Re: [PATCH] Improve configurability of ox-koma-letter

[Rasmus <rasmus@gmx.us>]

Viktor,

>> In a similar spirit to subject is firsthead.  First head is displayed
>> by default in scrlttr2 as far as I recall, which is annoying.
>
> Is firsthead something that you change on a letter-by-letter basis? Or
> do you configure it once for your letters and never change it? In the
> latter case the configuration can be moved of to an LCO file. In the
> former case an option would be nice. I had to actually change the
> options I introduced so far for different letters (e.g., I don't want
> foldmarks if I send the letter by email). 

How I use scrlttr2: I rarely use firsthead as I find ugly.  I use
firstfoot all of the time and I change it regularly.  E.g. for some
letters I include bank addresse.  Still, I see your point.

>> If we make foldmark an option should it not be a string accepting
>> values from "Table 4.3.: Combinable values for the configuration of
>> folding marks with option foldmarks" in the KOMA-Script manual?
>
> Same as above. If I read the scrguide correctly, the foldmark variable
> can be set multiple times, e.g.,
> #+BEGIN_SRC latex
> \KOMAoption{foldmarks}{blmtP}
> \KOMAoption{foldmarks}{true}
> #+END_SRC

This is true.

> The first line sets the behavior (and could be moved to an LCO file),
> the second line can be used to switch foldmarks on or off (and could be
> set by the exporter). Haven't tried this though.

Yeah, I guess it's true.  Still, since foldmarks depends on which
envelopes you have at hand it might make sense to have it accept a
string.  In lisp-terms a string is still t.  On the other hand the
current approach is consistent with your approach above so that's a
merit.


>> [...]
>> > +(defcustom org-koma-letter-use-subject "untitled"
>> > +  "Use the title as the letter's subject."
>> > +  :group 'org-export-koma-letter
>> > +  :type 'string)
>> 
>> Perhaps the default should depend on whether a title is present?
>
> The way I understand the org exporter, a title is always present. If not
> explicitly set using #+TITLE it defaults to the file name or the org
> heading.

Could be.  It seems you are right.  In the 'main' LaTeX exporter a
\maketitle is inserted only if title is not nil:

#+begin_src emacs-lisp
;; snip from ox-latex.el, Org-mode version 8.0-pre (release_8.0-pre-523-g7248fb @ /usr/share/emacs/site-lisp/org/)
     (format "\\title{%s}\n" title)
;;   [...]
     (org-element-normalize-string
      (cond ((string= "" title) nil)
	    ((not (stringp org-latex-title-command)) nil)
	    ((string-match "\\(?:[^%]\\|^\\)%s"
			   org-latex-title-command)
	     (format org-latex-title-command title))
	    (t org-latex-title-command)))
     ;; Table of contents.
#+end_src

>> Another approach (which I also somehow like) is to put it under the
>> LaTeX menu.  E.g. like beamer or like org-ravel ¹  
>
> I didn't know this was possible. I don't care either way, but I suppose
> an advantage would be that some commands could be then consolidated to
> save space.

I slightly prefer the "org-ravel" mode of adding it to LaTeX ("l").
It make sense in the way that Beamer is also there.  I don't know if
there exists an official opinion on such practices.
  
>> > +   (let ((with-place (plist-get info :with-place))
>> > +	 (place (plist-get info :place)))
>> > +     (when (or place (not with-place)) 
>> > +       (format "\\setkomavar{place}{%s}\n" (if with-place place ""))))
>> 
>> Wouldn't this work better: E.g. it wouldn't allow a place when place
>> is nil or when with-place is nil?
>> 
>> (when (and place (not with-place))
>>   (format "\\setkomavar{place}{%s}\n" place ))
>
> This will print the place if it is configured somewhere, even if 
> =place:nil= is set in #+OPTIONS.

I see.  I wasn't thinking of the possibility you outline below as I
would configure the variable with my default location (which
unfortunately isn't Berlin).

> Note that I use the empty string "" to
> surpress the space in my code. For example, I have in my LCO file
> #+BEGIN_SRC latex
> \setkomavar{place}{Berlin}
> #+END_SRC
>
> That is, the space is usually printed in every letter. 
>
> If I use #+OPTIONS: place:nil in a particular letter, the following code
> is emitted:
> #+BEGIN_SRC latex
> \LoadLetterOption{myletter.lco} % place is set here
> \setkomavar{place}{}            % this surpresses the printing of place
> #+END_SRC



> I would rather adapt the with-foldmarks option to accept a string rather than
> adding another foldmarks variable.

Agree.  I think it shold.  It's almost "free" to add since a string is
true in both and Emacs lisp sense and a scrlttr2 sense (in this case).

Thanks!

–Rasmus


-- 
And let me remind you also that moderation in the pursuit of justice
is no virtue