[O] Bug: transposition of words surrounded by forward slash [9.0.5 (release_9.0.5-266-gca31fa )]

2017-02-11 Thread Justin Kirby


An org file with the following content:

#+OPTIONS: *:nil

/simple/ /example/



During export org will transpose the word 'example' so the final output is

//examplesimple/ /



This will be the output regardless of export format. I have tried html,
txt, odt, md.

The words will be transposed as long as they are on the same line,
regardless of how many characters are between them. 

This was observed as early as org 8.3.5.

Changing the italic character in org-emphasis-alist to | does not change
this behavior.


Any ideas where to start looking for this?

thnx,
Justin





Re: [O] Using CEDET modules from Emacs core

2017-02-11 Thread Stefan Monnier
SM> For one, I'd like to see more major modes come with support for Semantic
SM> right in the major mode's own definition (rather than have it part of
SM> CEDET). E.g. for Elisp mode, CC-mode, ...
SM> The idea is to get to the point where Semantic support is just another
SM> thing that a major mode should aim to support alongside syntax-tables,
SM> indentation, font-lock, outline-minor-mode, ...
> Is the semantic support really at the point of warranting that?

What more would you want?

> Does it have many users currently?

Chicken, meet Egg!

> Is it something major-modes would want to include default support for?

It provides a functionality that's usually requested by users, so I'm
surprised you'd ask.

We don't have enough infrastructure support in general for "advanced"
editing functionality such as type/scope-aware completion.  So we have
various add-on thingies (like company-mode, cedet, and auto-complete)
which provide such support for specific modes, but really these should
move to core, so that major modes can themselves provide support for
such completion.

I've done some of that with completion-at-point-functions and moving
some company-.el to -mode.el, but there's a lot more to do.


Stefan



Re: [O] Sync up the org in emacs master to org maint branch?

2017-02-11 Thread John Wiegley
> "DE" == David Engster  writes:

DE> I find this whole core vs package argument strange. If you ship Emacs with
DE> Org, Gnus and CEDET, they are part of Emacs, so it's in the interest of
DE> all Emacs developers that they work well, whether they use them or not.
DE> The users won't care if they originate from a separate repo and are
DE> considered a "package". So if Paul is determined to fix all occurences of
DE> "compatilibity" in the doc-strings, why would he only do that for core?

If pulling Emacs.git also pulls ELPA.git (as it should, since it's within our
purview as well), then global search&replace should still affect all the
files.

I hear your other points, so I'm curious now as to what more people think
about this who work on Emacs core: Do you want more modularity with regard
ELPA, or does the monolithic model work better for you?

-- 
John Wiegley  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com  60E1 46C4 BD1A 7AC1 4BA2



Re: [O] Sync up the org in emacs master to org maint branch?

2017-02-11 Thread John Wiegley
> "PL" == Phillip Lord  writes:

PL> The alternative proposal is, essentially, to copy files into the Emacs
PL> core build structure and move from there.

PL> Reading the discussion reinforces my feeling that the latter is the wrong
PL> approach, because it reinforces a distinction between packages on ELPA and
PL> packages in core above and beyond the location that they are stored and
PL> versioned. I can't see the advantage of doing this.

PL> I will probably try to work a little further on my package.el solution,
PL> although there seems little enthusiasm for this as the way forward.

Correct, I don't want to move forward with the package.el option yet. Just
copying files into the core build structure should be very lightweight, and it
lets us focus on the ELPA separation first. I believe the package.el coupling
is something we can examine at a later date.

-- 
John Wiegley  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com  60E1 46C4 BD1A 7AC1 4BA2



Re: [O] Using CEDET modules from Emacs core

2017-02-11 Thread John Wiegley
> "SM" == Stefan Monnier  writes:

SM> For one, I'd like to see more major modes come with support for Semantic
SM> right in the major mode's own definition (rather than have it part of
SM> CEDET). E.g. for Elisp mode, CC-mode, ...

SM> The idea is to get to the point where Semantic support is just another
SM> thing that a major mode should aim to support alongside syntax-tables,
SM> indentation, font-lock, outline-minor-mode, ...

Is the semantic support really at the point of warranting that? Does it have
many users currently? Is it something major-modes would want to include
default support for?

The last time I tried it, for C++ code, it was far too slow. Are you saying
it's effective for other languages, like Python or Javascript or Go?

