text with an underscore seems to always produce <sub></sub> in markdown cells. I have tried
#+OPTIONS: _:{}
#+OPTIONS: ^:nil
and escaping (i.e., '..use the pd.read\_csv() function to...) with no luck. Is this reproducible for others? (I use only ox-ipynb from scimax, so please try to reproduce first). Thanks

Perl locale setting error

The following error message is produced upon opening scimax in a new window:

perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
    LC_ALL = (unset),
    LANG = "ENU"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
    LC_ALL = (unset),
    LANG = "ENU"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").

This is minor since the error seems to have no impact on functionality, and does not prevent files from opening correctly like previous errors. This only occurs on Windows.

Ispell/hunspelll setup

I am trying to set up hunspell with the en_US dictionary.

every time I open a project with projectile I get the following message in my mini-buffer
(ispell-phaf: No mathcing entry for en_US in 'ispell-hunspell-dict-paths-alist".)

I installed hunspell via brew with:

brew install hunspell

and when I check hunspell -D I get the following:
AVAILABLE DICTIONARIES (path is not mandatory for -d option): ./en_US Library/Spelling/en_US Can't open affix or dictionary files for dictionary named "default".

Now I thought of adding the link to the .aff file as suggested by Alex Gromnitsky like this:

(setq ispell-hunspell-dict-paths-alist '(("en_US" "~/Library/Spelling/en_US.aff")

without success.

I had a look at your packages.el file and noticed the following:
;; spell-checking on windows (setq ispell-program-name (expand-file-name "emacs-win/bin/hunspell" scimax-dir))

with the comment that it is spell checking under windows. I am on OSX with Emacs 26. How do I set up hunspell to work?

ox-ipynb.el thows error if no src blocks exist

Edge case: if no src blocks are present in org file, ox-ipynb-get-language() throws a wrong-type error.

I added a simple check to ox-ipynb-get-language(), and export-notebook produced an ipynb without specifying a kernel. Unsurprisingly, Jupyter didn't like that, so maybe an arbitrary kernel should be assumed if one isn't defined in the org file. Python?

ws-butler missing for fresh Windows installation

The following error occurs from a fresh installation on Windows:

url-insert-file-contents: http://melpa.org/packages/ws-butler-20150126.759.el: Not found

Commenting out the import of the package from packages.el is a temporary solution to the problem.

Allow exporting R code blocks to Jupyter notebooks with ox-ipynb

Jupyter supports running R as a kernel, so it can be used to present research in R code as well as Python.
At the moment ox-ipynb will export R code blocks as markdown, would be great to make it export them as code cells.

org-src-mode autocompletion clashes with elpy

Hi John,

If I do C-c ' to go into python-mode, both elpy and org_src autocompetion seem to be on, resulting in weird conflicts. See snapshot below. What do you think is a good fix?

Executing python in org-mode code blocks always outputs None

Not sure if this behavior is correct, but it is definitely different than old jmax. Any successfully run python code always outputs "None".

Checkbox colors

I noticed a little while ago that the checkbox colors were updated after a commit. The change went from grey backgrounds to white. It worked for blank and done boxes, but missed intermediates. See this pic.

Was this scimax coloring or did a theme maker change this?

"comment-or-uncomment-region" command looses cursor position

I noticed that the comment-or-uncomment-region "works" in org-mode now. Unfortunately, it also resets cursor to the beginning of the code block which is disruptive. Any way to have this information stored and later returned before switching into python mode?

Useful bug that might be a feature? Bold delimiters and starting new cells with ox-ipynb

Continuing to play with the awesome ox-ipynb, I discovered it will start a new cell if the first characters on a line are org-mode bold delimiters, e.g: **blah**. Turns out this is useful, but I'm wondering if there's an "official" way to start a new cell in ox-ipynb. If not, does it make sense to have something like #+ipynb-newcell?

Motivation: I tried the collapsible heading nbextension which, as suggested in #47 , makes long notebooks more navigable (I'm sold!). However, the extension collapses at the cell level; i.e., text is not collapsed if it's in the same cell as a heading. Starting a new cell immediately after the heading leads to a "more collapsible" notebook. (Also more along the lines of MMA if that's important :).

Example of org bold feature-bug working with collapsible headings extension:

* First heading
This text won't be collapsed using the nbextention.
* Second heading
The org parser will split the text below into a new cell. Accidentally useful!

** ** This kludge will start a new cell after the heading, enabling this
text to be collapsed. However, you get a non-breaking space indent in the text.

** A sub heading
Can also be split
**with** bold intact.

keybinding conflicts

I use ergoemacs-mode (https://ergoemacs.github.io/). But with scimax, a lot of my ergoemacs keybindings conflict with the ones defined by the packages preinstalled in scimax. Would you please suggest a solution? Thank you!

async pdf export + DocView

Hi John,

I was wondering if you can suggest a way to do this.

While writing a paper or a report, I would like to see a pdf on the side while I am typing, which automatically refreshes to sync with the changes I make in org-mode (perhaps in a periodic interval or every time I save the org file). I think this is pretty standard in something like overleaf or sharelatex, and it would add to the writing experience in scimax.