-- 
John Wiegley  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com  60E1 46C4 BD1A 7AC1 4BA2



[O] Latex summary

2017-02-11 Thread Russell Adams
I frequently write documentation with recommendations nested in
different sections, typically one line per recommendation. I then
summarize a list of those recommendations at the end of my document.

Today I use babel to grep the document for my recommendation format,
but given I'm exporting to Latex can anyone suggest a better method?
Something like a reversed table of contents?

I'm not as familiar with Latex packages.

--
Russell Adamsrlad...@adamsinfoserv.com

PGP Key ID: 0x1160DCB3   http://www.adamsinfoserv.com/

Fingerprint:1723 D8CA 4280 1EC9 557F  66E8 1154 E018 1160 DCB3



Re: [O] Help checking orgcard.pdf

2017-02-11 Thread Charles C. Berry

On Sat, 11 Feb 2017, Kyle Meyer wrote:


"Charles C. Berry"  writes:



[snip]




Also AFAICS, this one is not current either:

\key{toggle pretty display of scripts, entities}{C-c C-x {\tt\char`\\}}


Hmm, C-c C-x \ is bound to org-toggle-pretty-entities, so the binding is
correct.  Is the description no longer accurate?



My bad. I *misread* the TeX as rendering =C-c C-x `=. Sorry.

The rest of that section looks good now.

Chuck



Re: [O] Help checking orgcard.pdf

2017-02-11 Thread Kyle Meyer
"Charles C. Berry"  writes:

> Sorry not to be sending a patch, but I figured I should at least mention 
> that these are not current:
>
>
> \key{export visible part only}{C-c C-e v}
> \key{insert template of export options}{C-c C-e t}

With e50cd98f9, I've update these, as well as added keys for other
options listed in (info "(org)The Export Dispatcher").

> Also `C-c C-e' brings up the export dispatcher menu (by default) where 
> options to select the visible part or insert a template are revealed. So, 
> mention of the correct incantations need not appear in the refcard.

True, but because there's space on the page and because commands for the
agenda dispatch are mentioned, I chose to leave them (well, actually add
more).

> Also AFAICS, this one is not current either:
>
> \key{toggle pretty display of scripts, entities}{C-c C-x {\tt\char`\\}}

Hmm, C-c C-x \ is bound to org-toggle-pretty-entities, so the binding is
correct.  Is the description no longer accurate?

-- 
Kyle



Re: [O] org-time-stamp, day format

2017-02-11 Thread Uwe Brauer

   > Nick Dokos  writes:

   > And if you want to fix this in Emacs only:

   >   ;; System locale to use for formatting time values.
   >   (setq system-time-locale "C") ; Make sure that the weekdays in 
the
   > ; time stamps of your Org mode 
files and
   > ; in the agenda appear in English.

Aha, thanks that looks helpful, however the documentation for that
variable is nonexistent. How do I know its options and that "C"
corresponds to English?





Re: [O] Babel R source blocks problem

2017-02-11 Thread Charles C. Berry

On Fri, 10 Feb 2017, Thomas S. Dye wrote:


Aloha Chuck,

Charles C. Berry writes:


A look at the `slidify-example.org' file in the `org-9-plus' branch of my
(chasberry) `orgmode-accessories' github repo might be a good start with
.Rmd exports even if you do not want to use `slidify'.


Your ox-ravel exporter is a pleasure to use.  I'm farther along than I
thought I'd be at this point.



Thanks. I am glad to hear this.



I'm new to much of the R ecosystem, including slidify, so forgive some
noob ignorance in the following three questions:

1) I've specified the properties EXPORT_TITLE and EXPORT_AUTHOR, but
they don't show up in the .Rmd file.  Is this a limitation of ox-ravel,
or do I need to learn how to insert them in the Org mode source?



Both, I guess.

Arguably, it is a limitation in the markdown backends available that
there isn't one that will create the yaml header from the export
options in the org buffer.  I think the long term fix is to create a
backend derived from 'md, but this is far down my todo list.  For now,
either hand tool a yaml header like that in

: slidify-example.org::#tds-headline

(If you are viewing that on my github repo, be sure to click 'raw' or the 
yaml header will not be visible!)


or create a src block that will pull the title, author, et cetera and wrap 
them as a yaml header in an md export block and put it at the top of the 
output. If you want to do the src block, I can provide a skeleton version.




2) How do I instruct ox-ravel to include the graphics output of a source
code block?  I tried :ravel fig=TRUE as a header to the source code
block, but no dice.  Most of my experience is with ox-latex, so that
colors my expectations here; i.e. I'm expecting a [[file:myfile.pdf]]
link to produce a graphic in the output.


Often it `just works', but it depends on the vignette engine. `fig=TRUE' 
is correct for Sweave. See


: demos.org::#36234656-157D-4F1D-B441-E727DFCC0251

for an example of that. For knitr, you usually do not need to include 
anything if there is just one plot produced, but there are loads of 
options. See


https://yihui.name/knitr/options/#plots


I find myself using ~:ravel fig.cap=my_caption~ (where `my_caption' is
an R string from an earlier src block) a lot. See

: demos.org::#C341969B-90F3-4C27-AA7F-5352A9669506

for an example of a lengthy caption done in latex.

There is a src block that produces a plot in the slidify-example.org file. 
No special header was needed for ~slidify()~ to render it.  But I cannot 
be sure whether the yaml block does something special for the graphics.




3) I've been knitting the .Rmd output using RStudio, which is also new
to me.  I don't mind doing this (it is what the session organizer is
going to do with the .Rmd file I send next week), but I'm wondering
if you have an emacs-only solution?



I don't know RStudio, but here is a start:

#+BEGIN_SRC R :exports none :results silent
require(rmarkdown)
render("my-file.Rmd")
#+END_SRC

If you look at the =demos.org= file, you will see examples and the src 
blocks used to run ~knitr::knit2pdf()~.  If you export some of the 
examples to *.Rmd files and run the equivalent ~render(...)~ on the .Rmd 
files, you will get html by default.


---

Let me know how it works out!

Best,

Chuck



Re: [O] Adding #+LATEX: \newpage before section header using org-export-before-parsing-hook

2017-02-11 Thread Joon Ro
I spoke too early - if I have two headings with :newpage: tag in a row, my 
current code below skips the second heading.

It seems (setq org-map-continue-from (outline-next-heading)) in my code is the 
problem .. any help will be appreciated!



(defun org/parse-headings-newpage (backend)
  ; add \newpage to headings with :newpage: tag
  (if (member backend '(latex))
  (org-map-entries
   (lambda ()
 (insert-string "#+LATEX: \\newpage\n")
 (if (outline-next-heading)
   (setq org-map-continue-from (outline-next-heading)))
 )
   "+newpage"))
)

(add-hook 'org-export-before-parsing-hook 'org/parse-headings-newpage)




From: Emacs-orgmode  on 
behalf of Joon Ro 
Sent: Wednesday, February 8, 2017 8:58:40 PM
To: Nick Dokos; emacs-orgmode@gnu.org
Subject: Re: [O] Adding #+LATEX: \newpage before section header using 
org-export-before-parsing-hook


> So you'll have to manipulate org-map-continue-from appropriately.

Thanks a lot! Adding (setq org-map-continue-from (outline-next-heading)) after 
insert-string seemed to solve the problem.
(I added "newpage" to org-tags-exclude-from-inheritance, so the newpage does 
not get applied to subheadings)

(defun org/parse-headings-latex-newpage (backend)
  ; add \newpage to headings with :newpage: tag
  (if (member backend '(latex))
  (org-map-entries
   (lambda ()
 (progn
   (insert-string "#+LATEX: \\newpage\n")
   (setq org-map-continue-from (outline-next-heading))
   ))
   "+newpage"))
)

(add-hook 'org-export-before-parsing-hook 'org/parse-headings-latex-newpage)
(add-to-list 'org-tags-exclude-from-inheritance '"newpage")

Best,
Joon


[O] reverting to level 4 text

2017-02-11 Thread Sharon Kimble

Using the following example, how can I get the sentence starting "Data
confirms that..." to be under a level 4 heading please?

Or, in other words, how can I kill the level 5 heading after
"(cite:0580)" so that "Data confirms that ..." is under a level 4
heading please?

--8<---cut here---start->8---
 Legal Limits in the UK

* England and Wales drink driving limit

The maximum \gls{BAC} limit in England and Wales is -

- 35 micrograms of alcohol in 100 millilitres of breath, or

- 80 milligrams of alcohol per 100 millilitres of blood, or

- 107 milligrams of alcohol per 100 millilitres of urine (cite:0580).