I know I can asynchronously export org to pdf like this (org-latex-export-to-pdf 't) and get the pdf open in a split buffer using DocView. To refresh the pdf, I tried auto-revert-mode on the DocView buffer, but it doesn't seem to work. It can't convert a half written pdf source to png and refuses to work once the final pdf file is written. The solution is to kill the buffer and reopen the pdf.

It seems like it should be possible to automate this whole process (and preferably to open the pdf on the last read page, and shrink/zoom it to fit the buffer), but I don't know enough elisp to do this very easily. Hopefully you can help ;)

Error enabling Flyspell mode

I got an error message anytime I try to turn on flyspell mode after I run (org-agenda-to-appt):

Error enabling Flyspell mode:(ispell-phaf: No matching entry for en_US in ‘ispell-hunspell-dict-paths-alist’.)

However, no error will occur if I turn on flyspell mode first and then run (org-agenda-to-appt). Any suggestions how to solve this issue?

Thank you very much!

Disable forceful lisp editing

Is the comment completion and parenthesis generation from the lispy mode package?

How can you remove the lambda hooks found in packages.el after they have been created?

Turn an org-link into a PDF link?

If I have a link to another org file and then I produce a PDF, the PDF will still link to the org file.

Is there a way to turn the org-link into a PDF link given that the linked PDF is in the same folder as the org file linked and has the same name?

ox-ipynb generates Markdown blocks with a null source

Hi John,

I've noticed this behaviour a little while ago and not sure if you had experienced any of this.
Basically, if there's no text after the last src block in an org file, exporter generates a Markdown cell
with the following content

{ "cell_type": "markdown", "metadata": {}, "source": [null] }

The problem is that [null] is not considered valid structure by Jupyter and it does not allow to open the notebook, where [""] does work.

Here's the sample org file that might help you to reproduce the problem:

#+BEGIN_SRC ipython
print "Hello"
#+END_SRC

Windows error: Invalid Parameter - /Users

When navigating to an org file in windows, if that org file has image links the following error is displayed:

Invalid Parameter - /Users
File mode specification error: (error "Cannot determine image type")

Attempting to run C-c C-x C-v in that same org file produces the following error:

Invalid Parameter - /Users
image-type: Cannot determine image type

Clicking on a .png image will open the figure in a new buffer. However, clicking a .gif results the display of the image as text. Attempting to use C-c C-c to produce the images results in the following error:

Type C-c C-c to view the image as an image.
Cannot display image: (Invalid image specification)

The first two errors occur regardless of whether there are .gif files in the org document or not. None of these errors occur on a Linux machine.

Incorrect path for SCIMAX_ROOT with portable Git on Win 7

I installed portable git as a regular user on Win 7.

(You may add to the instructions that one should configure the User Name and E-Mail address in git before installing scimax.)

Inside the bash I ran the recommended incantation.

The SCIMAX_ROOT was "s:imacs ..." because the result of pwd was /scimax
Maybe I should have cded to ~ first.
(Using the non-portable version gives the right path.)

I added a cd to my documents folder in front of the runemacs thus I start at the right place.

So far everything else looks good.

Axel

wow..this is amazing :) user defined colors/bold etc possible?

wow..just wow..this is amazing :)

im now spoiled with both the new color links you created (which i use alot now for my org-reveal presentations!..posted on that in the org list and many colleagues here at uni ;-)) and now this which seems even easier to add highlights.

I have the same question i asked with color links before regarding user defined colors/bold etc, that is to add highlight types such as colorX that has a FG:blue and BG: white and BOLD and then a colorY with a FG:yellow and BG: red etc..

thx so much again, your hard work is really appreciated here in my lab :D

PS will this come to melpa eventually and or get integrated into org?

best

Z

pygments-ipython-console prompt at startup

I think this may be Python 2.7 related. I get the following message when opening scimax on my local Linux machine:

Downloading/unpacking git+git://github.com/sanguineturtle/pygments-ipython-console
  Cloning git://github.com/sanguineturtle/pygments-ipython-console to /tmp/pip-vy16Mr-build
  Running setup.py (path:/tmp/pip-vy16Mr-build/setup.py) egg_info for package from git+git://github.com/sanguineturtle/pygments-ipython-console

  Requirement already satisfied (use --upgrade to upgrade): pygments-ipython-console==0.2.8 from git+git://github.com/sanguineturtle/pygments-ipython-console in ./pygments_ipython_console-0.2.8-py2.7.egg
Requirement already satisfied (use --upgrade to upgrade): setuptools in /usr/lib/python2.7/dist-packages (from pygments-ipython-console==0.2.8)
Cleaning up...

As far as I can tell, this isn't limiting functionality in any way, but it would be nice to resolve what is causing it so I don't have to see it.

Python mode automatically spaces ( after function

The python-mode editor seems to add one white space automatically when typing a parenthesis (). I do not think this is PEP8, namely when the () is in front of a function. Is there a way to change this?

ox-clip package install doesn't install html-clip-w32.py

I installed the melpa ox-clip package but when I select a region of an org mode file and run

M-x ox-clip-formatted-copy

I get the error message

python: can't open file 'c:/Users/scott/OneDrive/scotto/.emacs.d/elpa/ox-clip-20161106.823/html-clip w32.py': 
[Errno 2] No such file or directory
Quit

The documentation says that html-clip-w32.py will be installed on Windows but it isn't in my case. I'm using windows 10 and emacs 25.1.1