Data confirms that aggressive behaviours are present for drinkers who have a 
\gls{BAC} of .08 or
higher, thus reminding us that 40%--60% of those jailed nationwide are 
convicted of crimes
committed under the influence of alcohol at this rate or higher.
--8<---cut here---end--->8---

Thanks
Sharon.
-- 
A taste of linux = http://www.sharons.org.uk
TGmeds = http://www.tgmeds.org.uk
Debian 8.6, fluxbox 1.3.5-2, emacs 25.1.1.1


signature.asc
Description: PGP signature


Re: [O] org-annotate/collaboration?

2017-02-11 Thread Uwe Brauer
>>> "Eric" == Eric Abrahamsen  writes:

   > Matt Price  writes:
   >> Does anyone use org-annotate actively? I'm wondering what your
   >> workflow is, how you incorporate comments, etc.  

   > I wrote it, and I don't use it that much. I do use it for quick
   > notes-to-self when writing, but footnotes do the job just as well.

   >> I'm hoping to embark on a book project with a colleague. I would like
   >> to use org-mode if I can, but I need to get a sense of the
   >> collaboration workflow. When you work on projects together, do you use
   >> annotations? Or git pull requests? If the latter, od you use any
   >> filters, or any magit tricks, to approve or modify suggested changes
   >> chunk by chunk?  

   > It's a huge problem, and one that org-annotate isn't going to solve. I
   > do a lot of manuscript editing, and passing files around, and have only
   > barely gotten some people to accept my "weird" workflow, which is to
   > send them a clean version of an edited file, and along with that an HTML
   > file containing htmlized word-diff output, where the insertions and
   > deletions are colorized. They make further edits on the clean copy, and
   > I do another go-around. It's a huge pain.

I did (and still do) the same, using latex and latexdiff, but found out
that a better solution is to use mercurial and bitbucket (I presume git
should be fine as well), since one of my collaborators agree to use it
as well. This is quite a relief to the former method relying on external
tools and email.  

-  Usually instead of comments I use issuesin bitbucket.
-  hg diff is not perfect but a good first approximation.




[O] Bug: org-eww-copy-for-org-mode Empty Link problem + easy fix [9.0.4 (9.0.4-elpaplus @ /home/andrea/.emacs.d/elpa/org-plus-contrib-20170124/)]

2017-02-11 Thread Andrea

Hello,

thanks hugely for org mode: the more I learn about it the easier is my
life organization!

I was trying to use org-eww-copy-for-org-mode (org-eww package in the
contrib/ directory) within eww in website like www.indeed.co.uk, and
every time I tried, I was receiving the error "Empty Link".

The error is due to the presence of input forms (e.g., in
www.indeed.co.uk where you can add the job you are looking for), that
are considered like urls by this function [1]. When the input form is
passed to org-make-link-string, org-make-link-string complains that the
link is empty.

Now I solved this by adding an if statement that makes the link only if
the info passed is not nil [2], but maybe someone knows a better check
for org-eww-goto-next-url-property-change than the one using 'shr-url?

Thanks a lot for org mode,

Andrea



[1] org-eww-copy-for-org-mode uses org-eww-goto-next-url-property-change
to get the position of the next link, which uses
(next-single-property-change (point) 'shr-url) which considers input
forms 'shr-url.

[2] this is the fixed version (the only change is highlighted by a FIXED
HERE comment):

#+BEGIN_SRC elisp
(defun org-eww-copy-for-org-mode ()
  "Copy current buffer content or active region with `org-mode' style links.
This will encode `link-title' and `link-location' with
`org-make-link-string', and insert the transformed test into the kill ring,
so that it can be yanked into an Org mode buffer with links working correctly.

Further lines starting with a star get quoted with a comma to keep
the structure of the Org file."
  (interactive)
  (let* ((regionp (org-region-active-p))
 (transform-start (point-min))
 (transform-end (point-max))
 return-content
 link-location link-title
 temp-position out-bound)
(when regionp
  (setq transform-start (region-beginning))
  (setq transform-end (region-end))
  ;; Deactivate mark if current mark is activate.
  (when (fboundp 'deactivate-mark) (deactivate-mark)))
(message "Transforming links...")
(save-excursion
  (goto-char transform-start)
  (while (and (not out-bound)   ; still inside region to copy
  (org-eww-has-further-url-property-change-p)) ; there is a 
next link
;; Store current point before jump next anchor.
(setq temp-position (point))
;; Move to next anchor when current point is not at anchor.
(or (org-eww-url-below-point)
(org-eww-goto-next-url-property-change))
(cl-assert
 (org-eww-url-below-point) t
 "program logic error: point must have an url below but it hasn't")
(if (<= (point) transform-end) ; if point is inside transform bound
(progn
  ;; Get content between two links.
  (when (< temp-position (point))
(setq return-content (concat return-content
 (buffer-substring
  temp-position (point)
  ;; Get link location at current point.
  (setq link-location (org-eww-url-below-point))
  ;; Get link title at current point.
  (setq link-title
(buffer-substring
 (point)
 (org-eww-goto-next-url-property-change)))
  ;; concat `org-mode' style url to `return-content'.
  (setq return-content (if (and ;; FIXED HERE added if statement
(if (stringp link-location) 
(string-blank-p link-location) nil)
(string-blank-p link-title)
)
   (concat return-content
   (org-make-link-string
link-location link-title))
   return-content   
   
   )))
  (goto-char temp-position) ; reset point before jump next anchor
  (setq out-bound t); for break out `while' loop
  ))
  ;; Add the rest until end of the region to be copied.
  (when (< (point) transform-end)
(setq return-content
  (concat return-content
  (buffer-substring (point) transform-end
  ;; Quote lines starting with *.
  (org-kill-new (replace-regexp-in-string "^\\*" ",*" return-content))
  (message "Transforming links...done, use C-y to insert text into Org mode 
file"
#+END_SRC





Emacs  : GNU Emacs 25.1.1 (x86_64-unknown-linux-gnu, GTK+ Version 2.24.31)
 of 2017-02-03
Package: Org mode version 9.0.4 (9.0.4-elpaplus @ 
/home/andrea/.emacs.d/elpa/org-plus-contrib-20170124/)

current state:
==
(setq
 org-capture-prepare-finalize-

Re: [O] Release 9.0.5

2017-02-11 Thread Fabrice Popineau
Thanks Nicolas for the confirmation.

I am running Emacs 25.2rc2 under Windows 10 compiled with Mingw64.
2 of the failures seem to be harmless.

This one is about the file system which behaves differently under Windows :

F test-org-export/file-uri
Test ‘org-export-file-uri’ specifications.
(ert-test-failed
 ((should
   (equal "file:///local.org"
 (org-export-file-uri "/local.org")))
  :form
  (equal "file:///local.org" "file://d:/local.org")
  :value nil :explanation
  (arrays-of-different-length 17 19 "file:///local.org" "file://d:/
local.org" first-mismatch-at 7)))

The absolute path has the drive name. So actually not a failure.

F test-org-pcomplete/keyword
Test keyword and block completion.
(ert-test-failed
 ((should
   (equal "#+startup: "
 (org-test-with-temp-text "#+start"
(pcomplete)
(buffer-string
  :form
  (equal "#+startup: " "#+STARTUP: ")
  :value nil :explanation
  (array-elt 2
(different-atoms
 (115 "#x73" "?s")
 (83 "#x53" "?S")

This one is because pcomplete is used to get completion, and the completion
depends on
the pcomplete-ignore-case variable, which makes pcomplete ignore case for
windows-nt and cygwin by default.

The last one is strange:

F test-org-element/link-parser
Test ‘link’ parser.
(ert-test-failed
 ((should
   (equal "id"
 (org-test-with-temp-text "[[id:]]"
(org-element-property :type ...
  :form
  (equal "id" "fuzzy")
  :value nil :explanation
  (arrays-of-different-length 2 5 "id" "fuzzy" first-mismatch-at 0)))

This is complete non sense, because when I evaluate the test form, I
actually get "id" and I have
no idea where this "fuzzy" may come from.

Do you also get a warning like this:
Making org-entities-user local to  *temp*-443480 while let-bound!
in the *Messages* buffer ?

Fabrice

2017-02-11 1:47 GMT+01:00 Nicolas Goaziou :

> Hello,
>
> Fabrice Popineau  writes:
>
> > Sorry for the noise.
> > I found the culprit and it was on my side.
> > Everything is working again
> > but there are still 3 unexpected failures to the tests.
>
> FWIW, i cannot reproduce it. Neither can the BuildBot.
>
> Regards,
>
> --
> Nicolas Goaziou
>



-- 
Fabrice Popineau
-
SUPELEC
Département Informatique
3, rue Joliot Curie
91192 Gif/Yvette Cedex
Tel direct : +33 (0) 169851950
Standard : +33 (0) 169851212
--


Re: [O] Feature request: lists with letters

2017-02-11 Thread Eric S Fraga
On Saturday, 11 Feb 2017 at 01:20, Nicolas Goaziou wrote:

[...]

> Again, I'm sure it is useful for some documents, but I don't think Org
> should provide such WYSIWYG feature.

Just to let you know that I completely support this statement.  If
people want a word processor, there are plenty about and I would not
want the effort of all those working on org to have their efforts
diluted in this direction.  I do, however, understand the desire for
such features so I am in no way denigrating those that wish for these
extra features.

Being able to customise back end exporters already provides an
incredible amount of tailoring to specific users' requirements.  

While I'm at it, Nicolas, thanks for all your work on org!

eric
-- 
: Eric S Fraga (0xFFFCF67D), Emacs 26.0.50.1, Org release_9.0.4-242-g2c27b8


signature.asc
Description: PGP signature


Re: [O] Release 9.0.5

2017-02-11 Thread Colin Baxter
On Fri, Feb 10 2017, Fabrice Popineau wrote:

> Maybe this explains it :
>
> Ran 670 tests, 667 results as expected, 3 unexpected (2017-02-10 
> 22:35:45+0100)
> 14 expected failures
>
> 3 unexpected results:
> FAILED test-org-element/link-parser
> FAILED test-org-export/file-uri
> FAILED test-org-pcomplete/keyword
>
> But it is still not a fix :-/
>
> Fabrice
>
> 2017-02-10 21:25 GMT+01:00 Fabrice Popineau :
>
>  Hi Bastien and list,
>
>  I have updated Org this morning from the master branch and I have a problem.
>  Inserting a link as the first element of a line result in a strange 
> behaviour. 
>  Sometimes, the link is not indented as it should relatively to the header 
> above, 
>  and the link seems to continue after the eol. Look at the printscreen :
>
> Images intégrées 1
>
>  I can't add a new line after the link.
>  Anybody has already seen something like this ?
>
>  Thanks for any clue in fixing this problem.
>
>  Fabrice
>
>  2017-02-10 14:53 GMT+01:00 Bastien :
>
>  Hi all,
>
>  Org 9.0.5, a minor bugfix release, is out.
>
>  Thanks!
>
>  --
>  Bastien
>
>  -- 
>  Fabrice Popineau
>  -
>  SUPELEC
>  Département Informatique
>  3, rue Joliot Curie
>  91192 Gif/Yvette Cedex
>  Tel direct : +33 (0) 169851950
>  Standard : +33 (0) 169851212
>  --

If it's any help, I get:

Ran 703 tests, 703 results as expected (2017-02-11 09:25:24+)
8 expected failures

with release_9.0.5-262-g0fe987.

In compiling, there is however the warning:

In org-create-formula-image:
org.el:19419:16:Warning: org-latex-make-preamble called with 3 arguments, but
accepts only 1-2


Colin.



Re: [O] Question about running the tests

2017-02-11 Thread Marco Wahl

> Perhaps the bundled org in stock emacs interfered with a newer
> downloaded version because load-path is not configured properly. If a
> mixed-version installation happened to Carsten, it is probably
> happening to lots of people. I also use a Mac and it's happened to me
> before. What else can be done to prevent this from affecting normal
> users? Would it be possible for org to do a quick self-consistency
> check on startup?

There could be done a lot, I guess.

But I think the situation currently is quite good.

There is http://www.randomsample.de:4457/waterfall which shows the
last build on a reference platform.

You can checkout a fresh orgmode source-tree and activate it with a
minimal configuration file.

If there are problems with an individually configured installation you
have no other choice than to somehow find out what is the culprit
AFAICS.  E.g. debugging and comparing to a reference.

Btw for the concrete issue in this thread about "ob-sh" and "sh" the
user-individual settings for BTEST_OB_LANGUAGES in the local*.mk
makefiles could be of interest.


Best regards

   Marco